No, a self-archiving mandate will not split up Canada...

From: Stevan Harnad <>
Date: Sat, 30 Oct 2004 23:10:35 +0100 (BST)

I am ambivalent about intervening here, but it seems so unjust that Alma
Swan -- who is always so rigorous in her reasoning, fastidious about her
facts, sparing in her comments and productive in her empirical research
on OA -- should have her few cautious and well-founded contributions on
this list repaid by such an unconstrained and unrelenting series of
arbitrary opinions and speculations, escalating lately to apocalyptic
auguries about the split-up of Canada.

The moral of the story (and this is not addressed to Alma!) is that if
one has nothing to say, one should try saying nothing. There is no need
to pronounce on everything and the contrary of everything. There is
careful thinking and planning to do, and genuine and urgent work to done.

I say this knowing full well that I comment a lot too. I am ready, though,
to submit to the verdict of history, as to whether my own frequent but
focussed, convergent, and evidence-based comments and critiques are on
a par with the ceaseless scattershot opining with which it must compete
for attention, understanding, and constructive action in trying to reach
the long overdue goal of OA at last.

Here are some facts (and they are facts!):

For a university to create an eprint archive amounts to the cost of a
$1000 linux server (unless there is one, or part of one, available
already), plus a few days' sysad set-up time. Maintenance time is
negligible; a few hours a month.

So when readers weigh the speculations about "What if a university
doesn't have an archive?" or "What if a university won't create an
archive?" these are the real-world facts underlying the answer, and what
is (and isn't) really at stake.

Creating an institutional archive is trivial, and effectively cost-free.
The only non-trivial part is how to get the archive *filled.* But if we
prefer to speculate instead about imaginary obstacles (such as universities
not having archives) we can effectively avoid that question, the real
substantive question -- to which the answer is that archives are filled
by adopting effective archive-filling policies.

Two effective archive-filling policy proposals are on offer, one the UK's
and the other the US's. Both are simple mandates (and if implemented, they
will be sufficient to generate 100% OA, at 100% of universities that
are publishing any research at all). France, Germany, Switzerland,
Australia, Norway, Japan, India (and, yes, Canada!) are among the other
countries currently contemplating similar archive-filling policies,
-- and a few universities have already implemented them, not waiting to be
officially told to do so by their governments or their research funders.

Some commentators prefer instead to discuss, endlessly, hypothetical
obstacles, hypothetical national differences, discipline differences,
etc. Anyone is free to continue shadow-boxing with conjectures and
counter-conjectures for another 10 years, but I would urge those who are
interested in maximizing OA now to desist from debating hypotheses and
get to work on actually implementing archive-filling policies. There is
nothing hypothetical about that:

Now some comments:

On Fri, 29 Oct 2004, Heather Morrison wrote:

> If the Canadian federal government were to refuse to fund research that
> could not be deposited in an institutional repository, I remain
> convinced that there would be significant opposition.
> This is an area where I do have expertise, since I am Canadian (and
> Quebecoise), have lived in several provinces, and work for a provincial
> coordinating body for post-secondary libraries. ELN is extremely
> successful and highly regarded in leading collaborative projects. We
> mandate nothing, however - all collaboration here is purely voluntary.

Mandated self-archiving has absolutely nothing -- zero -- to do with
ELN (Electronic Library Network)! Libraries can neither provide OA nor
mandate it. Only authors can provide OA and only their employers and
research funders can mandate it. Libraries *could* help in implementing
an OA mandate (by helping to create and maintain archives, and helping
authors to fill them), but not if they prefer to speculate instead
about what is and is not possible, a priori, when none of it is in their
hands in the first place!

> I completely support the development of institutional repositories,
> incidentally. It's just that a federal mandate like you suggest isn't
> an approach I would recommend for Canada or for B.C. It may well be
> the best possible option in the U.K. or other countries; there, I do
> not have the expertise to judge.

Let us now remind ourselves of the *reasons* adduced for this
expert judgment about the unsuitability of a self-archiving
mandate for Canada and BC:

> Since most institutions do not have institutional repositories at
> present, a mandate requiring self-archiving in institutional
> repositories would, in most countries, greatly slow down open access.

This confident speculation about institutions, countries, self-archiving
and OA growth rate is so far from anything even faintly resembling
the facts that it is difficult to judge whether it makes more sense to
reply with the actual data, or just to leave it as an opinion opined!

I will reply once (but will withdraw if it only succeeds in eliciting
still more speculation by way of response):

    (1) Many institutions already have eprint archives,
    and their problem is that they are not being filled.

    (2) Most institutions do not yet have eprint archives, but they
    are only a $1000 linux server and a few days sysad time from having
    them. Then their problem too will be that they are not being filled.

    (3) This is true in all countries that produce research at all.

    (4) The annual open-access rate currently is about 5% via OA
    journal publishing and 15% via self-archiving.

