The Infinite Resourcefulness of Human Intelligence

Abstract

We have already noted that even if a behavioural capability couldbe achieved mechanistically it far from follows that the capability in question is mechanistically determined in man. That arithmetical calculations can be mechanised in no way suggests that the calculations carried out by school children are mechanistically determined. Indeed, the long uphill struggle which many quite intelligent children have to go through before gaining a mastery of the basic arithmetical operations suggests rather that their arithmetical calculations are not mechanistically determined! My purpose now, though, is to argue that the kind of general understanding of things that people show in their use of language could not be achieved mechanistically. Computers can easily beat even the most gifted calculating prodigy, but no computer, however fast its switching operations or numerous its components, could be programmed to use language in the way people can in daily life. My argument for this is a development of one put forward by Descartes in his Discourse on Method, to which I shall advert later. Clearly, if man's language using ability could not be mechanised then, a fortiori, human speech is not mechanistically determined: Laplace's demon would be unable to compute what anybody would say next.