Chaire de recherche du Canada, Centre de neurosciences de la cognition,
Université du Québec à Montrèal
[Please see the powerpoints for this presentation at:
For the moment, I would ask you to use an
intuitive definition of the research
cycle, which I will explain in detail later.
archiving one’s research output in an institutional archive provides open access
by definition. If the 500
000 authors of articles that appear each year in the
scientific journals around the world were to archive their material, we would have
open access by definition. Self
archiving maximises and accelerates research impact,
and hence research productivity, progress and its rewards.
I would like to ta
ke this opportunity to acknowledge the work of Hélène Bosc, who
was one of the first persons to have had a vision with relation to open access. It is
thanks to her that we are all here today.
The open access initiatives are not aimed at comp
eting with or replacing publishers.
They constitute a parallel movement concerned with access and not with publishing.
There is no question of using open access to ruin or replace publishers, and those who
take this line have not understood the underlyi
Second, open access is not aimed at resolving the serials crisis faced by libraries,
although this may prove a by
product of the initiatives. The primary goal is to assist
researchers maximise the access to and the impact of their own re
search, not to
resolve the budgetary problems of libraries.
Third, open access not aimed particularly at providing access for teachers, students
and the general public. Of course, this will be a side effect but it will not of itself
provide the rationale
for engaging researchers in these initiatives.
Fourth, open access initiatives are not aimed at providing access for the Third World.
Again, this will be a side effect, but we cannot present the movement as motivated
primarily by this.
What, then, is open access aimed at?
The Classic Cycle
I will begin by describing the existing research
impact cycle which is based on toll
based access to material supplied by publishers. This limited research access causes
imited research impact, with impact being the main concern of researchers.
A complete cycle takes about 12 to 18 months. It begins with the conduct of the
research, which is based on the prior literature current research and can take months
or years to
complete. The researcher then drafts the “preprint” which is submitted to a
journal. The preprint is refereed by the experts selected by the journal editor in a
process known as peer review. The referees make corrections and suggestions for
t, or they recommend rejecting the article. If the revised draft is
accepted it becomes the postprint, certified as having reached the level of standards of
that journal. This will then re
launch the next cycle.
The Future Cycle
ill now describe the cycle in a system where all research is accessible to all
potential users, anywhere and at any time. First the preprint is self
Eventually the postprint too is self
archived. Access is maximised through open
quently, work’s impact is maximised and accelerated, creating more
impact cycles, sooner.
Open Access and Research Impact
The argument that open access leads to greater research impact is based on a number
of studies. For example, the Lawrence study o
showed that for
equivalent articles available by open
access compared to toll
access, the impac
increased by 336% on average. We can even go further. The Research Assessment
Exercise in the UK assesses and ranks all departments of all universities every four
years. It is a highly complex process, leading to scores that form the basis of
dies provided to the relevant departments. In psychology, for example, the RAE
scores can be predicted very closely using the citation impact of each researcher’s
articles. The system would be much simpler, richer and more sensitive if it was made
omprehensive if the full texts and reference lists of all those articles were
and hence assessable
Why do we want to maximise scientific impact? First, scientific impact indicates the
level of importance of a scientific
contribution. Second, it generates funding for
research. Third, it contributes to the prestige and overheads of institutions. Fourth, it
provides impetus to the careers of researchers, and their wage levels. Five, it leads to
progress in science.
e quality level of the journal where an article is published is already an index of
impact. The citation rate of the journal and of the article itself is also an indicator.
citations can also be included: who is cited with this particular author? Cit
rates for preprints are also of significance. Citations can be measured, as can usage
rates, through “webometrics.” Temporal analyses can be conducted to see which
indicators have an impact earlier or later in the cycle. Data are available on all
You may find the following websites of use in this area:
Harnad, S. (1990) Scholarly Skywriting and the Prepublication Continuum of
Scientific Inquiry. Psychological Science 1: 342
343 (reprinted in
Current Contents 45: 9
13, November 11 1991).
Harnad, S. (1994) A Subversive Proposal. In: Ann Okerson & James
O'Donnell (Eds.) Scholarly Journals at the Crossroads: A Subversive
Proposal for Electronic Publi
shing. Washington, DC., Association of
Research Libraries, June 1995.
Harnad, S. (2001) For Whom the Gate Tolls? How and Why to Free the
rch Literature Online Through Author/Institution Self
Harnad, S. (2001) Research access, impact and assessmen
1487: p. 16.
Question and Answer Session
Pieter BOLMAN (Elsevier)
Contrary to presentations you have made in the past, you have confined yourself
today to the essential points. However, you also note that there are side effects, such
g the library crisis or competing with the publishing business.
I did not say that competing with publishers was a side effect.
Possible side effects could be that the publishing business is undermined. We
discussed this po
I would not describe this as undermining the publishing industry. Rather, I would
say it is changing or evolving the publishing industry in a positive direction.
