Re: Electronic archiving and IIS talk

From: Stevan Harnad <>
Date: Fri, 8 Sep 2000 12:35:24 +0100

On Thu, 7 Sep 2000, Chris Armstrong wrote:

> [RE] <>.
> In your scenario, we would have pre- and possibly post-print versions
> of a paper co-existent with the final, formally published version - and
> it is here that I have to take issue with it. How is the user to know
> which is the copy of record? Which should be cited - and which will HEI
> authorities consider in RAE or tenure-testing exercises?

The copy of record for the refereed postprint is the draft published by
the refereed journal in which it appears, just as it has always been!

There is obviously no "copy of record" for the pre-refereeing preprint,
of which there might even be multiple updates

but that is of course normal at that prepublication phase of the
embryology of knowledge.

There may also be post-publication revisions of the refereed postprint.
But the canonical version will always be the one that was accepted for
publication and appeared in the journal, just as it always was.

In other words, NOTHING CHANGES except accessibility! In the first
phase of the "subversive proposal," the self-archived versions simply
provide a free means of ACCESS for those who do not have access to the
official journal version (whether on-paper or on-line).

How refereed journal publishing will adapt to the availability of free
access is a later phase, and one can only try to make educated guesses
as to the exact form it will take.

My own guess is that user preference for the free version will reduce
subscription revenues and cause publishers to scale down to providing
only the SERVICE of Quality-Control & Certification (QC/C: peer review
and certification as accepted by that journal) instead of providing a
for-fee product (the text itself, on-paper or on-line). There may
continue to be a market for ADD-ONs for-fee, but the final, accepted,
refereed draft will no longer be held hostage to those add-ons (and the
much lower QC/C service will be paid for by the author-institution in
the form of up-front QC/C costs per accepted paper out of its annual
savings from the cancellation of its
Subscription/Site-License/Pay-Per-View expenditures).

The QC-certified version will then have its own "official" sector in
the Eprint Archives, in which authors can neither modify nor remove
anything, and that will become the canonical version and locus
classicus, etc.

That's all there is to it. There are no substantive problems with
getting exactly the same level of authentication, protection and
preservation for digital data as for printed paper.

I think that you are merely confusing the first (subversive) parallel
phase, in which the free self-archived versions will be SUPPLEMENTS to
the official version, with the later phase, in which journal publication
practices adapt to this new PostGutenberg phase for this give-away
research literature.

> Further, what
> is to prevent accidental or intentional changes, not to mention further
> electronic copying and publishing - again with possible unauthorised or
> accidental alterations (see Graham, 1992/3) - of the electronic
> original?

In the first, subversive phase, nothing -- although the Eprint Archives
are designed so all versions remain, and authors simply add new drafts
to supersede them if they wish:

But this first phase is, as I said, a SUPPLEMENT, not yet a SUBSTITUTE,
for the journal version, which co-exists with it.

Once the free version drives the restructuring of journal publication,
the Eprints Archives can have unalterable official-version sectors
(authenticated, date-stamped, encrypted, backed-up, etc.) that will do
the job of what used to be the paper-on-shelves -- and will do it with
at least as good a level of protection and security and permanence as
paper-on-shelves did.

The unofficial pre-refereeing drafts and post-publication revisions
need not be as lapidary as the canonical version, but there is no
reason in principle why they too cannot be, if that is how the Learned
Community prefers it.

There is no problem here; you simply have to think in terms of a
digital archive rather than an analog one, a for-free archive rather
than a for-fee one, and a protected, dedicated sector for the "copy of

> We try to inculcate good practice which includes the
> evaluation of resources in users of the WWW, but it seems to me that
> your proposal, if widely adopted, would muddy the water, at least to
> the extent that indication of authorship and corporate source will no
> longer necessarily help to define authority, accuracy or currency.

Not at all! For now, it will simply be a free way to access versions of
papers that otherwise are only accessible for-fee. And when the system
restructures to adjust to this, the official version will be accessible
for free too, authenticated, protected and preserved (and suitably
sign-posted for navigators).

> Other issues that cannot easily be ignored are those of access (few
> _formally_ published electronic papers find their way into BNB or its
> equivalent) and legal deposit.

I do not understand what you are asking here at all. I don't know what
BNB is, but ACCESS is the core issue here! The objective is to free
access to both the official published literature and its precursor
and successor drafts. That is what subversive self-archiving is all

> Most arguments that electronic publishing will take over from print
> hinge on mechanical rather than social advantages; your imposition of a
> human-originating variable is an unknown.

Mine is not just a prediction, it is a strategy for bringing it about.

Nor is it merely about "electronic publication": Virtually all refereed
journals now have online digital versions already. It is about FREEING
ACCESS to this give-away literature; and the means of achieving this is
author self-archiving -- for the time being as a (subversive) for-free
SUPPLEMENT to the for-fee journal version, and eventually, after the
restructuring this induces, as an authenticated substitute for it too.

> Ginsparg has suggested 10 to
> 20 years before electronic supersedes conventional scholarly
> publishing, others and you, I think, somewhat less.

No one knows how long the restructuring will take; but the freeing
can literally take place overnight, if all authors simply self-archive
their current papers in an Open Archive. And Universities who wish to
increase their authors' research impact (while perhaps eventually even
saving on their serials budgets) need only mount the interoperable
Eprint Archives for their authors to self-archive in.

> The interregnum
> will prove difficult for users who are faced with such choices - I fear
> that all but the most committed will remain faithful to the tried and
> tested formal publishing.

Again, nothing of the sort. No "choice" needs to be made! Authors
continue to do everything they did before (do research, publish
it in refereed journals), giving up nothing, but merely take one simple
additional step: Popping a copy (of both the preprint and the postprint)
into their institutional Eprint Archives. Open archiving and

will take care of the rest.

Stevan Harnad
Professor of Cognitive Science
Department of Electronics and phone: +44 23-80 592-582
             Computer Science fax: +44 23-80 592-865
University of Southampton
Highfield, Southampton

NOTE: A complete archive of this ongoing discussion of providing free
access to the refereed journal literature is available at the American
Scientist September Forum (98 & 99 & 00):

You may join the list at the site above.

Discussion can be posted to:
Received on Mon Jan 24 2000 - 19:17:43 GMT

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:50 GMT