L. argues that there are activity-specific constraints on language usage. Both the nature of speech acts and the application of Grice's maxims are dependent on the nature of the activity in which talk is conducted. This is illustrated in detail by an examination of the functions of questions in discourse.