Re: Draft Policy for Self-Archiving University Research Output

From: Stevan Harnad <>
Date: Wed, 8 Jan 2003 20:19:11 +0000

> Date: Wed, 8 Jan 2003 13:57:41 -0000
> From: "Picciotto, Sol" <s.picciotto_at_LANCASTER.AC.UK>
>sp> It seems that copyright ownership could be an important obstacle to
>sp> archiving postprints.

It is only an obstacle if the author signs a copyright transfer agreement
that gives up the right to self-archive the postprint. Our advice to
authors is always to self-archive the preprint (which precedes submission,
refereeing, revision, postprint, and copyright agreement) and if/when it
is revised and accepted as the postprint, to change the wording of the
copyright agreement to allow self-archiving of the postprint too. If/when
the publisher insists, sign the restrictive agreement and then archive only
the corrigenda that list the substantive changes that need to be made
to the preprint to turn make it equivalent to the postprint.

In most cases, publishers will agree to the self-archiving of the
postprint if asked. For the shrinking number of cases where they do not,
the preprint+corrigenda will do for now.

It would be a *great* strategic error (indeed, it would be absurd)
to hold up the majority of the research literature, and to continue to
lose its potential research impact because of access-barriers, until
the postprint self-archiving right is officially incorporated into every
publisher's copyright transfer form!

For over 10 years now the physicists have had the good sense not to be
held back by this: The rest of the research community should now show
equally good sense. (See also the recent posting by Mark Doyle,
of the American Physical Society (APS), the highest
quality and highest impact publisher in Physics.
APS now explicitly allows postprint self-archiving -- but the physicists
did not wait for this announcement to do it!)

>sp> I have proposed at Lancaster that academic staff employment contracts
>sp> be modified to make it clear that the university asserts its rights as
>sp> employer to copyright in staff research publications, but only to the
>sp> extent of reserving the right to authorise non-commercial publication
>sp> on the internet, e.g. in an eprints archive. This would circumvent
>sp> a possible restriction resulting from any copyright assignment the
>sp> author signs. The idea has been met favourably here, both by the AUT
>sp> (professional association) and management, but both have referred it
>sp> for discussion at national level.

This is excellent and should certainly be done, but *in parallel* with
universal self-archiving along the lines described above, and definitely
not as a *precondition* for it.

[ A similar institutional copyright retention initiative was launched by
Provost Koonin of CalTech
as discussed in this Forum 4 years ago:
but CalTech has since become one of the leaders in research
self-archiving: ]

>sp> I think the university should be willing to forego any claim to income
>sp> from research publications, but should retain the right to authorise
>sp> non-commercial publication. The decision on when to publish, which
>sp> version, etc, should be left to the author(s), within a policy such
>sp> as that suggested here for Southampton, which would greatly facilitate
>sp> acceptance of eprints archiving as a standard practice.

Claim to income from research publications? The author receives no income
from papers published in refereed journals. But to put any of this
on the self-archiving/open-access agenda at all (except in-parallel)
is to invite still more delay when over a decade of potential research
impact and usage have already been needlessly lost. For this introduces
the gray area of books and of teaching materials (including textbooks):
All worthy topics, but *irrelevant* to the clearcut, open-and-shut case
of refereed research papers.

Yes, let university copyright retention initiatives proceed apace, but let
them not portray themselves or be perceived as necessary *preconditions*
before authors can safely self-archive all their research output! They
already can do it now -- and should, with no further delay.


> Date: Wed, 8 Jan 2003 15:52:47 -0000
> From: Fytton Rowland <J.F.Rowland_at_LBORO.AC.UK>
>fr> Before you can have a "postprint" you must have published the paper
>fr> somewhere. In many (most?) cases you will have transferred the copyright
>fr> to the journal. So how can the University then assert its ownership of
>fr> a copyright that you, the individual academic, have already given away
>fr> in the belief that it was yours to give?

See above. The idea of institutional copyright retention is to give
authors more clout when they propose changes to the wording of their
publisher's copyright transfer agreement, so as to allow self-archiving
of the postprint. Two minor down-sides are:

    (1) If the publisher refuses, it puts universities in the position of
    dictating to their authors where they may and may not publish their
    refereed research. (Not a good idea, nor one likely to be welcomed
    by authors.)

    (2) It needlessly retains *more* than is necessary for self-archiving
    and open access (see Mark Doyle, below).

