Astronomy and probability: Forbes versus Michell on the distribution of the stars

Summary

James Forbes' critical examination of the probabilistic reasoning, which led John Michell to infer a physical connection between optically double and multiple stars, is analysed. It is argued that despite the interpretations of its nineteenth-century defenders, Michell's reasoning has some force which does not depend upon questionable Bayesian principles. Attention is drawn to some of the ambiguities concerning the notion of randomness, and it is shown that these ambiguities render Forbes' objections less than conclusive.