INTRODUCTION 2 страница

(29) Peter is mad

and Mary, on some other occasion, says (or writes)

(30) Tour brother is mad.

Let us further suppose that 'Peter' and 'your brother' refer to the same person and that 'mad' is being used with the same sense (rather than on one occasion with the meaning "insane" and on the other with the meaning "angry"). Provided that they have indeed asserted a proposition, John and Mary will have asserted the same proposition, and will therefore have said the same thing in this other sense of 'say'. But, in saying, respec­tively, Peter is mad and Tour brother is mad, they will not necessarily have made an assertion.

To make an assertion, or statement, is not to perform a locu­tionary act of one kind rather than another; it is to perform a locutionary act whose product - an utterance-token - has one kind of illocutionary force, rather than another. According to Austin, as we have seen, the constative or descriptive function of a language is only one of its functions. We also use languages to ask questions, issue commands and make promises; to threaten, insult and cajole; and, of course, to do all those things for which Austin originally employed the term 'performative' -to baptize a child into the Christian faith, to plight one's troth, to sentence a convicted criminal, and so on. In short, there are many different language-functions and correspondingly many different kinds of illocutionary force.

But how many:' One way oftackling this question is to ask how many verbs in a particular language can be used in explicitly performative utterances in the same way that 'promise', for example, can be used non-constatively in the utterance of (3) of section 8.1 above, repeated here as:

8.3 Illocutionary force 249

(31) 'I promise to pay you Ј5'.

There are hundreds, if not thousands, of such verbs in English. Some of them are more or less synonymous: e.g., 'implore' and 'beseech'. Others, though obviously not synonymous, can be seen intuitively as falling into classes with common characteris­tics. For example, 'promise' and 'undertake' are semantically related in that, when they are used in explicitly performative utterances, their use commits the speaker to a particular course of action. All such verbs, and therefore the particular kinds of illocutionary force which they serve to make explicit, may be grouped together as members of one class. Similarly for other sets of semantically related performative verbs: e.g., 'order', 'command', 'request', etc., all of which have the common prop­erty that, when they are used in explicitly performative utter­ances, their use expresses the speaker's will that some other person, usually the addressee, should carry out a particular course of action. Such utterances (orders, commands, requests, etc.) are commonly referred to collectively nowadays as direc­tives.

Austin himself provided the outlines of one classificatory scheme of this kind at the very end of How to Do Things With Words. Other such schemes, differing to a greater or less extent from Austin's, have since been put forward by his followers. The very fact that alternative more or less plausible classifica­tions are possible constitutes a problem. How do we decide between one classification and another? There is no reason to suppose that the set of performative verbs in English or in any other language will give lexical recognition to every possible kind of illocutionary force. There is still less reason to suppose that there must be some uniquely correct analysis of such verbs, applicable to all cultures and to all languages. Indeed, the vast majority of performative verbs in English and other languages are obviously culture-dependent. For example, the meaning and use of the verb 'swear', in so far as it differs from that of 'promise' and 'undertake', on the one hand, or 'covenant', 'contract' and 'guarantee', on the other, depends upon the culturally established institution of the taking of oaths.

250 Speech acts and illocutionary force

Moreover, it is now clear that it is wrong to attach particular importance to performative verbs. Admittedly, they had a special status in Austin's original formulation of the distinction between constative and performative utterances. But this was because at that time he was mainly concerned to challenge the descriptive fallacy. Looked at from this point of view, sentences such as

(32) 'I promise to pay you ₤5'

were obviously of greater theoretical interest than sentences such as

(33) 'I will pay you ₤5'.

In terms of the later, more general, notion of illocutionary force which he developed in How to Do Things With Words, we have no grounds for confining our attention to declarative sentences con­taining performative verbs.

It is also worth noting at this point that the definition of 'expli­cit performative' which I gave earlier in this chapter ("one in which the utterance-inscription contains an expression which denotes or otherwise makes explicit the kind of act that is being performed") makes no reference to performative verbs as such. For example, the expression 'by Heaven' might be used by mem­bers of a particular group of English speakers as an equally expli­cit alternative to the use of the verb 'swear', in order to indicate that they are taking an oath. In which case, in the appropriate circumstances,

(34) By Heaven, I'll pay you ₤ 5

would count as an explicit performative: 'by Heaven' would make explicit, though it does not denote, the illocutionary force of the utterance. It is but a short step to the recognition of the further possibility that speakers should be able to make explicit the illocutionary force of this utterance, not by using a particular expression, but by using a particular modal particle, a particular grammatical mood or, even, a particular intonation-pattern. I will come back to this.

