Real Algorithms: A Defense of Cognitivism

Abstract

John Searle dismisses the attempt to understand thought as a form of computation, on the grounds that it is not scientific. Science is concerned with intrinsic properties, i.e. those features which are not observer relative, e.g. science is concerned with mass but not with beauty. Computation, according to Searle, presupposes the property of following an algorithm, but algorithmicity is normative, by reason of appealing to function, and hence not intrinsic. I argue that Searle's critique presupposes the folk notion of function, which is indeed normative. But this folk notion can be replaced by a purely descriptive analogue, thereby showing that algorithmicity can be construed as intrinsic after all.