The Raven Paradox Revisited in Terms of Random Variables

The discussion about the Raven Paradox is ever-renewing: after nearly seventy years, many authors propose from time to time new solutions, and many authors state that these solutions are unsatisfactory. It is worthy to be carefully noted that though most arguments in favor or against the paradox are based on the notion of ``probability'' and on the application of Bayes' law, not one of them makes use of the Kolmogorov axiomatic theory of probability and on the subsequent notion of ``random variable''. This seems to be due to a preference for the purely logical interpretation of the notion of ``probability'' which allows to assign a probability value to propositions expressing natural (universal) laws. This paper aims at presenting an attempt to contribute a solution to the paradox by the proposal of a different way to express universal natural laws and by the use of a language which does not allow to apply probability to universal statements.