Re: [evol-psych] Geertz' fallacy
- A note on Geertz's Fallacy: This has gone by the name of the fagot theory in
parapsychology. A fagot (or faggot) is a bundle of sticks. Each stick
represents some piece of evidence for a proposition, such as "ESP exists."
Individually, each stick has a weakness, but together the fagot cannot be
broken. Although all scientists use this approach to some extent (the
principle of the "preponderance of evidence," I suspect is basically the same
idea) it has an obvious flaw: a lot of bad evidence adds up to bad evidence.
University of Pittsburgh
William George Irons wrote:
> Gene Anderson and Keith Sutherland are exactly right about Geertz and
> interpretive anthropology. Interpretive anthropologists make a virtue out
> of believing what the want. Their anthropology is strongly associated with
> a feeling of self-righteousness. It is morally correct to believe what
> they want to believe and immoral to challenge it. This is a general
> characteristic of true believers including not only Mussolini, but also
> Hitler, Lenin and Stalin. There however also good true believers I
> think. The problem with true believers is that they close off thought. If
> it is for a good cause, you never know when it may morph into a bad cause
> and you're not allowed to criticize it when it does.
> The major reason why the true belief of interpretive anthropologists is a
> scandal is that they are scholars who should be pursuing the truth (which
> can sometime be know), but actually they interfere with the pursuit of
> truth. Scientific methods are the best tool for overcoming true belief and
> other common human biases, but they attack science and not only false but
> morally incorrect. They close off the best method for reexamining their
> own thinking and they encourage others to do the same (with some success).
> Shorter version of Geertz' fallacy: you can prove a point with bad data as
> long as you have a large mass of bad data.
> Probably in Geertz' case we need also not that you can prove politically
> correct points with just about any kind of data (even with pure assertion),
> but you need huge amounts of perfect data to prove a politically incorrect
> point. If you point is politically incorrect any imperfection in your data
> invalids you knowledge claim. If all else fail, imaginary imperfections
> will do.
> I also note that Geertz' was by no means the first to commit what I have
> labeled Geertz' fallacy. However, he get the honor--I would argue--for
> being so explicit in stating it as general principle--with the usual
> interpretive equivocation of course.
> Bill Irons
> William George Irons
> Northwestern University, Evanston, IL. USA