But, more important than these, the major problem with waiting for
changes in university copyright retention policies is:

    (3) It isn't *necessary* -- and propagating the misconception that
    it is necessary simply serves to further delay self-archiving and

> Date: Wed, 8 Jan 2003 16:19:09 -0000
> From: Tim Brody <>
>tb> If the author's employment contract states that their employer (the
>tb> University) reserves non-commercial distribution rights then that author
>tb> can not sign away those rights to a publisher (without the agreement of
>tb> the University).

Correct, and that would be a bad thing, because it would take away from
the author the right to be the sole determiner of where his research is
submitted for publication! (And it would do so needlessly.)

>tb> In my opinion I would rather the IPR were held by the institution -
>tb> who paid for the research, facilities & support - rather than with the
>tb> publisher. If not for any other reason than an institution will rarely
>tb> hold the same kind of monopoly as the big publishers.

All irrelevant when the goal is open access: Self-archiving ensures
that. And once the research is open-access, "monopolies" in the
toll-access version are of no interest.

> Date: Wed, 8 Jan 2003 16:55:34 -0000
> From: "Picciotto, Sol" <s.picciotto_at_LANCASTER.AC.UK>
>sp> You as author can only transfer any rights you actually have. If the
>sp> university as employer retains any rights to works which you produce
>sp> in the course of your employment, which it is entitled to do, then you
>sp> can only transfer the rights the university has agreed to allow you to
>sp> retain. My proposal is that the employment contract should specify that
>sp> the university retains the right to authorise non-commercial publication,
>sp> but allows the author(s) to negotiate commercial publication rights. Any
>sp> licence or assignment signed by the author could only validly transfer
>sp> rights which the author actually owns, so the author could only grant
>sp> non-exclusive rights to a publisher.

This may be useful, though it does seem to conflate the refereed-research
literature, which is always an author give-away -- regardless of whether
the publisher is commercial or non-profit, toll-access or open-access --
with other forms of literature, such as books, textbooks and teaching
materials, possibly also software and multimedia. There is no need for
the refereed-research literature to be mixed up with the rest of the
literature at all. It is a special case. And it does not *require*
copyright retention, by either the institution or the author; it
requires only the online self-archiving right, which is uncontestably
the author's for the unsubmitted preprint, and if ceded for the
postprint, easily remediable through corrigenda.

Holding out for anything else -- or more -- is simple delaying the
optimal and inevitable, needlessly, and at a growing cost in research
impact, usage, and, ultimately, productivity and progress.

>sp> The only problem I foresee is if the publisher's contract has a clause
>sp> requiring the author to warrant ownership of all the rights. This could
>sp> be dealt with fairly easily by a standard form clause that all staff
>sp> would be advised to send to publishers, pointing out that the university
>sp> has reserved the right to authorise free publication via the internet or
>sp> eprints archive, so rights acquired by the publisher are subject to this.

This should certainly be done -- but universal self-archiving should not
wait one microsecond for it to be done: It should proceed regardless.

Date: Wed, 8 Jan 2003 13:56:09 -0500
From: Mark Doyle <doyle_at_APS.ORG>

>md> The APS takes copyright for a variety of reasons (it has been a good
>md> thing lately for both scanning backfiles and for handling recent cases
>md> of scientific fraud and plagiarism).

Agreed. No harm in transferring copyright as long as the self-archiving
right is retained.

>md> The APS approach is to grant back to authors all of the rights they
>md> expect (such as putting material in centralized or institutional e-print
>md> archives, just not the APS formatted version).

That's plenty. The postprint need not be the publisher's PDF, in this
day and age. (In fact, it can be much better!)

>md> The question is whether there could be a mechanism for the publisher
>md> to also give these rights to the author's institution and even whether
>md> there could be an arrangement that allows institutional archives to use
>md> the publisher formatted version for their own archives (and how to ensure
>md> the record stays complete and up-to-date regarding for example, errata,
>md> retractions, corrections to online material, etc.). This last part is
>md> one problem with authors "self-archiving" - they may not be inclined to
>md> make sure that errata (or worse) get recorded properly.

Errata/corrigenda are a matter of scholarliness and scientific rigor
for refereed research and researchers. But it's always nice to have others
invigilating too. The referees were one, the publisher is another, and,
in the open-access world, the entire peer community is yet another.

>md> Another point worth making is that authors often have stronger
>md> associations over their career with the publishers of the journals
>md> they submit to than with their institutions. Authors might not be happy
>md> relinquishing rights to their work if they change employers.

Moot point, when the work is open-access. The author retains the
authorship (the "moral right") in any case. His CV goes wherever
he goes. There is no royalty for refereed research, just the indirect
rewards of research impact, which is maximized by open access. So what
are we talking about retaining or not retaining here?

Stevan Harnad
Received on Wed Jan 08 2003 - 20:19:11 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:48 GMT