Re: Decision on OA policy Lund University

From: Stevan Harnad <>
Date: Wed, 14 Dec 2005 20:16:20 +0000

On Wed, 14 Dec 2005, Lars Bjørnshauge wrote:

> You wrote: "Once the Lund Policy is optimised, I hope you will register
> it as a model for other institutions to emulate in."
> I do hope that we are allowed to register the existing policy!?

Of course, Lars! The Registry currently contains 5 institutions which require
self-archiving and 11 that recommend it.

My suggestion for optimising the Lund policy was based on all the evidence
to date (and you can check it by comparing the self-archiving growth rates
of institutions that require it with those that merely recommend it):
Requiring works, and works fast; recommending is slow, and it is not
clear whether it would ever approach 100% or even 50%. Arthur Sale's
forthcoming article, which I also cited, illustrates this quite clearly:

    Comparison of Deposit Policies

> You wrote: If one slight change is made to the Lund OA policy, the
> strongly recommends" can be uncontroversially changed into a "requires"
> I guess I do not need to mention that political processes within an
> organisation such as a university are not easy to manage. Therefore the
> wording: "strongly recommends" is not easily exchanged with "requires"
> mainly because it sends a slightly different message and because it
> reflects the way we do things in Lund.

Well I would say that requiring self-archiving is slightly different from
the way they do things at 10,000 - 5 = 9,995 universities (and research

But the five who have required it have shown that it could be done,
and it works. It is merely a record-keeping policy, it merely requires
deposit (not necessarily OA-setting, which is merely recommended),
deposit can be done by proxy (secretaries, library help), and takes only
a few minutes per paper -- not much more than a digital CV update.

Not so hard or unthinkable or impolitic as one might think if one has
not quite thought it through, or tried it!

> The bottom line is and will be that we have to implement services,
> workflows and show the benefits in such a way that it becomes attractive
> for researchers to act according to the policy.

The *benefits*, dear Lars, are already shown and known:

The trouble is that neither the benefits, nor a recommendation, nor even
library help are enough to generate a large enough self-archiving growth
rate. And that too is already shown and known. And researchers themselves
have as much as said: "I'm busy. I won't self-archive until/unless I
am required by employer or funder." But if required, 95% say they
would comply, 81% willingly (14% reluctantly), and that is exactly
what the actual data on requiring vs. recommending are confirming.

> You wrote: "Once the Lund Policy is optimised, I hope you will register
> it as a model for other institutions to emulate in."
> First of all I do not think it is important to optimise the policy
> for a foreseeable future, second it would be an own goal of dimensions
> to try to do it.

Perhaps you will find the above comparative evidence relevant; also the
evidence of the feasibility of a self-archiving requirement, after all!
Lund's IR is one of the first. It would be nice to see it filling with
100% of Lund's annual research output at last.

> What we want to do instead of fine tuning a policy is to concentrate on
> the huge job we have to do in order to actually implement the OA-policy
> then policy adjustments might come later, when hopefully everyone can
> see the actual benefits.

What you mean is that so far the Lund IR has been operating with neither
recommendation nor requirement, only library help. Now it will be recommendation
plus library help. I wish you the best of success, but as prior evidence
is that this is not enough, it might be better to hedge your bet!

Best wishes, Stevan

> ----- Original Message -----
> From: Stevan Harnad <harnad_at_ECS.SOTON.AC.UK>
> Date: Wednesday, December 14, 2005 7:30 pm
> Subject: Re: Decision on OA policy Lund University
> > To save time (and minimise multiple snippet-postings) I am posting and
> > replying to Michael Day's comment in the same message (and appending
> > Heather Morrison's at the end):
> >
> > On Wed, 14 Dec 2005, Michael Day <> wrote:
> >
> > > > SH: If one slight change is made to the Lund OA policy, the
> > "strongly> > recommends" can be uncontroversially changed into a
> > "requires":>
> > > Surely the main difference is that _requiring_ something means
> > that an
> > > organisation needs to set up some kind of monitoring and/or
> > enforcement> regime? This is not just a matter of semantics but
> > would have genuine
> > > resource implications.
> > >
> > > Michael Day
> > > Research Officer
> > > UKOLN, University of Bath, Bath BA2 7AY, United Kingdom
> >
> > Yes, there are resource implications, but not the (minor)
> > administrative ones
> > (for monitoring compliance) that Michael Day is alluding to here.
> > The major
> > resource implications are the researcher and institutional revenue
> > losses from
> > *failing* to maximise the impact of their research output (by
> > requiring it):
> >
> > Maximising the Return on the UK's Public Investment in Research
> >
> >
> > It is already quite clear, both from international survey data on
> > researchers' self-report data on future compliance with self-archiving
> > requirements
> >
> > Open access self-archiving: An author study
> >
> >
> > and from actual statistics for self-archiving rates for "recommended"
> > versus "required" self-archiving policies
> >
> > Comparison of Deposit Policies
> >
> > scientist-open-access-forum&D=1&O=D&F=l&P=90391
> >
> > that recommendations simply do not work and requirements
> > definitely do.
> >
> > Depositing papers in the author's institutional IR is a record-
> > keeping matter,
> > just like updating a CV for annual performance review. Yes,
> > something has to
> > be put into monitoring compliance, but a sensible policy will just
> > make it
> > a routine part of performance review (and, in UK universities, RAE).
> >
> > And the thing to keep in mind is the resourcing implications of *not*
> > requiring deposit (because all the evidence is that if you don't
> > require deposit, you don't get much depositing -- just as if you do
> > not require
> > publish-or-perish you don't get much publishing).
> >
> > Or is Michael worried about the resource implications of requiring
> > publishing too?
> >
> > Another opinion (this time favorable) about requiring vs.
> > recommending follows
> > below. I recommend setting aside opinions and looking at the
> > actual objective
> > evidence, which is all favorable.
> >
> > Stevan Harnad
> >
> > On Wed, 14 Dec 2005, Heather Morrison wrote:
> >
> > > SH: If one slight change is made to the Lund OA policy, the
> > "strongly> recommends" can be uncontroversially changed into a
> > "requires":
> > Good suggestion, Stevan - one worth highlighting and repeating. No
> > doubt universities wish to develop flexible policies to accommodate
> > faculty choice, however, a requirement is much more empowering for a
> > faculty member than a strong recommendation. In the rare instance
> > where a faculty member wishes to publish with one of the very few
> > remaining publishers with no green self-archiving rights, it is much
> > easier for the faculty member if they are required to self-archive.
> > A recommendation is more likely to invite negotiations, wasting the
> > researcher's time, while a requirement is more likely to invite
> > either a change or an exception to a publisher's policy.
> >
> > hope this helps,
> >
> > Heather Morrison
> >
Received on Wed Dec 14 2005 - 21:10:30 GMT

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