Tuesday, 26 January 2010
Similarity and Induction
Susan Sterrett, Duke University
12:05 pm, 817R Cathedral of Learning
Abstract: Discussions of induction in philosophy of science generally address the problem of making predictions about the properties of entities, i.e., generalizations of the form "All Fs are Gs" or predictions about particulars of the form: "a is F." The way that similarity is involved in the problem of induction on this characterization of induction is that when there is some F such that both "a is F" and "b is F", it is said that a and b are similar.
In this talk, I want to examine the assumption of similarity involved in the usual formulation of the problem of induction. I will not be questioning that similarity is the basis for induction; I will be challenging the usual assumption of what similarity is. I will examine several examples in science in which similarity is used to make inductive inferences, and show that a better characterization of similarity is called for.