    (5) This rate is far too slow, but it is hard to imagine how or why
    anyone would imagine that mandating self-archiving would slow it down
    still further, rather than accelerate it.

    (6) In the few cases that exist of institutions or departments that
    have mandated self-archiving, it has accelerated OA self-archiving

    (7) In Alma Swan's survey, the majority of authors reported that
    they would willingly self-archive if it was mandated (but not

    (8) Apart from institutional self-archiving, the only other source
    of OA is (i) central self-archiving and (ii) OA journal publishing.

    (9) There are far fewer central archives than institutional archives
    (of course), and, more important, creating and maintaining them is
    far more complicated and costly, because the cost and responsibility
    must be undertaken by a central entity, rather than being distributed
    and off-loaded onto authors' own institutions.

    (10) The biggest of the central archives, the Physics ArXiv, has been
    in existence for almost a decade and a half, contains 300,000 papers,
    but is still growing only linearly, as it has been doing from its
    very first year: That is not fast enough. It would mean that even
    in Physics, 100% OA is still at least a decade away.

    (11) Distributed institutional self-archiving, particularly if
    supported by a mandate, has the immediate potential to reach 100%
    OA virtually overnight.

    (12) Not only is creating and maintaining institutional
    eprint archives far cheaper and easier than creating
    central eprint archives, but it is *incomparably* cheaper
    and easier than creating or converting OA journals!

So unless Heather is recommending that we sit and wait instead for central
archives and OA journals to be created (and filled!), maybe she should
remain mute on self-archiving mandates rather than speculating that
they will slow down OA ("in Canada" or anywhere). Of course parallel
efforts to persuade and help authors to self-archive are welcome too,
but those are not reasons for opining (against evidence and reason)
that mandating self-archiving will slow OA either.

> It would also, in many jurisdictions, create a new form of opposition -
> not necessarily to OA per se, but to this particular approach - based
> on jurisdictional arguments (does the government have a right to say
> that the university must set up an institutional repository). If this
> were to happen, OA itself could get swept up in these other arguments.

Why would anyone prefer to speculate about hypothetical jurisdictional
disputes when the facts are so obvious? Canada has national as well as
provincial research funding councils. If they mandate self-archiving,
their fundees will do it (just as the respondents in Alma Swan's author
survey said they would). That's all there is to it. Shadow-boxing with
hypothetical jurisdictional disputes is just begging the question (and
prolonging the inaction).

> As an example: in Canada, as mentioned, higher education is a
> provincial, not federal matter. In one of our largest provinces -
> Quebec - the idea of separating from the rest of the country has been a
> rather popular notion for many years. For example, a few years ago a
> referendum was held on the subject - the idea of staying in Canada, but
> the vote split was 51:49. The vast majority of government
> representatives from this province on a national level are from the
> Bloc Quebecois - a federal separatist party, silly as this notion is.

Why are we being told all this? I live in Quebec. I teach at UQaM.
And we are busy building eprint archives and planning policy to fill
them. Quebec researchers are funded by both the national and provincial
research councils. If mandated, they will self-archive, like everyone
else. Why do we have to be regaled with Quebec's political history when
it is utterly irrelevant to the matter at hand, which is OA?

> If the Canadian government were to tell Quebec universities that they
> must create institutional repositories, there would be a negative
> reaction - from the universities themselves, the province (which sees
> this as their jurisdiction, not the federal government's), and the Bloc
> Quebecois. If the federal government were sufficiently insistent on
> this - it could literally split up the country. (Quebec would not be
> alone in this - none of the provinces like being told what to do).

I am at a loss. The proposal was to mandate self-archiving in order
to generate 100% OA, and here we are speculating that it might split up
the country.

I regret that I cannot call this anything but the idle, empty speculation
that it is. It would be far, far better for OA if we focussed on reality
and practical, positive action, and resisted the temptation to pile
conjecture on top of conjecture. And where we have nothing substantive
to say, it may be better to say nothing at all.

> That's the kind of situation you can run into when you take a solution
> that makes perfect sense in one country and try to impose it on another
> country, without considering that country's politics and history - even
> with two countries with very similar politics and history, such as
> Canada and the U.K.

No, that's the kind of situation you get into if you let your imagination
carry you off into fantasy instead of focussing on the specific, concrete,
practical problem at hand, which is to accelerate OA, not to voice an
opinion on everything and the opposite of everything.

> Nevertheless, you [addressed to Alma!] are entitled to your opinion, of course.

But is this really about generating an endless stream of opinions,
and opinions on opinions, or are we actually trying to get something done
here? Is this really about opening access or just opining about access,
and about everything else under the sun?

Stevan Harnad

A complete Hypermail archive of the ongoing discussion of providing
open access to the peer-reviewed research literature online (1998-2004)
is available at:
        To join or leave the Forum or change your subscription address:
        Post discussion to:
Received on Sat Oct 30 2004 - 23:10:35 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:39 GMT