MARK SIEBEL AGAINST PROBABILISTIC MEASURES OF COHERENCE ABSTRACT. It is shown that the probabilistic theories of coherence proposed up to now produce a number of counter-intuitive results. The last section provides some reasons for believing that no probabilistic measure will ever be able to adequately capture coherence. First, there can be no function whose arguments are nothing but tuples of probabilities, and which assigns different values to pairs of propositions {A, B} and {A, C} if A implies both B and C, or their negations, and if P(B)=P(C). But such sets may indeed differ in their degree of coherence. Second, coherence is sen- sitive to explanatory relations between the propositions in question. Explanation, however, can hardly be captured solely in terms of probability. Why does the proposition ‘Tweety is a bird’ fit ‘Tweety has wings’ much better than ‘Tweety cannot fly’? Because, one might argue, the probability that Tweety has wings, given that it is a bird, strongly exceeds the probability that Tweety cannot fly, given that it is a bird. Presumably, such considerations have led some philosophers to de- velop probabilistic measures of coherence. Tomoji Shogenji, Erik Olsson, Branden Fitelson, and Igor Douven and Wouter Meijs have presented functions which take as input certain probabilities relating to the propositions in question to calculate from them a number which is supposed to represent their degree of coherence. 1 In Sections 1–5, I point out some difficulties with these proposals as well as with Luc Bovens and Stephan Hartmann’s coherence quasi-ordering. However, even if these specific accounts prove to be deficient, this does certainly not mean that the notion of a proba- bilistic measure of coherence is fundamentally wrong. After all, this research project is still in its infancy, and it thus seems justifiable to adopt a ‘don’t worry it’ll sort itself out’ attitude. In opposition to this attitude, the last section provides some grounds for believing that no probabilistic function will ever be able to adequately capture coherence. Erkenntnis (2005) 63:335–360 Ó Springer 2006 DOI 10.1007/s10670-005-4002-4