SSHRC Open Access Consultation (Canada)

From: Stevan Harnad <>
Date: Mon, 22 Aug 2005 22:07:58 +0100

I have been asked to advertise widely the following survey on Open
Access by the Canadian research funder, the Social Sciences and
Humanities Research Council (SSHRC):

I am happy to do so, but not without advertising my own replies, which
contain an implicit critique of the confused and confusing way in which
some of the questions are being asked:

> Consultation on Open Access: Questions
> SSHRC formally adopted the principle of open access at
> the October 2004 meeting of its governing council.

Please define the "principle of open access" to make sure we are
referring to the same thing. To me it means making the full-text of
peer-reviewed research freely accessible on the web.

> However, questions remain about how to make this principle
> operational, how to revise current research
> dissemination and communication policies, and how to
> reshape research support programs to meet the needs
> of researchers within this new policy context.

Only one simple revision is needed, and the Research Councils UK are
already about to do it: Require the full-text of all peer-reviewed
research resulting from SSHRC-funded research to be made freely
accessible on the web:

> Policy Questions
> SSHRC requests your advice on the following general
> policy issues:
> Should SSHRC adopt a regulation requiring that one
> copy of all research results be deposited in an
> institutional repository?

Yes, definitely -- and immediately upon acceptance for publication, not
3, 6, 12 months or longer after publication; any embargo on research
findings is a gratuitous embargo on research usage, progress and impact,
with no research-based justification whatsoever.

> Should such a regulation apply to all forms of
> research outputs (i.e. peer-reviewed journal articles,
> non-peer reviewed research reports, monographs,
> data sets, theses, conference proceedings, etc.)?

All those that the author would otherwise have made public and would
not have sought revenue for its sale. SSHRC cannot and should not try
to force authors to make public what they do not wish to make public
(or wish to make public only to sell), but SSHRC *can* certainly make
publication of the findings a condition for funding the research
("publish or perish").

> Should there be exceptions for research outputs where
> there is an expectation of financial return to the author (i.e.,
> monographs where royalties are accrued)?

Yes, definitely. One step at a time! 85% of the refereed literature
is still not freely accessible online, and there the case for it is
incontrovertible, because it is all an author give-away, written solely
for research usage and impact. This is not the time to get bogged down
in gray areas: Encourage making monographs freely accessible online,
but don't *require* it, and certainly not now, before the primary target
itself (peer-reviewed research) has been made 100% freely accessible.

> Operational Questions
> In general, there are two accepted routes to open access:
> Self-archiving - depositing research results and
> materials in institutional repositories that can be
> searched by anyone with Internet access; and,
> Open access electronic journals - peer-reviewed
> journals that provide Internet-based access for readers
> without subscription charges.
> Both routes present SSHRC and the research
> community with operational challenges:
> Institutional repositories: Building a management
> and service platform

This presents no "operational challenge." SSHRC need do nothing except
provide a backup repository of its own for those researchers who don't yet
have an institutional archive. It costs $2000 plus 3 days sysad set-up
time and 3 days a year sysad maintenance time to for an institution or
SSHRC to create and maintain an archive. The rest is just getting the
researchers to do the keystrokes required to deposit their contents in it.

> Currently, not all Canadian universities provide an
> institutional repository service. Some 26 repositories
> are now in place, or are in development, but this does
> not yet provide the necessary services for all SSHRC-funded
> researchers.

What is missing is not repositories (which, as noted, are cheap and
trivial to create and maintain). What is missing is *policies* requiring
that they be filled. The 26 Canadian repositories are near empty
(relative to their own annual research output)..
They are not missing dollars, they are missing keystrokes, and the
keystrokes need to be required - by the SSHRC, amongst others.

> If required by SSHRC, would you be willing to send all
> outputs from SSHRC-funded research to an institutional
> repository?

Yes, I already do. And two international JISC studies confirm that 81%
of researchers would deposit willingly if required by their employer
and/or funder (14% more would comply reluctantly, and only 5% would not
comply) :
And the results from the (only) two institutions so far that already
have required self-archiving (Southampton ECS and CERN) confirm this,
with >90% compliance in both.

> What range of electronic publications and institutional
> repository services are needed to fully meet the needs of
> the scholarly community? See, for example É
> (, a Quebec-based electronic service
> provider. Should this model be extended across Canada?

