APPENDIX 11
Memorandum from Professor N J Hitchin,
University of Oxford
What are the consequences of increasing numbers
of open-access journals, for example for the operation of the
Research Assessment Exercise and other selection processes? Should
the Government support such a trend and, if so, how?
The open-access unrefereed archive http://uk.arxiv.org/
(the UK mirror) is a prime resource for mathematical papers. Not
all mathematicians put their stuff there, not all know about it,
but many of us download papers from there as a matter of course,
and a number of RAE submissions were of this nature. Being unrefereed,
the results may be plain wrong (which also happens in journals,
it has to be said) and also the good, the bad and the ugly all
sit next to each other. A journal puts a quality and accuracy
assessment on its contents, but in practical terms as often as
not I still download from the arXiv instead of going to the library
to look at the journal. If the library subscribes to the journal
I will go first to the electronic journal-version, but failing
that go to the arXiv.
From this standpoint a paper accepted by a good
journal has a quality mark placed on it, but then one feels free
to access the web version without hesitation. Some journals try
and stop any further alterations of the arXiv version after acceptance
and modification by the refereeing process but I can't see how
the internet can be regulated to stop the final version appearing
there.
For the RAE, the problem is one of assessing
quality of open-access unrefereed articles; panel members have
no time to do the refereeing job themselves.
The model of open access publication where payment
is made on submission could have consequences for disciplines
such as mathematics which are not big grant-earners and also for
independent researcherslatter-day Einsteins in their patent
offices. It could act as a deterrent to submitting papers to those
journals (which could be good or badthere are undoubtedly
too many journals). If the free arXiv continues, though, then
worldwide dissemination of knowledge would not be affected.
January 2004
|