Donald Hobson comments on Strong implication of preference uncertainty