On Sat, 2 Oct 2004, Ted Winslow wrote:
> Sultan's is also the insight of Foucault. This form of "science" to
> some degree (Foucault himself exaggerates it) hides a sadistic will to
> power. What Foucault fails to notice, though, is that the methods
> themselves are inadequate for understanding their object because they
> canalize the same instinct as the sadistic will to power. The methods
> deny that the object is "alive."
No, you're missing Foucault's point. He is not anti-psychiatry or anti-science, and it has nothing to do with sadism. F's most important insight is that modern power relations work by producing things, not denying them or censoring them. One example: we can better control mass populations by creating criminal deviant types (Discipline and punish). Scientific discourse facilitates this (criminology, etc). All this has nothing to do with "sadistic will to power".
> The exclusive reliance on "experimental" methods
> ignores this by implicilty assuming that the behaving entity remains
> unchanged with changes in its relations. So too do mathematical and
> statistical methods.
I have my concerns about experimentation, but these are nonissues. In an experiment, you can manipulate an independent variable that directly or indirectly involves changes in social relations or dynamics (e.g., assessments of social skills training for people who are shy). How behavior changes in different social settings or in different relationships is another research topic, not a fatal flaw of all experiments.
The skepticism of statistics baffles me. There's dynamic change and complexity in the nonsocial world too; are you just as critical of the use of statistical models in physics? Again, stats can be a tool to model the relational complexity of human behavior; it doesn't have to be used the way you fear it does.
Miles