S. questions Austin's belief that the force of speech acts is essentially conventional. He shows that there is a class of speech acts that cannot be said to be backed by conventions (other than purely linguistic ones). He claims that, instead of conventions, intentions (a Gricean notion which he adopts and adapts) typically underlie the performance of speech acts: in some cases the effectiveness of those intentions is guaranteed by conventions, but in other cases the effectiveness is not guaranteed at all.