Response to Yellman & Murray's comment on ‘The meaning of probability in probabilistic risk analysis’

Abstract In their comment on a recent contribution of mine, [Watson, S., The meaning of probability in probabilistic safety analysis. Reliab. Engng & System Safety , 45 (1994) 261–269.] Yellman & Murray assert that (1) I argue in favour of a realistic interpretation of probability for PSAs; (2) that the only satisfactory philosophical theory of probability is the relative frequency theory; (3) that I mean the same thing by the words ‘uncertainty’ and ‘probability’; (4) that my argument can easily lead to the belief that the output of PSAs are meaningless. I take issue with all these points, and in this response I set out my arguments.