Re: The True Cost of the Essentials

From: Stevan Harnad <harnad_at_cogprints.soton.ac.uk>
Date: Wed, 3 Apr 2002 15:39:22 +0100

On Wed, 3 Apr 2002, Eberhard R. Hilf wrote:

> It seems that [APS starts] from assuming that the short cut of free access
> full texts available on the web from the author to the reader
> (by either his server, distributed services or central archives)
> is inevitable. Thats fine.

As we will see, the meaning and even the coherence of Ebs's further
questions revolve on resolving a fundamental ambiguity here. Is this
assumption (let us call it Assumption F for Free) assuming free
full-text online access to peer-reviewed postprints or only to
pre-peer-review preprints?

> The future role of the learned societies such as APS...:
> a bundle of services to manage scientific documents:

Nothing hinges on this, but on the assumption of free online access to
all the peer-reviewed full texts, I, for one, if I were a publisher,
would be very reluctant to stake my future on products/services like the
following:

> * intelligent personalized retrieval

Can learned societies really do this better than free research
community service providers? And sell the product/service? (I ask this
in the spirit of openness and realism. Nothing is at stake for
open-access in this, as this is all predicated on the assumption, F,
that full-text access will be free.)

> * crossreference across all sources, including inst.-webservers,

Same question as above.

> * virtual subfield collections and alerting

Same.

> * professional offline refereeing (refereeing after dissemination of the
> documents, and independent of where it resides (even across publishers of
> course); closest to this is the successful 'living reviews of
> Gen.Relativity', although far too slow), that is referees who are experts
> (and might even be paid by APS) to oversee actively their field and the
> new papers and read them and referee/summarize openly.

Ebs is unfortunately smuggling in a little speculation of his own on
this: Are we speaking about offline refereeing of ALREADY PEER-REVIEWED
POSTPRINTS? In other words, is this 2nd-order review? If so, I again ask the
same question as above: Is this something learned society publishers
like APS can do better than the research community can do itself, for
free, and can they sell it?

A variant of this already exists, in the form of open peer commentary
journals such as Behavioral and Brain Sciences http://www.bbsonline.org
and Psycoloquy http://www.ecs.soton.ac.uk/cgi/psyc/newpsy but they
combine the functions of pre-publication peer review and
post-publication peer-commentary. Presumably these can be separated. In
that case the usual question remains (as it does for all the options
above, including the free access itself, to the peer-reviewed
full-texts), namely, how is the peer-review to be paid for?

Presumably there are answers there, e.g.:
http://www.publications.parliament.uk/pa/cm200304/cmselect/cmsctech/399/399we152.htm

But then is the commentary phase any different? Are the
commentaries/reviews peer-reviewed? If so, presumably commentary/review
journals will have the same cost-recovery model as conventional journals.

Or perhaps, in addition to journals for reviewed commentaries/reviews
there can be ad-lib commentaries/reviews, in which case the question
again arises: Is this something that learned societies could provide
and sell, or can the research community implement it for free? (I am
really only asking the questions here, not answering them!)

But, more important, Ebs may have been making the implicit assumption
above that the free full-texts referred to at the very beginning are NOT
peer-reviewed texts, but merely unrefereed preprints. And this ad-lib
post-hoc reviewing service is meant to replace classical peer review?

If so, we are no longer predicating this on the assumption, F, that
open access -- as described, for example, in the Budapest Open Access
Initiative http://www.soros.org/openaccess/ -- is already available.
Rather, we have made a transition from the question of open access
(to the current peer-reviewed literature, such as it is)
into some sort of untested and unspecified agenda for peer review
reform:

"Peer Review Reform Hypothesis-Testing"
http://www.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/0479.html

But in that case, all bets are off, and the subject has been changed
from the sure one (that free online access to the current peer-reviewed
literature is optimal and inevitable) to the hypothetical one (that
peer-review reform may be desirable, and it may be possible to couple
certain new, untested peer-review "reforms" with certain new ways to
access the resulting literature -- whatever that literature may turn
out to be).

> * permanently updating collection of authors tools to help them writing
> and transferring to MathML/XML in a 'state of the art' way.
> [Revtex 4 was in that respect at that time the best tool by concept].
> * etc.

Certainly XML authoring tools will be necessary. (Are learned societies
in the best position to develop and sell them? Each society it own
tools?)

> The business model follows from this:
> * registered users [Members of the society APS or associated societies
> (DPG) for free, who paid with their fee: thus contracts with the other
> major societies (how much per member would be an estimate?]
> * registered Institutes, Libraries
> * anyone from Industry by registration (highest fee,..).

Note that these are all questions about selling and recovering costs for
these hypothetical added services. I have no idea whether such services
are needed, are best provided by learned societies for a fee rather than
the research community for free, and sustainable. But the big question
is still about the open-access full-texts: Are we assuming them to be
peer-reviewed (as they are now) or not? And if not, what ensures that we
are even talking about the same literature?

> Example: we as a small physicist group of 10 pay at present 1.000,-$ per
> year and person for computer programmes, and a virtual share of about
> 100,-$ per year and person to the library for journals, and about 50,-$
> per year per person for society membership fees. We always use the
> computers, but miss a surrounding to use MAthML-Physics etc., never go to
> the library since the web exists, and see virtually nothing from our
> society (apart from the 17 docs per year of NJP..).

This is unfortunately not helpful. It describes a status quo, but it
does not specify the essentials on which this status quo depends (e.g.,
peer review) and how those essentials are to be paid for once full-text
access (to the peer-reviewed draft, presumably) is free. And it leaves
out completely the researchers whose institutions cannot afford all of
this.

(Are the essentials to be paid for by the options? But what if
institutions opt out, and access the open-access version anyway? What
about the Prisoner's Dilemma, or the "Tragedy of the Commons"?)

> Estimate for the above mentioned services 100,-$ per year and person,
> would add in physics to more than APS and those few other physics
> societies (IoPP?, JPS?, not DPG) who are willing to serve and improve
> those services would ever need? [500.000 physicists worldwide make up thus
> for 50 Mill. $ per year, enough?]

This is all too confusing and vague. What, precisely, is being
contemplated to pay to whom for what, under what conditions?

> Transition period: charge the printed copies of APS much higher, and
> reduce the online versions gradually to zero.

That's fine for phasing out paper subscriptions in favor of online
licenses, and perhaps also for phasing out online licenses in favor of
add-on services, while the free peer-reviewed text is free. But then who
is paying for what? Do you imagine that the add-ons, by subscription,
paid by those institutions that can afford it, will continue to cover
the cost of the essentials (e.g. the peer review) even after the
peer-reviewed full-texts are accessible to everyone for free?

Both the utility and the causality here seem to me unclear, if not
incoherent. Can you give a much more explicit map of the specific
services/products, cause/effect relationships, and costs/benefits you
have in mind here?

Stevan Harnad
Received on Wed Apr 03 2002 - 15:41:17 BST

This archive was generated by hypermail 2.3.0 : Fri Dec 10 2010 - 19:46:29 GMT