Re: Nature 10 September on Public Archiving

From: Stevan Harnad <harnad_at_COGSCI.SOTON.AC.UK>
Date: Fri, 11 Sep 1998 10:00:25 -0400

> John Unsworth <jmu2m_at_VIRGINIA.EDU> wrote:
> there's a difference between a pre-print archive and a journal...
> one of the major differences is editing--not just the peer review
> process, but copyediting, source-checking, etc.. If we were to turn
> things like copyediting and source-checking over to authors, the
> consistency and quality of what we produce would go down-hill rapidly.

Completely agree. Archives and refereed journals are different, and
under discussion is the funding of precisely the quality control you

But free public eprint archiving applies both to unrefereed preprints
and refereed reprints, don't forget; that is their subversive
dimension. And that it why it is essential to plan the transition
strategy carefully and collaboratively.

> My estimate is that this costs about $5K an issue, in our case...
> Each issue of PMC contains roughly half a dozen articles

That would work out to about $400 per paper, which would be in the
ballpark we were considering for page charges.

> --plus notices,
> reviews, and other things which have value to readers, but which I
> can't imagine authors paying to have published (would you pay to have a
> review published?). If the cost of the whole were assessed to authors,
> the per-article cost to authors would be roughly $933 per article.

$1000 sounds a bit on the high side; maybe the extras that have nothing
to do with the articles should not be wrapped into their costs. And if
the extras are add-ons that the publisher can sell, let them be sold
separately, while the articles are available free, their costs covered
by author page charges. Read on.

> Humanities research of the sort that results in articles we would
> publish is almost never grant-funded, nor is it university-funded: in
> short, there is no budget from which authors would normally be able to
> draw...

If the per-page savings from transforming to on-line-only are indeed
70% then arithmetic says the remaining 30% could be paid out of the
100% saving from journal subscription cancellations, as stated in the
Nature article.

> If Universities elected to establish such a fund, all
> at once and everywhere, and if journals in the humanities all at once
> and everywhere agreed to begin operating on this basis, then all would
> be well--but the devil's in the interim.

The unstable transition period will indeed require careful thought,
advance planning, cooperation, and some one-time tide-over subsidy:

See the following thread in this forum:

The Urgent Need to Plan a Stable Transitio

> IF all journals don't operate this way, those that do will be at a
> distinct, probably fatal, disadvantage--authors will not submit
> articles to them.

Yes, trying to recover costs from page charges before the practise is
established would fail (and has failed already , in a few existing cases).

But that is the wrong way to go about it. Tide-over subsidy is needed
during the transition period, both for new on-line-only journals and
for hybrid paper/on-line journals trying to make the transition from
S/SL/PPV to on-line-only supported by page-charges.

> Even if all journals, tomorrow, adopted the new
> economic model, if Universities didn't simultaneously and universally
> establish funding for authors to pay these charges, then some, maybe
> many, authors would be at a distinct, probably fatal, professional
> disadvantage.

Indeed. And the universities will have no funds for the funds (sic)
until they have in hand the windfall from the S/SL/PPV cancellations.

So clearly a managed, subsidised transition period is necessary; but it
need not be a long one. And the sooner we plan it the better, because
otherwise subversive erosion will take place from de facto user
practises (as with xxx) and instability or even a chaotic period could
ensue instead.

> And even if universities did universally adopt this
> model, where would that leave independent scholars, creative writers,
> and others who do (at least in the humanities) publish in academic
> journals?

Whereas it would be undesirable in the extreme (and subverted in any
case by author self-archiving) to have other services and unnecessary
add-ons factored into author page charges, there would be nothing wrong
with averaging in the page costs for unaffiliated scholars. The only
problem here would be disciplines where unaffiliated authors represent
a significant proportion of the authorship. (Does anyone actual have
data on this, by discipline?)

Please, let's leave "creative" writers and the trade literature out of
this. We are talking about the refereed journal literature: no
royalties or fees to authors.

> I am all in favor of authors retaining rights to the intellectual
> property they produce, and I certainly wouldn't object to authors
> posting their own work on their own sites

That is all that is necessary to ensure that user practise (author and
reader) goes on subversively to bring everything to its optimal and
inevitable conclusion. We can do what we like with the fire-walled
S/SL/PPV literature; users will simply choose free access
overwhelmingly, as they are doing with xxx.

> --but I don't think this
> substitutes for uniform search access across an entire journal archive
> or across a whole set of journals: in fact, it throws us back out on
> the Web search engines, where our chances of finding what we're after
> are not very good.

ot so, I'm afraid. All that the search engines need in order to become
perfectly adequate conduits to the literature are simple tags:
+/-REFEREED and +/-JOURNAL-BRAND-NAME. Authors are perfectly capable
of supplying that.

> If the answer to that is to allow people to deposit
> materials on a central server, then we're back to issues of maintenance
> and paid support

xxx, supported by NSF/DOE, could handle not only all of the Physics
literature, but all of the refereed journal literature (all 14K
refereed journals covered in Ulrich's) without even a change of scale
(and I'm betting that that would in fact be the most sensible way to
go, rather than needlessly duplicating noncompeting resources discipline by

> as well as to issues of consistency in the data
> (since inconsistently entered or formatted data will decrease the
> effectiveness of tools like search engines).

Patience. That problem will be sorted out by the pressures of user
practice itself: As xxx becomes the locus classicus, authoring tools
will improve and standardise. No one wants one's articles missed because
they are nonstandardly formatted or tagged.

> If we could start the world over, perhaps this would be a good model,
> but if we have to get from where we are to something better, I think
> it's a very long shot. Another case of "old habits find[ing] no end of
> rationales for perpetuating both themselves and the status quo"? My own
> view is that one is more likely to change the status quo by
> understanding than by dismissing the physics of inertia.

Understanding is what this Forum is for. But perhaps it is better to
subvert than to be inert...

Stevan Harnad
Received on Tue Aug 25 1998 - 19:17:43 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:45:26 GMT