The “raven paradox” in the light of probability theory

This paper deals with the well-known raven paradox, also named the paradox of black ravens, one of the most renowned paradoxes arising from the confirmation procedures for universal statements of empirical theories. Its aim is to show that the paradox owes its origin to a number of very drastic positions about the form in which such universal statements should be expressed, and the confirmation value the indirect observations (i.e., observations not explicitly carried out on ravens) should be deemed to have. More specifically, we suggest that the use of the formal language of calculus of predicates (and, correspondingly, of set theory) is uncorrect in the framework of empirical theories, and that, when using only statements about probabilities of events rather than about probabilities of truth of statements, and choosing a suitable confirmation procedure based on the law of large numbers, a comparison between the confirmation values of direct and indirect observations becomes clear and the paradox, at least as it has been described, disappears.