To: "Harold Varmus (NIH Director)" <execsec1 who-is-at od.nih.gov>
This is a reply to the objections voiced by The American Association of
Immunologists to the E-biomed proposal.
http://www.nih.gov/welcome/director/ebiomed/com0627.htm#aaoi185
Similar objections have been raised by others in the E-biomed
discussion; the following replies apply to those too, mutatis
mutandis.
> The American Association of Immunologists, June 23, 1999
> Jonathan Sprent, M.D., Ph.D. President
> Frank W. Fitch, M.D., Ph.D. Editor-in-Chief, The Journal of Immunology
> M. Michele Hogan, Ph.D. Executive Director
>
> Peer-Review: First and foremost, we find that this proposal compromises
> the cornerstone of scientific method: peer-review. The process
> described in your proposal is vague, but if taken at face value it does
> not ensure a rigorous peer-review process. Without this we compromise
> our excellence, (at best) and (at worst), pose potential harm to the
> scientific community as well as the public at large. Furthermore,
> scientists depend on the current peer-review process to give their work
> legitimacy and guidance; they do not want to be held to lesser
> standards.
It is correct that the E-biomed draft proposal is vague and somewhat
ambiguous on some points, but this is all easily remediable. Indeed, a
simple remedy was already proposed in the first set of comments on the
draft:
http://www.nih.gov/welcome/director/ebiomed/com0509.htm#harn45
The remedy is to make it clear that the Archive is intended for the
SELF-ARCHIVING of the refereed biomedical journal literature by its
authors, in the first instance. It is not meant to be a journal; it is
certainly not meant to provide peer review; nor is it meant to bypass
peer review. If all authors self-archive their peer-reviewed articles,
it is evident that peer review is in no way being compromised or
sacrificed.
Once the archive has established its raison d'etre in this way, a formal
relationship with journals will also be possible, in the form of official
journal overlays "authenticating" the authors' self-archived drafts,
as has evolved in the case of the Los Alamos Physics Archive and the
American Physical Society, publisher of the most prestigious and
highest-impact journals in Physics.
http://www.cogsci.soton.ac.uk/~harnad/Hypermail/Author.Eprint.Archives/0006.html
Initially, however, there is no need for this official overlay, nor can
such official relationships be established before the Archive itself is
established, with a body of contents that the user community puts to
heavy daily use, as it does in the case of the Los Alamos Physics
Archive, the model for E-biomed:
http://xxx.lanl.gov/cgi-bin/show_weekly_graph
http://xxx.lanl.gov/cgi-bin/show_monthly_submissions
In addition to self-archived peer-reviewed eprints, E-biomed will, like
Los Alamos, contain self-archived non-peer-reviewed preprints as well.
Because such unrefereed papers in biomedicine (unlike physics) could
conceivably pose public-health risks, E-biomed has proposed a novel peer
"filter" to screen them. This filtering is not a substitute for peer
review, it is a supplement to it, in the NON-peer-reviewed domain of
self-archived preprints.
This ambiguity too, is easily resolved.
> Please do not dismiss out-of-hand a process that has taken 300 years of
> trial and error to evolve...
These admonitions seem entirely unnecessary as the E-biomed proposal
takes pains to make it quite explicit that it is not dismissing peer
review!
Moreover, it is a false opposition to imply that the central goal of
E-biomed, which is a free, self-archived version of BOTH the
peer-reviewed and pre-peer-reviewed literature (clearly and
unambiguously tagged as one or the other) can only come at the cost of
compromising peer review in some way. It is quite clear that one can
have both a peer reviewed literature AND a self-archived, free version
of it: One need not sacrifice one for the other.
> Creation of a Monopoly: We are concerned that the proposal would create
> a monopoly. The publishing world successfully operates on free-market
> principles and there is no evidence that monopolies guarantee a better
> product at a lower cost in any market. There may be cause for the
> government to supply a product or a service if there is evidence that
> free markets are unable to do so. But this is not the case in
> scientific publishing. The federal government has the prerogative of
> perpetuating established monopolies, e.g., national defence, but they
> have seldom replaced a successful free-market effort. A single review
> and publication source offers no options to investigators. Has the NIH
> considered the risks of creating a monopoly to replace a diverse and
> successful enterprise?
No monopoly whatsoever has been proposed. The Archive is not intended
to be a Mega-Journal or set of Mega-Journals, replacing the established
journals. It is intended to be a reliable, permanent, free repository
in which the authors of the articles in the established journals can
make their research available online for everyone for free.