Why is the question of providing institutional repository services
(which costs little, especially per paper, and requires only an SSHRC
policy) being conflated with the question of providing electronic

Most journals have an electronic edition these days, but
very few journals are freely accessible (about 5% of journals are). The
cost per paper of publishing in that 5% of journals may vary from
$500-$3000 per paper, if they are not subscription-based. And subsidising
the journals may cost even more.

But why are these questions being mixed up with the (minimal) cost of
SSHRC-required self-archiving, which can already generate > 90%

> Open access journals: Revising the SSHRC Aid to
> Research and Transfer Journals Program
> Although SSHRC financially supports the majority of social
> science and humanities journals produced in Canada, the Aid to
> Research and Transfer Journals Program does not provide support
> for non-subscription based journals.

It's not clear why SSHRC supports any journals, but surely if they do,
they should want them to be as widely accessible as possible?

> Scholarly peer-reviewed journals play a crucial role in
> the certification of research knowledge. In the context of
> open access, institutional repositories must be able to
> distinguish between peer-reviewed and non-> peer-reviewed
> research outputs. Therefore, the continued existence, and
> financial viability, of journals is clearly a critical
> issue.

Again, one has nothing to do with the other! SSHRC-mandated
self-archiving is the self-archiving of peer-reviewed journal articles.
Self-archiving has not generated any subscription cancellations, even in
the fields (e.g., physics) where it is most advanced and has been going
on longest. Why are the issues of self-archiving (and its minimal costs)
and journal publishing (and its costs) being conflated here?

> Please comment on each of the three following possible
> ways to tackle this challenge, taking into consideration
> the fact that there are limited resources for the support
> of research:
> A "moving wall" system where journal articles are
> available only by subscription for the first six
> months, and then made available free of charge.

An awful system, throwing away 6 months of research usage, progress and
impact for no reason whatsoever.

> A publication fee, charged by journals to authors,
> to be considered an eligible expense within a SSHRC
> research grant. This would require researchers to
> have access to SSHRC or other grant funds.

Fine, but this is the solution only to the problem of Open Access
for a small percentage of Canadian research output -- the output that
is published in Open Access journals (including the SSHRC-subsidised
ones, if SSHRC decides to go this route). What percentage of Canadian
research output do you think that covers? In the world, 5% of journals
are currently OA. In addition, about 15% of research output is already
self-archived spontaneously by its authors. That leaves 80% of research
output still to be accounted for -- and only a self-archiving mandate
can take care of that.

> A modification to the SSHRC support program for
> journals -- which currently covers 40 to 50 per cent of
> journal expenditures?to allow grants to cover all
> peer review, administration and manuscript
> preparation costs, but not costs associated with
> distribution.

These are minor details, concerning perhaps 20% of Canadian SSHRC
research output. What about the other 80%, published in international
journals, not OA journals, and not supported by SSHRC? (Like worrying
about monographs, this too is the tail wagging the dog: Focus on the 80%
of Canadian SSHRC research output that this does not cover!)

> As journal editors, do you allow your contributing authors
> to place their accepted articles in an institutional
> repository or on a Web site not connected with the
> journal? Why, or why not?

As a Cambridge University Press journal editor (of Behavioral and Brain
Sciences) for 25 years, I certainly did (from the moment the Net came
into existence)

And over 90% of journals worldwide now do too:
That's not the point either. The problem is not the lack of journal policy
allowing it, but the lack of institution/funder policy requiring it.

> As researchers/authors, would you be willing to comply
> with a SSHRC regulation that requires peer-reviewed
> articles to be published in an open access journal and/or
> placed in a publicly-accessible > institutional repository?

Did you not just ask half this question already above? "If required by
SSHRC, would you be willing to send all outputs from SSHRC-funded
research to an institutional repository?"

You are now asking it in an illogical and/or form (worthy of a dodgy
secession referendum!) :

If you ask me separately whether I or any other author would tolerate
being required to publish in any journal other than the one I judge
optimal for my work, the answer is definitely not.

Why conflate an open-and-shut yes-question with an open-and-shut no

Stevan Harnad
Received on Mon Aug 22 2005 - 22:28:32 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:58 GMT