On Mar 24, 2004, at 8:33 AM, Wei Dai wrote
The paper makes the point that what psychology views as mental
diseases in
many cases can be interpreted simply as extreme or unusual preferences,
and in those cases involuntary psychiatric treatment can not be
justified
as a benefit for the patient.
It seems to me that a clear exception may be where there's an extreme
preference to harm others.
This makes sense to me, but perhaps involuntary psychiatric treatment
can
still be justified as a benefit for the younger version of the patient
(i.e., before he became "sick") who presumably had more normal
preferences, and who would have prefered that he be given treatment to
reverse any radical preference changes.
I'm confused. How does one decide whether the younger version's
preferences are more "right" than the elder's?
The issue of how to treat insanity is a thorny one; it's not clear to
me from Bryan's paper, for example, whether medication, psychotherapy,
or locking certain insane people up are all bad ideas. I know Szasz
(and probably Caplan) have definite ideas about the misuse of these
remedies. But I'm not sure how that rules them out as treatment for
the insane. The point well-taken, however, is that what is observed as
insanity is NOT irrationality in the economic sense of the word. For
the insane, not only does thin rationality hold, but insanity is in
many ways closer to very thick rationality, rather than irrationality.
I've discussed this a bit with Eric Crampton, who's suggested that
statistically rare and extreme preferences or beliefs could serve as a
new definition of insanity, and that possibly that's sufficient
justification for their involuntary treatment (correct me if I'm
mistaken on your position, Eric). My concern is that a purely
statistical standard, rather than some standard of harm, is what
allowed Deirdre McCloskey's sister to have her locked up twice, or
could allow psychiatrists to forcibly medicate anarcho-capitalists like
David Friedman and Bryan Caplan.
I think that thin rationality always holds, even for the "insane." If
that's true, then it can be argued that the way to treat insanity is to
raise the costs of extreme preferences. The question is one of how to
raise those costs. It may be that medication, psychotherapy, or some
combination of the two can raise the costs of those preferences. But
how does a younger/elder multiple-selves model justify treating the
"insane" involuntarily? Because the younger's preferences (or beliefs)
are more statistically normal? Then you run into the same problem
mentioned above.
