Re: Role of arXiv

From: Stevan Harnad <amsciforum_at_GMAIL.COM>
Date: Mon, 18 Oct 2010 23:26:29 -0400

On Mon, Oct 18, 2010 at 10:33 PM, agentilb Gentil-Beccot
<anne.gentil-beccot_at_cern.ch> wrote:
> On Oct 15, 2010, at 3:10 AM, Sandy Thatcher wrote:
>
>> ST:
>> Anticipating Stevan's response, I'd ask how these disciplinary
>> repositories are to be filled since they lack the power that
>> universities have to mandate submissions?
>
> Without the power of mandates, successful subject repositories
> instead wield the immense power of incentives to attract
> scholars.

Unmandated subject repositories are just as empty of their target
content as unmandated institutional repositories are. The illusion
that they are not comes from the denominator fallacy: Yes, an
unmandated repository for all of, say, biology, from all institutions,
will have more content than a single unmandated institutional
repository (across all of its output subjects). But the denominator
from which the deposit rate for the institution is derived is the
institution's total annual output (across all subjects).

The denominator for the subject repository is all the output in that
subject *from all institutions*.

Prediction: Unmandated, the percentage will be the same for both: about 15%.

There is one known, longstanding (since 1991) exception, Arxiv, and
this is because of a habit of sharing preprints that HEP physicists
and astro physicists had had and had exercised since before the
Internet. Yes, when the Internet and then the web were created, these
subfields flocked to it, and their output is virtually 100% for years
now. But the trouble is, that's not happening in any other field
(though computer science and economics come close). Most fields are
still at 15%, almost two decades after the possibility of OA was first
demonstrated by the HEP and astro physicists.

The issue is not central vs. institutional locus of deposit. It's
getting paralyzed authors to deposit at all. And for that, the
institutional locus is the one that is in a position to mandate
deposit. Not all research is funded, but virtually all research comes
from institutions. Convergent, one-time deposit (for non-depositors)
is certainly easier to mandate than divergent, multiple deposit. (How
many subjects -- hence subject repositories -- might a piece of work
fall into?

All the benefits of central search are there with central harvesting;
but institutions are the places to deposit, once. They can mandate it.
They have no interest in mandating multiple, institution-external
deposit. And funders should help the process converge, systematically,
rather than forcing it to diverge, by mandating institution-external
deposit. That should be accomplished by harvesting, *after* 100%
depositing has been ensured, by convergent institutional and funder
mandates.

> Subject repositories do something crucial for
> scholars: they enable them to communicate faster, they make their
> work more visible to their peers, and they give their work more
> impact.

And you have them with local depositing and central harvesting.

But what we lack is enough depositing anywhere at all. That's what
mandates are for. And central repositories, as noted, cannot mandate.
And (except for Arxiv, in some subfields), central repositories don't
get more than the usual baseline 15% (except when mandated by
funders).

> This is epecially the case when large-scale digital
> libraries integrate the repository content with the scholarly
> record. We have done a case study of arXiv.org for High- Energy
> Physics, in symbiosis with the SPIRES/INSPIRE service (which
> incidentally we launched earlier this week: see
> https://listserv.nd.edu/cgi-bin/wa?A2=ind1010&L=pamnet&D=1&T=0&O=D&P=37998
> inspirebeta.net and projecthepinspire.net)

None of that has anything to do with locus of deposit.

> Our results could be of interest in this conversation,
> spotlighting motivations for scientists (whose behavior in this
> respect is known to be relatively insensitive to mandates)
> http://arxiv.org/abs/0906.5418 (apologies for those who already
> spotted it on this list in the past).

SUMMARY: Evidence confirming that OA increases impact will not be
sufficient to induce enough researchers to provide OA; only mandates
from their institutions and funders can ensure that. HEP researchers
continue to submit their papers to peer-reviewed journals, as they
always did, depositing both their unrefereed preprints and their
refereed postprints. None of that has changed. In fields like HEP and
astrophysics, the journal affordability/accessibility problem is not
as great as in many other fields, where it the HEP Early Access impact
advantage translates into the OA impact advantage itself. Almost no
one has ever argued that Gold OA provides a greater OA advantage than
Green OA.The OA advantage is the OA advantage, whether Green or Gold.

> We proved that clear
> advantages exist in the speed of communication and impact of work
> in subject repositories, to the point that subject repositories
> become the (sole) location for scholarly discourse.  Eventually,
> scientists with these field-specific tools rarely read journals
> but acquire most information directly from these resources.

It is uncontested that OA content will be consulted jointly, not
institution by institution. It is a historic fact that physicists
sharing preprints began with central deposit and have been doing it
ever since. But the habit has not generalized, and that is the
problem. And institutional repositories and mandates are the solution.

None of the benefits of central collections are based on a need for
central locus of deposit.

And, as noted, regardless of where researchers are consulting refereed
research (and, a fortiori, regardless of their locus of deposit), it
is still the journals that are implementing the refereeing today.

Stevan Harnad

>
> Kind regards,
> Anne Gentil-Beccot (CERN - Serials Librarian)
> Travis Brooks (SLAC/Stanford - INSPIRE Director of Operations)
> Salvatore Mele (CERN - Head of Open Access & INSPIRE Strategic
> Director)
>
>
Received on Tue Oct 19 2010 - 04:27:06 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:50:16 GMT