You use the term evolved. I would say that i
t is an evolution directed by you.
Not by me, but by the research community.
That may be the case but I am not sure that you speak on behalf of the entire research
community. (However, it should be noted that when the side
effects of a drug are
found to be very negative, the drug may have to be withdrawn. Perhaps you could
comment on this later. ) The second point I would like to make is that the importance
of the impact is questionable. The Lawrence article does not pro
ve that open access is
necessarily beneficial for impact. A system such as the one in Ohio where licensing
systems have been developed and accepted, and everyone has paid access, has
absolutely the same effect.
You are quite right that if
every researcher in the world were somehow granted
prepaid access to all refereed research, then everything would be fine. However,
there are at least 20
000 peer reviewed journals on the planet, and not even the richest
institutions can afford subscrip
tions to more than a minor subset of those 20
journals. Therefore, prepaid access is
universal, nor is it a realustic option (is
there to be a global click
through oligopoly licensing the planet?). Thus, from the
point of view of the author of
an article, today, the universal paid
simply does not exist. What is available to the researcher today? Rather than waiting
for an economic model to be developed that would perhaps allow everyone to have
prepaid access someday/somehow,
researchers can already provide this open access
right now. Every year that is lost represents a year of lost impact, that is, a year of
lost prizes and research support. We have already lost ten years in this context.
Given time, I beli
eve that publishers will succeed in designing licensing schemes that
allow everyone access. Open access is very much in its initial stages and many
people do not practise it.
That is correct. As I noted yesterday, the villains and the vic
tims in this area are the
research community. For the past ten years, they have had the means to provide open
access but have not yet chosen to do so. However, we are working on this. That is
impact measurement is so important in showing r
esearchers the causal
connection between access and impact and progress in research.
You mentioned the possibility of using bibliometry to assess research. It is well
known that when we measure something, we also influence it. Have we an
effects on the research community of using bibliometric methods in making funding
decisions? This could have very significant side effects on our community.
I did not say that bibliometry could be used to measure the quality of re
only its impact. In our scientific culture, impact has a very significant effect on
results and scientific progress as a whole. In the current system, impact is also a
significant factor in rewarding researchers and in the promotion of resear
ch.. It would
be completely illogical not to increase impact when we have the capacity to do so.
Of course, abuse is possible and the system can have negative effects. However, our
aim is to maximise impact and not to change the current system of asses
Marc MINON, Liège University
In order to promote my own laboratory, I work with a French language publishing
company in the human sciences. Given that they represent half of the titles present in
French universities, it is regrettable that they a
re not present today. Publishers should
not be perceived as predators. On the contrary, researchers solicit them to publish
their articles. For years, we controlled the peer review process, printing and
dissemination of articles. Publishers do not acce
pt to publish articles for economic
reasons given that the margins generated by these journals are minimal. Without
public subsidies, these publications would not exist.
These publishers are currently concerned by the fact that French language human
ence journals do not have a significant presence in the networks. Researchers are
thus asking that these journals be placed online in order to increase access and impact.
To this end, there are five possibilities and I would ask you to advise on which of
these five is preferable. First, the moving wall option. Second, placing articles online
for free, subsidised by public authorities. Third, require authors to pay to have their
article published. Fourth, modify contracts with authors, authorising them
archive. Should subscriptions drop by 10%, publication will be stopped. Fifth,
consider that scientific journals are now obsolete. From now on, they will no longer
be published online and authors will be responsible for publication themselves.
As I stated earlier, the aim of open access is not to replace publishers. At the same
time, our role is not to protect the economic viability of small journals, which has
always been problematic. I would definitely not recommend that you
journals by placing them on open access. We have to allow researchers to self
archive. Even if you are not in favour of this, you will not be able to prevent it
occurring. It is possible that this could lead to some journals going out of
but there are solutions to this. I cannot predict what is going to happen in the future.
However, open access is the optimum solution and it is possible today.
From the floor
It was stated yesterday that scientific results are rediscovered e
very five years. A
Wisconsin study showed that open software was of better quality than traditional
software. I have been hearing about the benefits of open access for years now, and
yet we still do not have a system of open access.
nalogy with free software is not a pertinent one as those who write free
software are paid for their work. In contrast, researchers are not paid for placing their
From the floor
I was impressed by the presentation on the work being carried o
ut to digitise our
mathematical heritage. However, I note that all of the institutions you mentioned are
located in either the US or Europe. My understanding of our mathematical heritage is
that it comes from China, India, Persia and Egypt. To what exte
nt are those
mathematical works being digitised?
Digitisation centres are currently located in the Europe and the US. However, work is
also being carried out in China and South America, for example, and there is no
intention to exclude
any area from this work. It is merely a question of time.
Christine DESCHAMPS, President, IFLA
Given the interest raised in the previous discussions, we have decided to continue the
debate through the Conference’s website. To this end, we will establish
a list of
discussions in which you can participate.
