On the interpretation of intuitive probability: A reply to Jonathan Cohen