A critique of the bayesian solutions to some paradoxes of confirmation irrelevant conjunction hempel s paradox and old evidence
