The semantics and pragmatics of imperative sentences

One of the basic functions of language is to get people to do things. An understanding of how we manage to provoke action via words is fundamental to an understanding of human communication. We need to understand how it is that we can use words not merely to describe the world, but to change it. The centrality of this function is shown by the fact that, throughout the world's languages, we repeatedly find linguistic forms whose role seems to be to indicate that the speaker's goal is that the hearer to act in a certain way.
'Imperative' is the name we give to the types of sentences we have in our languages that appear to be designed to signal our intention to get someone to do something. In English, we recognise sentences of this type by the absence of a subject, while, in many other languages, a special form of the verb is used. So, the Spanish version of (1) is (2), which is marked by a verb form that can only be used in this type of sentence.
(1) Go away!
(2) Vete!
Although basic and fundamental, imperative sentences are poorly understood. They pose a number of problems for linguists tackling problems of meaning and interpretation. Firstly, just specifying their meaning is far from problematic. The most obvious hypothesis is that their meaning just is that a command (or request) is being made. However, there are considerations that argue against this position. For example, we find sentences such as (3), in which the first clause is an imperative, but it can't be interpreted as a command: you can't command someone to be tall.
(3) Be tall and you'll get picked for the team
Second, there is the problem of how to characterise the meaning of imperatives within semantic theory. Linguistic semantic theories generally try to explain meaning in terms of truth: the meaning of a statement is typically expressed in terms of the conditions that would have to hold in the world for that statement to be true. Imperative sentences, however, are not open to judgements of truth and falsity. This raises the problem of how to explain their interpretation.
Over the last 50 years, a number of authors in linguistics and the philosophy of language have attempted to provide an account of how the meaning of imperative sentences should be characterised. While a great many important insights have been achieved, there has yet to emerge a consensus on this issue. Nor, moreover, has there been any attempt to systematically bring together this range of perspectives in one volume.
For these reasons, the time is ripe both for a book that brings together and critically evaluates the range of proposals made in the last 50 years and for a new theory that builds on their insights and overcomes their shortcomings.
The present proposal seeks to address both of these needs through a book and a journal article.
The book, a contract for which has been signed with CUP, will first detail the semantic and pragmatic challenges presented by imperative sentences, before focussing in detail on the issues raised by non-commanding uses. Previous theories will then be discussed in the light of their ability to deal with the data discussed in the first part of the book. The book will take a critical approach throughout, and will be marked by its consideration of the typological data on imperatives, thus avoiding the focus on English that is a feature of so much work in this area.
The book and paper will be co-authored with Dr Mikhail Kissine of Université Libre de Bruxelles. We will be working independently on separate chapters before revising each other's work, as the work plan shows. Funding is thus sought for three meetings to assess progress towards the final book.

An understanding of the use of language to achieve goals has potential impacts in many areas. This ranges from a better understanding of interpersonal relationships to the study of power deployment by and within institutions. Consequently, the work described in this proposal has potential impacts for anyone working in areas (academic or otherwise) where the aim is to get others to act.
Moreover, the cognitive approach adopted in our research means that it has a broader range of potential impacts than much linguistic theory. For example, one aspect of my co-author Dr Kissine's work concerns the comprehension of indirect speech-acts (commands etc.) by children on the autistic spectrum. The processing model that we propose will be informed by this research and is therefore likely to have practical implications for those working with autistic children. Furthermore, a cognitive approach, especially when complemented by formal rigour, lends itself towards applications in artificial intelligence. Indeed, my own earlier work on the Spanish subjunctive has been cited in the machine-translation literature. The approach taken to the imperative in the proposed paper will be directly transferable to practical applications in machine intelligence.