Abstract

The Aristotelian tradition in science, dominant before the advent of modern science, saw real modalities in nature: necessity, possibility, contingency, potentiality, and essence. Throughout the modern period and the early twentieth century, empiricists struggled to maintain that there was nothing to be found between matters of actual fact on the one hand and relations between ideas or words on the other. Probability has the logical form of a modality, but until the twentieth century, it could be construed as a measure of ignorance, or of actual proportions and frequencies, not a basic parameter in nature itself. That changed with the advent of quantum mechanics and the admission of indeterminism as a theoretical possibility. The legitimacy of irreducible probabilities was admitted in the modelling of natural processes. The challenge to the empiricist now is to interpret probability as it appears there, and this includes the larger challenge of displaying an empiricistically acceptable role for modalities in general in the scientific image. This chapter provides a systematic critique of frequentist and propensity interpretations of physical probability, but also characterizes the models of probability theory as constructible from families of sequences constrained by limiting frequencies. This formal agreement with a certain kind of ’ideal frequency’ account is embedded in a philosophical view of modalities in general that turns on an account of the language of science on the level of pragmatics.