Novel Predictions as a Criterion of Merit

Abstract

Any contribution to a volume in honour of Imre Lakatos should, I suppose, be polemical. I have selected the question of novel predictions as a criterion of merit of theories because it brings out a difference of views which is, in my opinion, more important than the shift from ‘theory’ to ‘research program’. The difference I am referring to is between what might be called ‘objectivist’ philosophy of science and what might be called the ‘sociological’ or ‘externalist’ approach. The former view is ‘internalist’ in that its criteria of evaluation are internal (i.e., refer to the theory itself), except that it also uses the criterion of ‘agreement with experience’. The latter view introduces psychological, sociological or historical criteria (possibly in conjunction with criteria acceptable to an objectivist).