« 次回POB | トップページ | 3月6日POB »

2008年2月13日 (水)

Graduate Seminar

2月19日(火)での院生ゼミで話すことになったので,その予告です.
時間は16:00-,場所は哲学研究室横のPriest客員教授研究室(予定)です.
ご都合がつく方はぜひどうぞ.

Speaker:
Jun OTSUKA

Title:
A better inference to the best functional explanation

Why does the Stegosaurus have heavy plates on its back? What was its
function? Inference of function from fossil records is a key component
of paleobiology (study of extinct organisms). The difficulty of this
task comes from the fact that we have no way to "see" the actual use
of the trait which remains only as a fossil record. To tackle with
this problem, paleobiologists adopt a form of "inference to the best
explanation" to decide which functional hypothesis is to be accepted.
Function determines form. So the functional hypothesis which predicts
the form closest to the actual structure must be the right story.
However, this method (called "Paradigm Method") has been strongly
criticized, for lacking any formal criterion to measure "similitude"
between the prediction and the actual structure, the result of the
test tends to be highly qualitative or subjective.
In this talk I want to propose a new method of testing functional
hypothesis, based on the Bayesian probability calculus. Evaluating
plausibility of each hypothesis by looking at multiple traits rather
than one, this treatment will reduce the difficulty of measuring the
exactness of prediction and give a clearer result. Moreover, based on
the Baysian account of unification (cf. Myrvold 2003), this method
provides a way to evaluate the correlation among several functions as
well as the degree of developmental constraints, all of which were
entirely missing in the previous paradigm approach.

|

« 次回POB | トップページ | 3月6日POB »

コメント

コメントを書く



(ウェブ上には掲載しません)




« 次回POB | トップページ | 3月6日POB »