From: Stevan Harnad <[log in to unmask]>

Physicists have been spontaneously self-archiving in Arxiv since 1991,
but most other disciplines have not followed suit, despite the
demonstrated benefits of providing open access in terms of research
uptake, usage and impact.

It is for this reason that research funders and institutions worldwide
are (at last) beginning to mandate (i.e., require) that their fundees
and faculty self-archive.

For open access mandates to work, however, it has to be possible to
systematically monitor and verify compliance.

Not all research is funded (and there are many different research
funders); but virtually all research comes from institutions
(universities and research institutes), most of which now have
institutional repositories for their researchers to self-archive in.

Institutions are hence the natural (and eager) partners best placed to
fulfill the all-important role of monitoring and ensuring compliance
with the requirements of their own researchers' grant requirements,
via their own institutional repositories. (This also gives
institutions the incentive to adopt open access self-archiving
mandates of their own, for all their research output, funded and
unfunded, in all disciplines.)

Researchers, in turn, should only need to deposit their articles once,
institutionally -- not willy-nilly, and multiply, in diverse
institution-external repositories.

The solution is simple, since all open access repositories are
interoperable, meaning they share the same core metadata-tagging
system, and hence each institution's repository software can
automatically export its metadata to any other institution-external
repository desired.

That way researchers need only deposit once, in their own
institutional repository; institutional and funder open access
mandates areconvergent and cooperative rather than divergent and
competitive; and mandate compliance can be reliably and systematically
ensured by the author's institution.

So Biorxiv is a welcome addition to the growing list of disciplinary
repositories for centralized search and retrieval, but deposit in
Biorxiv should not be direct: researchers should export to it from
their institutional repositories. (Biorxiv can also harvest from
institutional repositories, just as Google and Google Scholar do.)

Biologists and biomedical scientists, unlike physicists, do not have a
culture of spontaneous self-archiving. Hence open access mandates from
funders and institutions are needed if there is to be open access to
their research. And those mandates have to be readily complied with;
and compliance has to be readily verifiable.

So let us not lose another quarter century hoping that biologists will
at last do, of their own accord, what Arxiv users have already been
doing, unmandated, since 1991. In 1994 there was already a "Subversive
Proposal" -- unheeded -- that all disciplines should do as the
Arxivers had done. Harold Varmus made a similar proposal ("e-biomed")
in 1999, likewise unheeded.

Let us start getting it right in 2013, the year that funders in the
US, EU and UK have begun concertedly mandating open access, along with
a growing number of institutions worldwide. But let us harmonize the
mandates, to ensure that they work:

Arxiv has certainly earned the right to remain the sole exception,
insofar as direct deposit is concerned, being the only
institution-external repository in which authors have already been
faithfully self-archiving, unmandated, for almost a quarter century:

For Arxiv, institutional repositories can import instead of export.
But for the rest: Deposit institutionally, export centrally.