Now there is no doubt whatsoever that this service will force the
established journals to restructure themselves in certain ways. (My own
prediction would be that it will make journals scale down to providing
only the service of peer review and authentication, and that this
service will be paid for on the author-institution end instead of the
reader-institution end, but THAT is for the market to decide. E-biomed
merely provides authors with an infinitely more powerful and useful way
of distributing their peer-reviewed findings to everyone for free.)
> Conflict of Interest: In step with creating a monopoly is what we find
> to be a conflict of interest in NIH becoming a sole, centralized
> publisher. That is, that the funding agency charged with carrying out
> the assessment of the scientific accomplishment of an investigator now
> also carries out one of the most important signifiers of that merit:
> publishing. Additionally, it seems that rather than stimulating
> experimentation, oversight of both functions by one organization could
> narrow the scope of what is "acceptable" science.
Again a misreading (though, again perhaps resulting from some of the
vagueness and ambiguity mentioned already, but easily resolved):
E-biomed is not intended to be a publisher but a free Archive of both
the refereed and the unrefereed literature. Journals will continue to
be the publishers -- but their role may well shrink to that of
providing and then certifying the quality control. The rest of
publishing will vanish (for refereed journals).
> Service To the Community: The JI is a major scientific publication
> ranking in the top 10 percent of peer-reviewed publications
> concentrating on immunology. Approximately 9,000 subscribers receive
> 13,000 text pages annually in both print and electronic formats. Almost
> 4,000 new manuscripts are received per year, requiring about 12,000
> reviews. The peer-review process and the quality of data published is
> overseen by a 60 member editorial board and a dedicated in-house staff.
> We continue to acquire the latest in technologies that employ the
> Internet and enable rapid delivery of information to the community in a
> cost effective manner without jeopardizing the peer-review process nor
> diminishing our standards of quality. This is most certainly the goal
> of major scientific publications and we believe we have accomplished
> this it very well.
All of this is irrelevant. JI is not at risk if it indeed intends to
provide essential services to the biomedical community. Those essential
services consist of implementing the peer-review of the literature, and
certifying those final drafts that successfully pass that filter as
accepted and published in JI.
But now we face the real heart of the potential conflict between
E-biomed and the journals: Are the journals prepared to try to prevent
their authors from self-archiving their refereed papers? If so, on what
grounds? What service is being provided to the biomedical community by
preventing authors from distributing for free what they have given to
the journal for free (unlike authors of, say, books or magazines, who
provide their texts in exchange for a fee or royalties)?
http://www.cogsci.soton.ac.uk/~harnad/Hypermail/Author.Eprint.Archives/0006.html
THIS is the real question to confront, and, I believe, the real
motivation behind the attempts to find fault with the E-biomed
initiative on "free market" grounds: Let us face directly the question
of whether this very special and unusual subdomain of publishing -- one
in which the authors themselves want only the eyes and minds of their
readers, not royalties or fees for their articles -- has any
justification for trying to continue to hold this literature hostage.
We must ask what advantage SCIENCE derives from having its users denied
access to its research findings, when the researchers themselves wish
to give them away. This is the REAL conflict of interest lurking behind
this entire question, and it needs to be brought out.
For a discussion of the "Faustian bargain" at the heart of all of this,
see:
Okerson, A. & O'Donnell, J. (Eds.) Scholarly Journals at the
Crossroads: A Subversive Proposal for Electronic Publishing.
Washington, DC., Association of Research Libraries, June 1995.
http://www.arl.org/scomm/subversive/toc.html
> The JI has one of the highest acceptance rates among the larger
> journals and it is still only 40%. This leaves thousands of manuscripts
> not accepted for publication in the JI. We consider the publication, or
> even posting, of these unaccepted manuscripts to be clutter and a
> disservice to busy scientists who rely on expert endorsement to present
> the progress of their field.
This is a separate issue, and can be discussed separately, and examined
empirically. It concerns the non-peer-reviewed sector of the Archive.
It is likely that clearly tagging that sector as non-peer-reviewed (and
the peer-reviewed sector as peer-reviewed, with journal name, etc.) will go a
long way toward allaying such concerns, and E-biomed's proposed extra
peer-filter superimposed on top of that will go still further. (In
principle, all unrefereed papers in E-biomed could be made retrievable only
with an accompanying "health warning," if that was really deemed
necessary.)
But at this point dwelling on the unrefereed preprint sector is a red
herring. Launch E-biomed exclusively for self-archiving the
peer-reviewed literature, if we like, and meanwhile work out the
details about the preprints. They need delay nothing.
> On another level, scientists depend on the hierarchy of journals to
> help them select the most important studies in the plethora of
> information available to them. It is unclear how a single information
> source would assist this sorting process.
