On Psychomimesis

Abstract

In this paper some aspects of the mind-brain problem, as approached from the standpoint of mimesis, are examined. Such studies are usually prefixed by the adjective "artificial", as in "artificial intelligence"; "artificial life", etc. A key assertion of such approaches is embodied in the familiar "Turing Test"—that two systems which behave "enough" alike are alike; specifically, that a properly programmed finite-state device (i.e. a Turing machine) which behaves "sufficiently" intelligently is intelligent; or, contrapositively, that any system which behaves intelligently can be replaced by such a device. Here, such mimetic approaches are placed into an historical context, and are contrasted with parallel scientific approaches to the same questions. It is argued here that there is no finite threshold, beyond which "enough" commonality of behavior allows us to conclude an equation of causal underpinnings with finitely generated syntax typical of algorithmic devices, and hence that assertions like Turing's Test are false.