Newton’s Invention of the Problem of Induction

Abstract

Why is it that Locke, who possessed a full ontology of contingent causality, never raised the problem of induction, as Hume would do only when, within four decades, the scientific revolution would be over? I will suggest now that the reason for this puzzling fact is that the problem of induction is a detrimental difficulty only for the aristotelian, but is a standard feature of science for the platonist. If it is the problem of finding logically necessary connections between phenomena, then the 17th century platonist dismisses this as a non-problem, since there are no such connections in a world created by the Judeo-Christian God. If the problem of induction is that of discovering the unobservable causal layers of the world, then he declares this to be a difficulty so obviously unimportant in comparison with the irreducible irrationality of an informative world as to make any worry about it ridiculous. For science is necessarily hypothetical and the possibility of error is one of its necessary features not because of unobservability of micro structures at all, but because of its informativity. In so far as the problem of induction is identical with the problem of possible error about the identification of necessary links in nature, it is dismissed as a problem at all, since it is logically non-soluble and only soluble problems are really problems.