Sanjay Chandrasekharan Abstract
Institute for Interdisciplinary Studies, Carleton University, Ottawa, Canada; and Pre-doctoral research fellow,
Center for Adaptive Behavior and Cognition, Max Planck Institute for Human Development, Berlin, Germany
Supervenience is a fundamental concept for non-reductive physicalist theories of the mind (theories which hold that the physical level is the fundamental level of understanding, but which also hold that the mind cannot be reduced to the brain or any other physical level). Most computational theories of the mind belong to this category. The idea of supervenience applied to the philosophy of mind says that mental properties (M) cannot be reduced to physical properties (P), but M is dependent on P in such a way that all changes in M are changes in P, but not vice versa. In this paper, I outline two kinds of such supervenience relationships and the problems they face in explaining the mind-body problem.
I then argue that the second kind of supervenience, the more plausible version, is in conflict with multiple realizability, the idea that the mind can be instantiated by things other than the brain, which is the philosophical basis of all Artificial Intelligence (AI) efforts. A possible way of saving both the supervenience relationship and multiple realizability is then suggested, taking inspiration from the Indian philosophical concept of autoreflexivity of awareness (svasamvedana).