Re: Eprint versions and removals

From: Stevan Harnad <>
Date: Tue, 10 Aug 2004 09:29:50 +0100

On Mon, 9 Aug 2004, Ted Bergstrom wrote:

> I have been wondering about the same issue that concerns Barbara Quint
> in the message quoted below. It seems to me much more satisfactory if
> an open archive contains an exact copy of the publisher's pdf version.

More satisfactory than what? (It is *extremely* important that you ask
yourself this question!) At the moment, we have 10-20% OA. Is it more
satisfactory to keep holding out with 10-20% OA until and unless we can
have 100% OA with publishers' PDF? or should we proceed to 100% OA now,
with the author's peer-reviewed final draft (postprint), forgetting
about the publishers' PDF version?

Of the 8000+ journals sampled so far, 61% of them have already given
authors their green light to self-archive the peer-reviewed final
draft (postprint) and 23% more (total: 84%) have given their (pale)
green light to self-archive the unrefereed preprint. Only 16% are gray
(no green light at all yet).

I have no idea what subset of these 84% green publishers happens to have
given its green light for the PDF (though it can be checked in each case
by any interested author, via the "view policy" button at the
SHERPA site).

But when 80-90% of the annual 2,500,000 articles published are still not
OA at all, who cares about PDF? The cupboards are bare and the kids are
hungry: Should we keep them bare and hungry until we can fill them with
factory-wrapped food? Will generic wrapping not do?

> Two good reasons for allowing this approach are accuracy of the
> scholarly record and convenience for authors.

Two good reasons for whom? This approach (PDF) is not *disallowed* by
proponents of author self-archiving. It is just that many publishers
have not given their green light to self-archive their PDFs. Far more
important, the PDF version is not *necessary* -- whereas OA itself is.

The users whose institutions have toll-access to the articles in a given
journal today are not the ones with the access problem; the ones with
the access problem are the would-be users who do not have access. For
them, this is not merely a matter of convenience or accuracy, it is a
matter of necessity. Access to the content of the peer-reviewed, accepted
final draft (the postprint) fulfills their needs, completely. It is the
content they need, not the form. The PDF would be welcome too, but it
is not necessary, and certainly not something to hold back and keep
waiting for.

In order to maximise their own research impact, authors need to make the
*content* of their peer-reviewed research articles accessible online to
all would-be users. Self-archiving the contents of the peer-reviewed,
accepted final draft is sufficient to achieve this. *And infinitely better
than what we have now!* That is the essential fact that we have to keep in
mind, as we contemplate whether we need to hold out for any more
"convenience" or "accuracy." (Nothing wrong with trying to get permission to
self-archive the PDF too, but nothing we should wait for, or make anything
conditional on, or worry about.)

> It seems to me that publishers should be given a stronger incentive
> to permit putting the publisher's pdf in their institutional archive.

Eighty-four percent of the 8000+ journals sampled have now given their
green light to self-archiving because of the research community's strongly
expressed need for OA. There is definitely this strong need for OA in the
research community. But there is no desperate need for publisher's PDF! 100%
OA can be had without it. Publishers know this. The call for OA was
substantive; a call for PDF would be gratuitous, and hollow, and certainly
nothing for which to delay or reject having the immediate benefits of
OA on the publishers' existing terms -- terms and benefits of which most
authors have not taken advantage yet, despite all the clamour for OA!

The self-archiving mandate by their research-funders and institutions
will induce authors to take advantage of that green light, in their own
interest, their institutions' interest, and in the interests for research
progress and productivity itself. But to first insist instead on PDF
too, instead of self-archiving, would really cast doubt on the seriousness
and understanding of the authors who purport to desire the benefits of
OA. Research usage and citation impact do not come from the form (PDF)
but the content (the refereed final draft) of the research.

> At least one commercial publisher already allows this, so it is not
> beyond possibilities that others could be induced to.

I cannot follow this logic! If the majority of publishers allow self-archiving the
refereed final draft, but at least one of them also allows self-archiving the PDF,
what is supposed to follow from this? We are already *not* self-archiving. Should
we continue not-self-archiving now in order to wait for permission for the PDF too?

