Re: Guide juridique CNRS

From: Stevan Harnad <>
Date: Tue, 29 Nov 2005 23:48:03 +0000

Another knowledgeable colleague from France has sent this. I will
paraphrase with some interpretation:

    << Nous avançons beaucoup au niveau national vers la définition d'un
    réservoir unique de nos publications qui sera le vecteur d'une
    politique commune, dans le domaine des archives ouvertes, à tous les
    établissements français (avec le soutien de l'académie des sciences
    ;-)). Pas à pas, nous préparons notre communauté de recherche à
    utiliser HAL dans leurs pratiques courantes, comme tu le constateras
    sur le site du CNRS: Les guides
    légaux sont nécessaires pour accompagner une telle politique de
    déploiement à grande échelle, afin de nous prémunir contre l'effet
    "Royal Society" que vous avez observé récemment en Grande Bretagne. >>

Apparently France -- a country highly centralised since Napoleonic days,
with its unique national research institutions, of which the CNRS is the
biggest -- is in a structural and functional position to do something,
at a national level, at one stroke, that no other country is quite
in the position to do:

The idea is to create one national OA archive for all disciplines and
all French national research institutions and French universities to
deposit all their research output into: HAL (Hyper Articles on Line)

This unified national effort is meant to avoid the kinds of divisiveness
that in the UK led to a self-archiving recommendation by a parliamentary
committee, subsequently rejected by the government, subsequently taken
up again by the 8 UK Research Councils (RCUK) and now opposed by the
Royal Society (Britain's academy of science). In France, there is now
the hope that not only will the big national research institutes (CNRS,
INSERM, INRA, INRIA) unite in this national enterprise, but so will the
French Academy of Sciences, as well as all the French Universities.

If it succeeds, it will be an enormous coup for France, and a terrific
direction-setter for the rest of the world -- though probably no country
would be in a position to emulate it exactly, for lack of corresponding
centralised national institutions.

The test of the success of the French national OA initiative will of
course be the rate at which France approaches 100% OA in all disciplines,
in all institutions. The world average for that now hovers somewhere
around 15%. So to claim success, France will not just have to create a
central archive (HAL) and invite all French researchers to self-archive
in it. We already know that creating archives -- whether many small,
distributed local institutional ones, or big central ones -- is
definitely not enough. The archives remain near-empty in the absence
of a self-archiving mandate. It is over reaching agreement about a
self-archiving mandate that the various UK players have been progressing
so slowly.

To demonstrate success, France will have to fill its mega-archive,
and fill it at a rate significantly higher than the current
worldwide baseline rate of 15%, and its slow current annual rate
of growth.

To measure and document its success, France will first have to specify
very explcitily what its target population is. It may very well be every
research-active French scholar or scientist in every discipline at every
research institution and university in France. If so, it will be necessary
to estimate what the annual research output is for that total population,
either in terms of journal (and refereed conference-proceedings) articles,
or, if still more ambitious, also in terms of theses. Let us call that
(1) annual French research output FR.

The figures that will be needed are (2) the annual percentage of FR
deposited in HA (hence OA) and (3) the annual growth rate of that
annual OA percentage of FR. It will be interesting to see whether the
centralisation proves to be an asset or liability. It is impossible
to predict, because the case of France is so unique, but a UK study
on central vs. institutional self-archiving came up with a different
recommendation for the UK, suggesting that not only would distributed
local institutional and university archives, plus harvesting (including
central harvesting) be cheaper than a central archiving model in the era
of OAI-interoperability, but, even more important, the local institutional
incentives and culture are likely to be more conducive to archive-filling,
because of the shared interests and benefits between researchers and their
institutions. (On a larger scale there is less of a sense of identity
and common cause, and more competitiveness.)

    Swan, A., Needham, P., Probets, S., Muir, A., Oppenheim, C., O?Brien,
    A., Hardy, R., Rowland, F. and Brown, S. (2005) Developing a model
    for e-prints and open access journal content in UK further and higher
    education. Learned Publishing 18(1):pp. 25-40.

But who knows? No one has a winning formula for reaching 100% OA as
quickly and reliably as possible until a formula has been demonstrated
to win. Perhaps France's unique configuration makes a formula possible
there that would not work elsewhere. Moreover, we live in a virtual
world, and each institution's and university's sector of HAL could
be customised and given the look of, and put under the control of,
each local content-providing institution.

The real question is whether France's national OA mega-archive will
be backed up with a self-archiving mandate, or it will rely only on
voluntarism, for voluntarism has proved to be a very poor deliverer of
OA so far.

The benchmarks for measuring France's success are already in place:
They are the five institutions world-wide that already have a
self-archiving mandate in place, some of them already at or near 90%
annual OA: CERN, U. Southampton ECS, U. Minho, Queensland U. Technology,
and U. Zurich:

If France can match their growth rate, France wins, because it can
multiply its success by the number of French universities and research
institutions providing the content. But that grand total of RF is also a
handicap, because whereas HAL may already have a larger absolute number
of papers (40,000) than most other individual OA archives worldwide
its success must be reckoned as a percentage of RF, and that's a rather
more daunting target!

It is in the interests of us all to see France succeed in this
grand initiative, and we accordingly wish it every success, and with

Stevan Harnad
Received on Wed Nov 30 2005 - 00:10:36 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:07 GMT