On the Logic of the Social Sciences