Introduction

Imperatives constitute one of the major clause types of human language; most if not all languages have a grammatically distinctive class of sentences which are associated closely with directivemeaning. For example, in English we have (1):

(1) Ben, feed the bird!

The sentence is grammatically distinctive in that it lacks an overt subject (Ben here is a vocative) and employs a verb which is nothing but the verb stem (we can say it is the bare infinitive form), both properties which root clauses cannot usually have in standard English. Moreover, the sentence obviously has directive meaning in that it is naturally used with the aim of getting the addressee, Ben, to feed the bird and can barely be used with the aims associated with other clause types (Truckenbrodt, 2006): one cannot use it to assert that someone will feed the bird, has fed the bird, or anything of the sort; one should use a declarative like (2) if that is one's communicative goal. Nor can one use it to ask whether the bird will be or was fed; interrogatives such as (3) stand ready for English speakers with that purpose.

(2) You will feed the bird.

(3) Will you feed the bird?

Although (1) cannot be used to assert or ask for information in the ways that (2)–(3) can, it is possible for (2) to be usedmore or less as (1) is. The declarative sentence here can easily be used with the aim of getting the addressee to feed the bird, and it is even possible for (3) to be used directively as well (in the situation where the addressee is supposed to feed the bird, yet he has apparently forgotten to).

While the description of imperatives just outlined is nothing surprising, it is worthwhile bringing it up explicitly because the most basic issues for the semantic analysis of imperatives concern how to understand better and make precise the intuitive ideas which underlie it.