Imperatives as semantic primitives

There are three different ways in which nondeclaratives have been integrated into formal theories of meaning. These are: the performative analysis, which treats them as explicit or modal performatives and attempts to reduce them to statements; the parametric analysis, distinguishing a sentence type expression M from a ‘sentence radical’ p, together forming a meaningful expression [M,p]; and the operator analysis, according to which (nondeclarative) sentence type expressions scope over an statement, yielding M(p). The difference between the second and third approaches is that the parametric analysis considers all statements to subordinate its semantic content, whereas on the operator analysis simple propositions p are treated on a par with sentences containing a sentence type expression. According to the operator analysis an assertion-operator is unnecessary, because an utterance of a proposition (under appropriate circumstances) in itself constitutes an assertion. The idea that binds these three accounts of imperatives is that the intuitive connection between declaratives and imperatives, as in (1), must be explained by reference to a shared semantic entity contained in an analysis of those sentences.