---------- Forwarded message ----------
Date: Sun, 11 Aug 2002 04:12:03 +0100
From: Stevan Harnad <[log in to unmask]>
Reply-To: [log in to unmask]
To: [log in to unmask]
Subject: Archiving Journals in CogPrints
Resent-Subject: Archiving Journals in CogPrints

The following exchange is posted with permission. It concerns questions
about what might be the best way to archive the full contents of a
refereed journal (Medical Education Online, MEO) in a generic Eprint
Archive (CogPrints). The Editor of MEO had originally asked CogPrints for
a category for refereed MEO postprints as well as one for unrefereed MEO
preprints, but, as you will see, the system evolved somewhat from there.

> At 06:40 PM 8/10/2002 +0100, Stevan Harnad wrote:
>
> > In CogPrints subject area
> > "JOURNALS: Medical Education Online: MEO Peer Reviewed"
> > I think there is no need to call the category
> > "Medical Education Online: MEO Peer Reviewed."
> >
> > Calling it "Medical Education Online" should be enough.
> > The "refereed" tag will take care of the rest.

On Sat, 10 Aug 2002, Medical Education Online wrote:

> I am to blame for the category labeling. I asked CogPrints to set up one
> category for published articles in MEO and one for preprints (manuscripts
> submitted to MEO for publication).  I agree, one category makes more sense.
> I can use the refereed tag to differentiate and the update feature to
> handle manuscripts that are later published.
>
> Sorry about the confusion.
>
> Dave Solomon
> Medical Education Online
> http://www.med-ed-online.org
> E-mail: [log in to unmask]

Dear Dave,

No blame, no problem! We are all improvising with a new medium and
powerful new possibilities!

If you were using the eprints.org software yourself, exclusively for
a Medical Education Online (MEO) Journal Archive, you would have been
right to have a separate preprint and postprint sector, because you may
want different access permissions on the preprint sector: MEO or the
authors may not yet want public access to submissions before acceptance;
you might want access only by MEO editors and referees.

But as you are using the multidisciplinary CogPrints Archive as the public
archive for both MEO preprints and MEO postprints, it does not make sense
to have separate sectors with separate access privileges: You intend both
the preprints and the postprints to be openly accessible. In this case,
the refereed/unrefereed tag should be enough, together with the category
MEO for the journalname (for the postprints)

The only point I'm not sure of is whether the CogPrints software is
currently so configured as to allow a journalname to be specified for
an unrefereed preprint! In general, depositors are strongly discouraged
from naming the journal when they deposit a preprint, as it is quite
possible that that particular journal will not accept their paper --
in which case there is a permanent record of the name of the journal
in which their preprint was NOT accepted! (In addition, we strongly
encourage authors to treat CogPrints as archival, not removing their
papers except under exceptional circumstances.)

So it is possible that CogPrints set up the special MEO/refereed and
MEO/unrefereed categories to accommodate these two contingencies.

What I would suggest instead is that you simply not name the journal
when submissions are deposited as preprint, and -- if/when the preprint
is accepted -- deposit the refereed, final, accepted draft again
as an updated postprint (i.e., refereed), this time specifying the
journal-name. It will then be embryologically linked to its predecessor
preprint.

This may seem 100% convenient if you are in fact using CogPrints as
your refereeing buffer (i.e., the electronic locus where you direct your
referees when they are invited to referee the submission), but only
inconvenient in that it is the MEO editorial office, rather than
CogPrints, that will signal which papers are indeed submissions to MEO
-- which may not be a bad thing!

(This is one of the minor inconveniences of using CogPrints as your
journal archive, rather than mounting an MEO-specific version of Eprints
of your own, which could then be configured for and reserved exclusively
for MEO journal editorial use.)

I think this inconvenience is very minor, though, and is in fact a
sensible way of having one's cake (public accessibility for MEO preprints)
and eating it too (concealing that they are for MEO until/unless they
are accepted).

