Abstract: | Empirical prediction entails syllogistic reasoning, the major premise being a “proposition of regularity”. This latter can itself be tested empirically, as the hypothesis “Past regularity implies future regularity”. In the social sciences (in contrast to the physical sciences), the deterministic version of this hypothesis is demonstrably untrue. Though the “statistical” version of the hypothesis stands, the likelihood of its truth has never been expressed in probability terms. Thus empirical prediction in economics confronts not only uncertainty, which can be expressed in statistical terms, but an added element of “ignorance” which (at least as yet) cannot be so expressed. |