The UK and US mandates have wisely ignored these matters of luxury, and focussed on
matters of necessity only (content, not form). I quote from the UK recommendation:

    "We recommend that the Research Councils and other Government
    funders mandate their funded researchers to deposit a copy of all
    their articles in their institution's repository"

No mention is made that that copy must be the publisher's PDF. And it
needn't be; this is the digital age, and what matters is the content,
not the form.

> Wouldn't it be a good idea for Sherpa and the scholarly community to
> distinguish between "full green" publishers who do allow self-archiving
> the puplisher's pdf version, and "pale green" for those who allow the
> author to prepare her own pdf version of the final copy?

I think this would be an exceedingly *bad* idea!

The SHERPA version of the Romeo self-archiving policy list already
codes far too much unimportant and distracting information, with far to
many superfluous colors (green, blue, yellow, white, red and gray!).

In reality, the only fundamental distinction is green vs. gray: Green
publishers *have* given their green light to author self-archiving,
Gray publishers have not.
That's what prospective self-archiving authors need to know (and by
*journal*, not by publisher!).

Then there is the lesser distinction between pale-green (preprint only)
and full green (postprint).

There is definitely no reason to subdivide the postprints further in
terms of PDF or no. Any author who cares about that specific detail can
look it up in the publisher's policy link. It is of no interest to the
objective, which is 100% OA, as soon as possible! (And no further colour
codes, please!)

> If self-archiving does catch on, authors are likely to find it important
> that they can archive exact copies of their final version.

When self-archiving catches on and generates 100% OA there will be *far*
more important consequences than anything to do with PDF or non-PDF
versions. (And as XML prevails, the notion of an "exact copy" as a PDF
page-image will be seen to be the obsolescent papyrocentric relic it
already is!)

> Accordingly, publishers who want to attract good authors will be under
> some pressure to accommodate this desire.

The only responsible and progressive response on the part of publishers
to the growing and legitimate pressure from the research community's for
the benefits of OA can be -- and already is -- to go green. That will
allow the the research community to provide for itself exactly what it
wants and needs (through self-archiving). The rest of the details are
unimportant and will take care of themselves. Nothing more can justifiably
be demanded from publishers at this time.

But we are not there yet! and it is not the publishers' fault! No
self-archiving mandate has yet been legislated or implemented. So why are
we worrying here about factory-wrapping when there is an access/imact-hungry
research community that can be fully nourished with our generic postprints?

> Might it also be a good idea for archives to develop a convention
> that makes it clear when an archived copy is an exact replica of the
> published version?

No harm. But shouldn't we fill them first, before we worry about tagging
the ones that are factory-wrapped?

Stevan Harnad

> Ted Bergstrom
> >Do the "green light" publishers send digital copies back to the authors?
> >
> > [Some give the green light for the author to use the publisher's
> > PDF version, others only to use the author's own digital versioni.]
> >
> >
> >As an editor, I'm assuming that some of the text has been changed as it
> >goes through the editorial process.
> >
> > [Correct, especially from unrefereed preprint to refereed postprint.]
> >
> >
> >So are the self-archives of the edited "final" versions or of the author's
> >initial submissions?
> >
> > [If the green light is for the refereed, edited final version,
> > that is what is self-archived; otherwise the preprint plus
> > corrections.]
> >
> >
> >Or would there be more than one version, e.g. an author's edition (like
> >a director's cut video) that includes portions that didn't appear in the
> >"printed" article as well as the published one?
> >
> > [Authors can self-archive all significant stages of their work,
> > including pospublication corrections, updates and enhancements.
> > The essential target of OA, though, is contents of the peer-reviewed
> > final draft.]
> >
> >
> >>that's what i thought. a librarian's/archivist's nightmare.-- bq
> >>
> >>[But a researcher's dream. -- SH]
Received on Tue Aug 10 2004 - 09:29:50 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:47:33 GMT