We can either (a) define probability in a purely mathematical way and so obtain a unique solution which may be both consistent and correct (although, in my opinion, Sir Arthur Eddington's solution ...