Harnad reply to Odlyzko

Subject: The Separation of Archiving from Peer Review

That was only one of the many brilliant initial decisions Paul made, consciously and unconsciously, most of them contributing to the colossal success of his Archive. (In fact, the only BAD initial condition Paul introduced that I can think of -- and it is easily remedied -- is the cowboy front-end: The service was so stupendously valuable that the world physics community took to it despite this, but now it's time to give it a major face-lift, as the revolutionary capability extends from the Wild Wild West to the more civilized and conservative environs of other disciplines... The CalTech Consortium and the E-biomed Initiative will, en passant, no doubt do this. But apart from that trivium, Paul has shown the way, the whole way.)

Paul has eo ipso shown that REprint distribution can be handled very inexpensively. This is an absolutely critical point, and just as important a consequence of dissociating peer review and distribution -- indeed it will prove to be far more important, once its implications for the continuum from preprint to reprint are properly appreciated. (That is the essence of my Subversive Proposal.)


To put it another way: The Archive needs to be separate from peer review, but its CONTENTS can certainly be both peer-reviewed and not. This is absolutely central, and needs to be understood clearly.

But because of the unique and extremely nonstandard reward structure of this very peculiar and anomalous subset of the world of publication -- namely, refereed journal authors GIVE AWAY their papers: they do not seek fees or royalties, but eyes and minds -- because of this, the simple subversive act of self-archiving (of both preprints and reprints), once it is performed LANL-scale by all disciplines, will be sufficient to radically restructure the entire domain. This is why it is so critically important to provide the Archive, and to promote the rationale for using it.

Fortunately, however -- and this is the gist of my recommendations for both the CalTech and the Ebiomed Proposals -- this theoretical difference of opinion about what is the optimal future form of peer review has no bearing whatsoever on the Archiving initiative.

A free, public archive of the entire peer-reviewed literature, such as it is today, would be a monumental asset to research and researchers in all disciplines. No one disputes that. (Not even the publishers who say it is impossible!)

Hence archiving initiatives should not be yoked to any particular hypothesis about the future course of peer review; for if the hypothesis should prove false (or even if the status quo should prove recalcitrant to the hypothesis) it will needlessly delay the availability of a resource that we all agree would be optimal right now!

So we can agree that it's good to separate peer review from distribution without insisting that archives should commit themselves to any particular form of peer review: They should be above the fray, providing a means of archiving unrefereed preprints, reprints refereed by classical peer review, and reprints refereed by experimental peer review, alike. The Archive should not commit itself to any particular from of peer review.


I think the usage and citation patterns for LANL already constitute overwhelming evidence (if the mere reflection, by any author -- that the reach and impact of their work could only be greater if it were available online, for free, on the desk of every potential reader -- was not already "evidence" enough!):



Stevan Harnad harnad@cogsci.soton.ac.uk

Professor of Cognitive Science harnad@princeton.edu

Department of Electronics and phone: +44 1703 592-582

Computer Science fax: +44 1703 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/