The imperative and pragmatics

This article is a tentative exploration of the following question. In an account of the imperative construction in English, what should be accounted for in the syntax or semantics and what in the pragmatics? 1 In the literature, we find descriptions of an imperative construction with certain specific syntactic properties. For example, there is a subjectless form and also a form with a second person pronominal subject and corresponding to both there are second person pro- nominal forms in reflexive and tag counterparts. Abstract underlying structures, in this case a you subject, which subsequently may be deleted by IMPERATIVE or EQUI, are postulated to explain these properties (Postal, 1964; Katz & Postal, 1964: 75; McCawley, 1968). My question is: if we postulate a general pragmatic theory, that is, a theory of the use of utterances in context (separate from but related to theories about the syntactic or semantic properties of sentences), how many of the properties of the imperative can be explained in such terms instead of in the syntax or semantics? My conclusion is that the proposed abstract structures are syntactically and semantically unmotivated and unnecessary for pragmatic interpretation. Each property that the abstract elements explain is better explained either as a non-arbitrary property of main clause infinitives when they are used to utter commands (non-arbitrary in that the facts could not be otherwise), or inherent properties of main clause infinitives in all their uses. In the former case at least, they are facts about men in situations, not about syntax. Thus, syntax and semantics require only a single level of representation and there is no imperative transformation.