Computation in mind and world: A realist account of computation in cognitive science

2005   PhD thesis, University of Cambridge

Last updated 13 December 2005

This Ph.D. thesis concerns the truthmakers of the computational theory of mind (CTM). The CTM is the view that some, or all, of our cognitive processes are computations. A potential problem with the CTM is that it is not clear what it means for a system, such as the human brain, to perform a computation. This problem is often glossed over in the literature – computation is assumed to be just like what electronic computers do. Unfortunately, what electronic computers do is far from clear. My thesis addresses this concern. Specifically, it answers the question: under what conditions does a real-world system, such as a human brain or electronic computer, perform one computation rather than another? Searle (1992), Putnam (1988), and Kripke (1982) have argued that the only possible answers to this question are anti-realist. On their view, the computation that a system performs is a function of how human agents interpret that system, rather than of the mind-independent world. If this were true, then the consequences for cognitive science would be dire. Cognitive science would not explain cognition, it would presuppose cognition. In my thesis, I defend cognitive science from these anti-realist attacks. I argue that mind-independent facts can determine the computation that a system performs. My argument consists of two parts. First, I develop a semantics for computation talk in the context of cognitive science. This provides an account of the semantic content of our computational claims in that context. Then, I present an account of the metaphysical facts that make that semantic content true or false. I argue that these metaphysical facts are, at least possibly, mind-independent. The resulting notion of computation is compared to the alternatives of Chalmers (1996), Copeland (1996), and Mellor (1991).

Abstract:

This Ph.D. thesis concerns the truthmakers of the computational theory of mind (CTM). The CTM is the view that some, or all, of our cognitive processes are computations. A potential problem with the CTM is that it is not clear what it means for a system, such as the human brain, to perform a computation. This problem is often glossed over in the literature – computation is assumed to be just like what electronic computers do. Unfortunately, what electronic computers do is far from clear. My thesis addresses this concern. Specifically, it answers the question: under what conditions does a real-world system, such as a human brain or electronic computer, perform one computation rather than another? Searle (1992), Putnam (1988), and Kripke (1982) have argued that the only possible answers to this question are anti-realist. On their view, the computation that a system performs is a function of how human agents interpret that system, rather than of the mind-independent world. If this were true, then the consequences for cognitive science would be dire. Cognitive science would not explain cognition, it would presuppose cognition. In my thesis, I defend cognitive science from these anti-realist attacks. I argue that mind-independent facts can determine the computation that a system performs. My argument consists of two parts. First, I develop a semantics for computation talk in the context of cognitive science. This provides an account of the semantic content of our computational claims in that context. Then, I present an account of the metaphysical facts that make that semantic content true or false. I argue that these metaphysical facts are, at least possibly, mind-independent. The resulting notion of computation is compared to the alternatives of Chalmers (1996), Copeland (1996), and Mellor (1991).

HTML version of full text not available
PDF available above