Jorn Sonderholm (2005) has argued that Simon Blackburn's commitment semantics for evaluative discourse is unable to explain the validity of simple inferences involving disjunction. This is true insofar as the basic rules which Blackburn suggests are not strong enough, but it is relatively simple to augment those rules so as to meet Sonderholm's challenge, whilst respecting the spirit of commitment semantics. One way of doing this is to add a reduction rule such that if accepting p commits one to inconsistent commitments, one is committed to accepting ¬p. Thus Sonderholm has not provided any reason to doubt the adequacy of commitment semantics to explain validity in evaluative discourse.