For the September Forum:
There was a case in Massachusetts a few months ago of a surgeon who
inexplicably left the operating room for 35 minutes to cash a check
at the bank. About a month later the doctor, David Arndt, lost his
medical license. The patient was unharmed, although he was chagrined
that for the entire month the hospital did not tell him that Arndt had
been replaced as his doctor; Arndt continued to see the patient.
You could say that a case like this doesn't prove anything. Some doctors
are incompetent, some hospitals are second-rate, no harm was done,
and eventually corrective action was taken. This is all true. But you
could also say that the case is so patently ludicrous that Massachusetts
hospitals should do a sanity check, even if it doesn't surprise insiders.
This is how the Bogdanov case looks to many of us in math and physics.
The Bogdanovs made a mockery of the one of the main formal credentials,
journal publication, that most of us must use to get promoted. Granted,
it's no surprise. It's no surprise because I have seen other travesties
of research get published in supposedly good journals. It's also no
surprise in that when I do serious research I don't care where or even
whether papers are published - I can evaluate them myself. Not only
is the Bogdanov stuff of no consequence to me as a researcher, it of
no consequence to anyone who granted the Bogdanovs their credentials.
No one on their thesis committees has coauthored or even cited any of
their papers.
The problem is that outsiders, such as promotion committees and science
reporters, understandably can't evaluate technical papers themselves.
So instead they rely on formal credentials. Obviously when people
get publications and PhDs out of complete nonsense, it cheapens those
credentials.
It is absolutely untrue that the journals and universities that certified
the Bogdanovs are second-rate. That's just a knee-jerk reaction.
It's true that the University of Bourgogne is not quite Harvard and
that some of the journals involved are not top-flight either. Still,
we should expect any such authority to reject utter travesties. The first
journal to publish the Bogdanovs, Classical and Quantum Gravity, has not
only published many good papers, it also responded to the Bogdanov affair
in a first-rate way: by publicly admitting error. Usually journals have
no comment when they publish really bad papers.
To conclude, yes, this is evidence that there is something wrong with
the current implementation of peer review. I have made the same point
before using other examples. Coincidentally I recently proposed some
reforms in this opinion piece:
http://front.math.ucdavis.edu/math.HO/0210144
--
/\ Greg Kuperberg (UC Davis)
/ \
\ / Visit the Math ArXiv Front at http://front.math.ucdavis.edu/
\/ * All the math that's fit to e-print *
Received on Tue Nov 19 2002 - 12:12:06 GMT