Conférence « Ockham’s Razor and the Mind-Body Problem »
Elliott Sober, University of Wisconsin, Madison
La conférence est co-organisée avec le Réseau montréalais de philosophie des sciences, le Concordia’s Philosophy Department et le Centre de recherche interuniversaire sur la science et la technologie.
Philosophers often view the competition between the mind/brain identity theory, functionalism, and dualism as nonempirical, since all three theories are compatible with the discovery that a mental property and a physical property are perfectly correlated. Given this, it has been suggested that parsimony considerations can justify the choice of a mind/body theory even though observations are unable to do so. In this talk I’ll describe an epistemological framework used in science (AIC, the Akaike Information Criterion) in which parsimony is relevant to estimating a theory’s predictive accuracy. AIC cannot show that the simpler of two predictively equivalent theories is better. However, the old idea that the identity theory is more parsimonious than dualism turns out to be right within that framework (and functionalism turns out to be intermediate in its parsimoniousness). The three mind/body theories are general and abstract; when they are applied to a given pair of mental and physical properties, the result is three specific theories about those two properties. These specific theories are predictively nonequivalent, and AIC applies to them.