And this hierarchy will remain intact, with a clear tagging hierarchy
in E-biomed, including JI.
> Logistics As we stated earlier, the JI alone receives almost 4,000 new
> manuscripts a year for review and 60% of those are reconsidered. Unless
> some other screening event is put in place, this will continue and
> likely increase without barriers such as manuscript submission fees,
> page charges, and rigorous peer-review. How would the NIH handle the
> volume from thousands of scientific journals in a timely and quality
> assured manner? What are the personnel requirements to carry out this
> effort? What sort of editorial oversight would be required? Would this
> oversight be voluntary or would professional level FTE's be required?
> The lack of details in the proposal regarding the logistics make it
> impossible to assess your plan.
This was indeed the heart of the vagueness and ambiguity of the first
draft, but the answers should be quite clear now: The established
journals will continue performing this function, funded by the current
sources of reader-institution-end access-tolls --
Subscription/Site-License/Pay-Per-View (S/L/P) -- until such a time as
the free versions in the archive capture the user market as they have
in Physics, and S/L/P cancellations begin to make themselves felt. At
that time an alternative way to recover the much scaled-down costs of
providing only the service of peer-review and certification will be
author-institution-end publication-charges, provided by authors'
institutions out of only a small portion of their annual S/L/P savings.
http://www.cogsci.soton.ac.uk/~harnad/nature.html
So this too is a false opposition.
> Budget for E-Biomed: The proposal states that scholarly publishing is
> costly. The implication is that, somehow, these costs are an
> unnecessary barrier and can be mostly eliminated if only all
> publication took place electronically. As much as we wish it were not
> so, the reality is there are large associated costs in scientific
> publishing regardless of the medium; at best your proposal appears to
> only shift, rather than eliminate, cost.
This is a controversial issue, much discussed, for example, in
the American Scientist's ongoing September-Forum:
http://amsci-forum.amsci.org/archives/september-forum.html
Suffice it to say that no publisher has yet done a realistic estimate
of what it would cost to provide only peer review, dropping all other
functions, and all expenses associated with them.
At the very worst, the answer could be: not a penny less -- in which
case the reader-institution budget would be entirely shifted to the
author-institution budget in exchange for freeing the literature for one
and all. But the truth is probably much closer to a saving of 70% or
more, in which case the institutions too (and not only all authors and
readers, and research itself) will be much better off.
> However, our real concern is there is no substantive budget to support
> your proposal. We request that the NIH carry out due diligence and
> provide a realistic, detailed budget analyzing the start-up costs
> including personnel, infrastructure support, outsourced editorial
> support, hardware, development of software, and redaction as well as a
> projection for continued costs and support for assurance of longevity
> of the project and archiving the acquired information.
The budget need be only for the provision of a reliable, permanent
self-archiving facility, like Los Alamos. The cost will prove to be
remarkably modest, as the Los Alamos experience already reveals.
This too was a false opposition, making it appear that the free archive
can only be had if NIH pays the equivalent of all biomedical journals'
full current operating budgets, whereas in reality the cost of
providing a permanent, free, global archive is infinitesimal, and will
also force a downsizing of journals, whose remaining essential costs
will be recoverable out of institutional S/L/P savings.
> Support in Perpetuity: The final effect of this proposal, were it
> successful, would be to destabilize existing publishers. If truly
> successful, this would eliminate journals as we know them today.
> Without the backup of these journals, the NIH would have to carry on
> scholarly publishing in perpetuity and commit to this venture
> regardless of the funding circumstances befalling the NIH. In times of
> limited budget commitments to the NIH, how would the financial support
> of this venture be assured?
Nothing of the sort. Initially, E-biomed will make it possible for
authors to self-archive their refereed research findings for one and all
for free. Once that free Archive starts to cut into S/L/P revenues,
publishers will downsize and eliminate obsolete and inessential
services. Quality control will continue to be an essential service, however,
and its costs will be recoverable from S/L/P savings by abandoning the
reader-institution-end trade model for an author-institution-end
publication-charge model that is far more suited to this very anomalous
and special form of literature: the refereed research corpus.
--------------------------------------------------------------------
Stevan Harnad harnad who-is-at cogsci.soton.ac.uk
Professor of Cognitive Science harnad who-is-at princeton.edu
Department of Electronics and phone: +44 2380 592-582
Computer Science fax: +44 2380 592-865
University of Southampton http://www.cogsci.soton.ac.uk/~harnad/
Highfield, Southampton http://www.princeton.edu/~harnad/
SO17 1BJ UNITED KINGDOM ftp://ftp.princeton.edu/pub/harnad/