Theory Assessment and Coherence

Abstract

One of the most important questions in epistemology and the philosophy of science is: what is a good theory and when is a theory better than another theory, given some observational data? The coherentist‟s answer would be the following twofold conjecture: (i) A theory is a good theory given some observational data iff that theory coheres with the observational data and (ii) a theory is better than another theory given some observational data iff the first theory coheres more with the observational data than the second theory. In this paper we show that this answer is a good answer. More precisely, we argue that the coherence measures of Olsson (2002) and Shogenji (1999) are good measures for the purpose of comparing and evaluating theories, i.e. assessing theories. We do so by clarifying the sufficient and necessary conditions every assess-ment function must satisfy for being a good measure of the goodness of theories. Afterwards we will show that the coherence measures of Olsson and Shogenji are indeed good assessment functions for assessing theories.