ABSTRACT
How shall we construe Russell's assertion that attention to finite frequency models alone provides an adequate interpretation for the calculus? In one way that is true: axioms I and II hold for the finite frequency models. The fraction ofA U B in K, ifA and B do not overlap, is clearly the sum of the fractions ofA in K and B in K, as axiom II says, for example. But there are many possibilities left open by the axioms that are not excluded by the finite frequency interpretation. For example,
P(A) = ~! is logically absurd if construed as an assertion about a finite frequency model.
