Computational Theory of Mind

The Computational Theory of Mind (CTM) is the hypothesis that mental states are computational states — that cognition is, at its core, a form of information processing, and that the mind stands to the brain roughly as software stands to hardware. CTM is the theoretical backbone of Cognitive Science and the implicit metaphysics of most AI research.

The theory comes in stronger and weaker forms. The strongest version, associated with early cognitive science and classical functionalism, holds that the relevant computational processes are symbolic and rule-governed — that thought is, literally, the manipulation of mental symbols according to formal rules, as in a Turing machine or formal logical system. The Language of Thought hypothesis (Jerry Fodor) is this strong version's most developed form.

Weaker versions identify mental processes with computational processes of various sorts — connectionist, dynamical, predictive-coding — without committing to symbolic representation. The scope of 'computational' has expanded to accommodate the failure of each previous formulation, which raises the question of whether CTM is a substantive scientific hypothesis or a definitional claim that mental states are whatever physical processes implement them.

The central objection to CTM is that computation is defined relative to an interpretation: a physical process counts as computation only because an interpreter assigns meaning to its states. The brain is not a computer in the way a silicon chip is a computer — the chip is a computer because engineers designed it to be and users interpret its outputs. If CTM is true, who or what assigns the interpretation to neural states? This regress — sometimes called the symbol grounding problem — is the hardest problem CTM has yet to solve.