RE: [EuroscienceOA] A personal view about institutional archives

From: Alma Swan <>
Date: Sun, 18 Dec 2005 12:25:56 -0000

Dear Colleagues,

In his discussion of Pierre's clear explanation of the situation in France,
Stevan refers to our study on the development of a national system for
delivering open access to research output in the United Kingdom and I would
like to add some points of my own. (Our report is at

In that study we recommended a model based on distributed (institutional)
repositories for two reasons. First, it represents the most efficient and
efficacious way to gather together content that may reside in a multitude of
different sites (our remit was to develop a model that would deliver *all*
UK open access research articles). Second, it works *with* what was proving
to be the most effective way of getting those articles deposited in the
first place, that is, by means of mandatory requirement to deposit. We knew
from a previous study of ours
.pdf) - and this was confirmed by a subsequent one - that researchers will
deposit articles when required to, almost without any resistance, but that
great numbers will not do so, or will deposit only patchily, if the matter
is simply voluntary. This is just human nature. Most people will not bother
to do something that takes time and effort if it is not actually required of
them, even if it is demonstrably to their benefit: we were all familiar with
the beneficial effects of wearing seatbelts while travelling in cars, but
few of us bothered to use them on every trip until a law required this

We also knew that although large amounts of UK research is externally
funded, large amounts are not. The Wellcome Trust found that 31% of research
articles in biomedicine cite no specific funding source at all; my own small
survey of articles in physics shows a similar finding, with 35% of articles
citing no specific funding source. We concluded that although funder
mandates could play a big part in achieving a good rate of article
deposition, we would never achieve 100% deposition relying on funders alone,
even if they were prepared to issue a mandate, as the Wellcome Trust
subsequently did.

Employers, in the form of institutions, however, *could* issue mandates that
would affect near-100% of UK researchers if all employing institutions
operated such a policy (the tiny fraction that is an exception being those
individuals who research and publish on an independent basis). Moreover, it
was clear that institutions have a significant amount to gain in
establishing a repository and requiring researchers to deposit all their
output in it. Such a repository provides a showcase for the institution that
can be used for various purposes, as well as providing the institution with
the means to easily and quickly gather data about outputs for their own
adminstrative purposes and for any research assessment purposes that are
required. In all, then, the likelihood that institutions will see these
advantages, establish repositories and mandate deposition seemed perfectly
possible and we recommended that this should form the basis of the UK-wide
model, with service providers - existing ones and new ones that will come
along - harvesting content from this network of institutional repositories
and presenting to the user (searcher) appropriate resource-discovery tools.

The *critical* issue that arises in discussing whether this or that model
might work best in one country or another is this: where is the most
effective locus of the mandate in that country? In the UK it is definitely
at institution/employer level. This may not be so in France, where
differences in the organisation of science funding and employment have been
explained by Pierre, nor in some other countries where similar, more
centralised arrangements operate.

I would still suggest, however, that there must be a good number of
researchers in France who do not come under the influence of CNRS, INRA,
INSERM, etc, and sooner or later some means has to be found to reach those
researchers, too, with a requirement to deposit their output on an open
access basis. Who will best do that? That is the fundamental question. If
the answer turns out to be that only employing institutions can do this,
then those institutions are the answer to how to achieve a complete OA
corpus for France's research output and it is at their level that deposition
should be organised. We must always remember is that there is no country on
the planet where researchers will deposit all their output without a mandate
from somebody. The gene for inertia is dominant at its locus in all parts of
the globe!

I completely understand the argument about researchers feeling 'comfortable'
with collections of articles in their own discipline or sub-discipline. This
is, however, a researcher-as-searcher issue, not a researcher-as-depositor
one, and should be viewed as a separate operation in the process of open
acess provision. It, too, was built into our model in that we envisaged a
layer of services provided across the open access corpus, harvesting
possibly the whole of the corpus, but more likely harvesting in certain
disciplines to provide open access collections in, say, linguistics or
immunology or medieval history. In time, researchers in disciplines with
such a service at their disposal will find it just as natural to use 'their'
service on a regular basis as the physicists who currently visit arXiv each
morning before they begin that day's search for the Higgs boson.

Alma Swan
Key Perspectives Ltd
Truro, UK

EuroscienceOA mailing list
Received on Sun Dec 18 2005 - 12:42:46 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:48:09 GMT