Question and Answer Session
Olivier BERGER, ENT
Is the conceptual fingerprint technology that you discussed a patented technology or
will it be freely available to researchers?
the technology is patented, that is, the actual structure of the fingerprint. The
rest of the technology is more or less open source, as it is based on very standard
technology in the text retrieval analysis fields. Collexis is very willing to work
her with academic organisations and encourages people who use the technology
to place their fingerprints in the public domain.
In order to set up one of these systems, it is necessary to have a licence to use the
abstraction and fingerprint generating so
ftware. Collexis wants to open up research as
much as possible and academic users would only need to pay a very low cost licence.
How do your projects relate to open access? It is not clear to me who the customer is
for Figaro, nor what t
he relevant product is. In relation to E
BioSci, I understood that
this was a search engine for proprietary and open access databases. In this context,
what does “free to the user” mean?
Figaro customers are academic e
Figaro back office will remain in
the background and invisible to the actual authors and users of scientific content. We
want to facilitate publishing front offices, enabling people to carry out self
It promotes technology such as Eprint that
can be bought on the shelf so that people
do not need to reinvent the wheel.
In terms of the value of a new search engine to the open archive movement,
independent archives often have problems with the quality of metadata and cross
ity. We offer biology archives the ability to apply fingerprint technology to
the contents of their archives. Furthermore, we improve connectivity by using
sophisticated concept based search technology. The engine remains within E
which stores f
ingerprints generated by different archives and allows them to connect
into related resources, taking people to the correct archive or database.
From the floor
You discussed the shortcomings of URNs and DOI. What are the solutions to this
issue? Why di
d you not consider the URN
We are considering URN
Seki. I only wanted to warn you against the potential
dangers of DOI.
From the floor
I am grateful that you mentioned the many possibilities for re
use of th
which is an important part of what open access is about. Embedding literature in
databases allows us to carry out meta
analyses etc. I have been very impressed by my
experience of the fingerprint technology you presented. I understood that
was free for public or government funded organisations.
Collexis grew out of an EU project, Share, that was aimed at developing health
information networks within developing countries. They provide a percentage of their
to projects focused on that goal, and provide their technology free to Class I
countries. I am not sure of the situation with respect to government organisations.
Paul DUCASSE, Lyon University
In a French language project, we developed a platform th
at accepts, respects and
produces valid documents using DOI. The goal is to provide scientific institutions
with tools for archiving and disseminating documents. In this context, we also
provide an indexing tool. What we really need is to have all of th
ese initiatives come
together to avoid new systems being re
created every six months.
In this context, I suggest you come to the Open Office meeting in Hamburg this year.
From the floor
I hope that you will be considering “Shibboleth” f
or access control. In relation to D
Space, I have some concerns in relation to persistent identifiers for online resources.
What will be the licensing arrangements for other parts of the E
BioSci system? Do
you have any other commercial partners working
Ingenta is a partner, and is building the interface to the system.
We have considered “Shibboleth” and are in contact with Ingenta.
From the floor
SPARC supports open access, and its flagship publication is “Organ
published by the American Chemicals Society. When will it become an open access
Hopefully, as soon as possible. The world is moving very quickly and open access
and institutional repository concepts have only taken o
ff in the last few years. The
rival journal idea was a good one initially. Given that the technology and tools are
changing, it would be foolish for any organisation to not take advantage of
developments in order to best serve the scientific community.
From the floor
You highlighted a very important point: what is even more important than access to
archives is open access to databases. What actions are you taking in relation to
Having looked at search engines and text mining tec
hnologies in general, we found
the Collexis technology to be the best in terms of price and quality. We therefore
decided to enter into a collaboration with that company on the understanding that the
use of that tool in an academic setting would be absolu
tely free. We hope to sustain
our activities by charging commercial organisations for use of the system. Given that
the technology is evolving at a fast pace, we are prepared to change to a new system
in the next few years.
n concerns identifying and pointing to elements in databases. I recently
read of a proposal to use EOI for datasets. I would be strongly opposed to this.
In relation to journals supported by the OSI, it is very clear that in order to
their value worldwide, we have to improve the value of the publication by promoting
systematic links between documents and databases. This enables researchers to
access the relevant trials directly. Another option is to review the system of peer
From the floor
This illustrates the problem related to open access for databases. ACS’s attitude is
based on the fact that they have a very expensive database. Should they provide this
on open access, all chemists, including commercial chemists
, would have access to it.
What are your views on the evolution between the various foundations: the Soros
Foundation was the first one to support open access and has now been followed by
many other foundations?
In the US and Canada,
the issue of the public domain is of growing interest, due to a
background of encroaching private interests. People are now very wary of the trends.
The conference held at Duke University in Autumn 2001 resulted in a series of very
interesting articles o
n protecting the public domain, which is a guarantee of
biodiversity and intellectual diversity. In order to find these papers, use “duke” and
“benkler” on the Google search engine.