[GOAL] Re: Harnad Comments on Proposed HEFCE/REF Green Open Access Mandate

[Stevan Harnad]

On Fri, Mar 15, 2013 at 11:56 AM, Graham Triggs <grahamtriggs at gmail.com>wrote:

> On 14 March 2013 22:14, Stevan Harnad <harnad at ecs.soton.ac.uk> wrote:
>
>> Why is it an absurd requirement to deposit immediately in the author's
>> IR, regardless of whether the
>> journal is subscription or OA and of whether the deposit is embargoed or
>> immediate OA?
>>
>> That simple, natural, uniform local deposit procedure is precisely what
>> makes it easy for an institution
>> to monitor compliance.
>>
>
> imho, there are some significant unanswered questions regarding the
> HEFCE/REF proposals, which ultimately boil down to a couple of points. The
> main one actually being covered by what you've said above.
>
> Sure, an institution can monitor compliance. In fact, as they run the
> repository, they are the only ones that can effectively monitor compliance.
> So how exactly are the requirements going to be audited and enforced?
>
> There is no requirement to make the metadata public. There is no
> requirement to have the metadata harvested (whether public or not). There
> isn't even a requirement to have a "request a copy" feature (without which,
> the usefulness of immediate deposit is rather lost).
>
> And nor can there be in any useful time period for the first post-2014
> REF. These things will take time to build and/or implement. So there isn't
> any effective way to audit that deposits were made, much beyond actually
> being fully open access when the embargo ends at best, and possibly even
> only at the time of the return at worst.
>

I think you are mistaken -- and that you are vastly under-estimating the
reach of this simple REF/HEFCE policy:

(1) It is *institutions* that have always shown intense eagerness and
initiative in ensuring that their researchers comply with all RAE and REF
conditions.

(2) The proposed REF mandate makes it very explicit that *REF submissions
are ineligible if they are not deposited immediately upon publication*. (No
waiting till near the end of the 6-year REF cycle to deposit.)

(3) Compliance is based on two objective, verifiable data-points:
publication date and IR deposit date.

(4) Institutions, in monitoring and ensuring compliance will simply require
-- at least annually -- a list of articles published, together with
publication date and deposit date.

(5) If the publication date and the deposit date are not the same, the
article is ineligible for REF.

(6) With deposit, the metadata are immediately accessible web wide (though
the full-text might be embargoed for the allowable interval).

(The request-copy feature will be implemented by IRs as a natural matter of
course, once immediate-deposit is effectively mandated.)

The second issue comes from the requirements on what must be deposited -
> namely, the final post peer-review material, and under a licence that
> allows text-mining, etc. This has two implications - the first on authors,
> as far from being unrestricted in the journal choice, they will need to
> ensure that the journal policies will allow them to fulfil their HEFCE
> requirements. Understanding this is not straightforward with the resources
> currently available to authors.
>

You are right about this, and I have recommended to HEFCE not to press the
re-use rights too hard in order not to constrain journal choice with
non-quality constraints. Re-use rights should be treated the same as the
embargo: Zero embargo and full re-use rights preferable, but some embargo
and some re-use restrictions are allowable, if necessary.

*It is the immediate-deposit that is crucial, Once that becomes universal,
embargoes and re-use restrictions don't stand a chance and will soon
crumble.*
*
*

> The other implication is on the repositories themselves. The proposals
> will create another class of material in repositories. Deposits for REF
> will have 'liberal' licences - but what about existing content, which won't
> have the same licences attached to them? How will the licences be
> communicated? How will users be aware of the rights? This mix of licences
> will either diminish the value of the HEFCE requirements, or increase the
> costs of fulfilling them, or both.
>

The license issue is a red herring, and premature. Get all the content in
there, immediately on publication, reliably, and the rest will take care of
itself of its own accord soon after. Fuss instead about the rest,
pre-emptively, and you won't even get the content.

Having an IR (at a basic level) may be low cost, and voluntary deposit
> 'cost free'. But mandating Green OA, and in particular monitoring,
> enforcing and auditing compliance - especially when those requirements are
> as specific as the HEFCE proposals - does have a cost. And
> enforcement/auditing is something that will be needed - and will need to be
> effective - to achieve high compliance. Otherwise it may well fall short,
> regardless of the consequences.
>

RAE and REF compliance has always entailed some cost to institutions, but
they have willingly undertaken it in order to maximise their chances of the
benefits of a high RAE/REF ranking and top-sliced funding as a reward.


> This is at least, one thing that can be said in favour of Finch, etc. -
> they are proposals that are easy to understand, easy to follow, and easy to
> audit.
>

Not in the least! (I am astonished to hear you think that!) Finch mandated
Gold or Green. The mechanisms for auditing and dispensing Gold are not yet
worked out, and certainly not going to be simple or easy. The mechanisms
for auditing and ensuring Green are non-existent: HEFCE/REF provides them,
and they are easy to understand, easy to follow, and easy to audit.

Stevan Harnad
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ecs.soton.ac.uk/pipermail/goal/attachments/20130315/b5dc1496/attachment.html