Behavioral and Brain Sciences (BBS) -- http://www.bbsonline.org/ -- a
journal I have edited for 25 years, now also uses Eprints software for
its archive and refereeing, and we have had to create separate sectors
for those preprints whose authors do and do not wish to make it public
that their papers are under refereeing by BBS: Those who do not want
the BBS connection made public do not have their preprints made public
at all in the BBS archive: Only BBS referees can access them. This is
more conventional for a journal, but not the optimal use of the archive,
which, being dedicated specifically to BBS, cannot make preprints public
without revealing which journal they were submitted to!

Food for thought, for prospective users: The obvious solution is
that authors should archive their unrefereed preprints in their own
institutional Eprint archives; the journal can just use the author's own
institutional archive as an "overlay," directing referees to that URL to
review the (locked, date-stamped) submitted draft, and eventually, if all
goes well, "cerifying" the final, accepted postprint through a link to
their own journal-archive site, on which a definitive copy of that
final, accepted draft also resides as a mirror of the author's
self-archived postprint. Indeed, the journal's "definitive" copies need
not be in an Eprint archive devoted exclusively to that journal; the
journal-archive could be a whole collection of publisher-authenticated
versions of the accepted final drafts of many journals (i.e., a
publisher-authenticated collection of postprints for specific journals).
Indeed, the journal-archives could be virtual, consisting of an
official, password-protected system of distributed mirroring across the
institutional archives themselves!

Such a system of coordination between author/institutional archives and
publishers' authoritative archives (real or virtual) could serve the
multiple purposes of authentication and redundancy (for preservation).
And would would be incomparably cheaper and more efficient than the
current system of refereeing and dissemination.

And this would help hasten that optimal/inevitable day when we realize
that a PostGutenberg refereed-journal -- an open-access journal -- is
really just a peer-review service-provider and certifier. "Publication"
consists of the affixing of the journalname tag, the publisher's official
certification of the final peer-reviewed draft that has resulted from the
the dynamic refereeing/revision process. We are all looking for an
"economic model" that will allow this essential service -- peer review
and certification -- to be "unbundled" from all of the other things that
publishers did in the Gutenberg age, and are still doing now, such as
providing (and hence selling) a text (along with various value-added
extras that may or may not be wanted or needed, over and above the
peer-reviewed, certified final draft).

Just thinking aloud here...

Cheers, Stevan

> Date: Sat, 10 Aug 2002 21:05:55 -0400
> To: Stevan Harnad <[log in to unmask]>
> From: Medical Education Online <[log in to unmask]>
> Subject: Re: Archiving Journals in CogPrints
>
> Dear Stevan,
>
> Thank you. Your discussion above made obvious what I should have realized
> in the first place. I probably should not be posting submitted manuscripts
> as preprints, at least without the explicit permission of the authors. I
> certainly do not want to prevent authors from resubmitting the manuscripts
> to journals that do not accept manuscripts distributed as preprints.
> What I will probably do is change the submission form to allow authors
> to indicate their preference.
>
> I couldn't agree more about institutions having their own preprint servers.
> Unfortunately mine does not at the present time despite my prodding. I am
> hoping over the next few months to set up a server get a copy of Eprints
> running here. For convenience and because it is so inexpensive, I have been
> using a commercial server for my journal and though it is UNIX-based and I
> have access to Perl, after looking over the Eprints documentation, I don't
> believe I can get it to work without root-access on the server.
>
> FYI, my main reason for wanting to use CogPrints is not so much for
> archiving (the manuscripts are available on the journal web site) as for
> getting the metadata for the manuscripts available through an OAI
> registered archive so they can be more easily be located. I do very much
> appreciate your letting me use CogPrints and it never hurts to have the
> redundancy of a second copy available.
>
> Please feel free to post this on American Scientist Forum.
>
> I'm not sure much of an economic model is needed for peer-review as
> (obviously) all the real work is done for free already. The mechanics of
> running the process can be automated with Web forms and server-side
> databases much as you have done with Eprints.  The real challenge seems to
> be changing attitudes which hopefully will just take some time.
>
> Dave Solomon
> Medical Education Online
> http://www.med-ed-online.org
> E-mail: [log in to unmask]