FROM CASES OF HYSTERIA TO THE THERAPEUTIC SOCIETY
241
I belive Sokal does physics a great service. But I think that you were
perhaps a little too harsh on the editors of the journal, with which I am
not terribly familiar. I hope that they will learn a lesson about dealing with
things they really know nothing about and are not in a position to evalu–
ate. The reason I say you are a li ttle too harsh is due to an article about
fraudulent leukemia research, which was presumably submitted to qualified
referees. At least five articles had
to
be retracted. As a chemist, I believe in
the scientific method, and in peer review. But I think we sometimes do
ourselves a disservice when we don't admit the human fallibilities that are
associated wi th it.
Leon Cooper:
I agree with you, scientists are human beings. When I was
a teenager, my mother read an article in
The New York Times
in which some
scientist was caught in an embarrassing position; she told me it shocked
her. Teenage wise-guy that I was, I said, "so what, everybody does it," and
she said, "but he is a
scientist."
It doesn't seem especially surprising to me that scientists lie, steal, and
cheat the way everybody else does. And before you all, I declare myself
against lying, stealing, and cheating.
But what should we do if a scientist falsifies data. I can assure you none
of us in the game like that, because it can waste time and drive us crazy. The
question is how to deal with it. The worst thing to do is to bring in lawyers
and the FBI. The best thing to do isjust to repeat the experiment, that's all.
Just do it over again. Fraud is not the problem. We have to decide in every
article we read whether the experimenter is reporting exactly what he saw
or what he thought he saw, or what he believed he ought to have seen. We
worry about that all the time. Of course, if someone deliberately falsifies,
that can be aggravating. But it happens rarely and it is self-correcting. If a
scientist reports experiments that no one else can repeat, very soon his rep–
utation sinks below that of fraud-to incompetence.
Gunther Stent:
The issue of fraud did not arise in the Sokal case men–
tioned here earlier. The manuscript submitted by Sokal was a spoof whose
absurdity the editor should have noticed, if he had been a serious scholar.
But the issue of fraud did arise in the Baltimore case because it was alleged
that his co-author, Imanishi-Kari, had not observed the experimental data
they published. The case against Imanishi-Kari was eventually dismissed,
not because her data were vindicated but because on review the evidence
for fraud was found to be insufficient.
Leon Cooper:
I agree with you, as is said in Scottish courts: not proven.
Suppose, Gunther, I claim that you committed fraud in one of the exper–
iments you did twenty-five years ago. The only conclusion one could come
