Propensities and Probabilities

Abstract

In 1953 or thereabouts, Karl Popper brought forth into the world a new interpretation of probability: the propensity interpretation. According to this view, probability represents something objective and empirical, as it does on frequency or limiting frequency interpretations of probability; but it is a property not of a sequence or class of trials, but rather of a set of circumstances. It is a theoretical property of something (of exactly what we shall have to consider in due course), and more than that it is an explicitly dispositional property. In recent years quite a number of philosophers (Mellor, Levi, Hacking, Gillies, Fetzer, among others) have become persuaded that in principle this is the right approach to interpreting probability. Not only that, in recent years the propensity view of probability has been attributed, retroactively, to such writers as Cramér and Braithwaite and Hempel. There is, to be sure, a ‘theoretical counterpart of relative frequencies’ view that may be attributed both to Cramér and to Braithwaite (among others) as well, I think, as to Levi; whether this attribution is correct, and whether it is appropriate to characterise it as a brand of propensity theory is a question that would involve me in historical and terminological questions I should prefer to avoid.