8.3 Illocutionary force 251

For various reasons, then, there seems to be little point in drawing up comprehensive and allegedly universal schemes for the analysis of illocutionary force based on the existence of a par­ticular set of performative verbs in particular languages. There is perhaps even less point in trying to establish a watertight clas­sification of all possible speech acts in terms of the necessary and sufficient conditions that they must satisfy in order for them to count as instances of one class rather than another. Most speech acts, as I have said, are culture-specific in that they depend upon the legal, religious or ethical conventions and practices institutionalized in particular societies. If the society is one, like our own, with firmly established principles for deciding at law whether something is or is not, let us say, a breach of contract, it may be relatively easy to propose necessary and sufficient con­ditions constitutive of speech acts of this particular kind. But we are deluding ourselves if we think that all speech acts are regu­lated, in this way, in the societies in which they operate. Even the act of promising, which looks as if it might be readily defin­able in terms of the conditions that regulate it, turns out to be problematical from this point of view. It certainly cannot be assumed without argument that promising, in the sense in which we understand the verb 'promise', is an illocutionary act (i.e., a locutionary act with a particular illocutionary force) that can be performed with all languages and in all cultures. And yet assumptions of this kind are commonly made in some of the more specialized work in the theory of speech acts.

Although most speech acts are culture-specific, there are others that are widely, and perhaps correctly, assumed to be uni­versal. They include making statements (or assertions), ask­ing questions and issuing directives. It has been argued, on philosophical grounds, that these three classes of illocutionary acts are not only universal, but basic - in two senses of 'basic': first, that no human society could exist in which acts of this kind have no role to play; second, that many, if not all, culture-specific illocutionary acts can be seen as belonging to a more specialized subclass of one of the three basic classes. For example, as I mentioned earlier, swearing on oath that some­thing is so is obviously a culture-specific act. But swearing that

252 Speech acts and illocutionary force

something is so is also one way oi making a strong statement; and statement-making, it is argued, is basic and universal.

I will not go into the question of the relation between basic and non-basic speech-acts. However, I would emphasize one point: even if the allegedly basic acts of making statements, ask­ing questions and issuing directives are universal, they too are regulated, in all societies, by more or less culture-specific institu­tions, practices and beliefs. One recognizable dimension of cul­tural variation, in this respect, is that of politeness. It is impolite, in all societies, to speak out of turn: that is, to speak when the social role that one is playing does not grant authority and precedence or, alternatively, when the rules that govern turn-taking in that society do not grant one the authority to speak at that point. It is also impolite, in some societies, to be too assertive in the exercise of one's locutionary and illocution­ary authority. For example, it might be considered impolite, in certain circumstances, to make a straightforward unqualified assertion or to issue a blunt and unqualified command. The ori­gin and more or less conventionalized used of various kinds of indirect speech-acts can be explained in such terms as these, as, for example, in English, where Would you pass the sugar? (origi­nating as a question and commonly so punctuated in its written form) is used in preference to Pass the sugar (a direct command).

Politeness, however, is but one of the dimensions of cultural variation that regulate the use of the allegedly basic speech-acts. Furthermore, though it has a certain cross-cultural validity and under a sufficiently general interpretation of 'politeness' may be universal, it does not manifest itself in the same way in all societies. One must be careful, therefore, not to assume that generalizations made on the basis of one's experience of one kind of society will be valid in respect of all human societies. This point should be borne in mind in all that follows. For the discussion and exemplification of the part played by politeness and other factors in the regulation of language-behaviour in dif­ferent cultures, reference may be made to recent work in socio-linguistics and pragmatics.

8.4 Statements, questions and directives 253

8.4 STATEMENTS, QUESTIONS AND DIRECTIVES

We are assuming that all languages provide their users with the means of making statements, asking questions and issuing direc­tives: i.e., of producing utterances with these kinds of illocution­ary force. It does not follow from this assumption, however, that all languages will grammaticalize these differences of illo­cutionary force. As we saw in Chapter 6, it is quite possible for sentences to exist which are neutral in sentence-type or mood: sentences (or clauses) which are neither declarative nor interro­gative, on the one hand, and are not indicative, subjunctive or imperative, on the other.

Nevertheless, it may well be clear enough what illocutionary act is being performed when one of these sentences is uttered. This may be clear not only from the context in which it occurs, but also from the prosodic structure that is superimposed upon the resultant utterance-inscription. For example, if English had no interrogative sentence, so that

(35) 'The door is open'

was not declarative, but neutral in sentence-type, it would be possible to utter this sentence (as The door is open) with, let us say, a falling intonation-contour to make a statement and rising intonation-contour to ask a question. This point was made earlier. It may now be generalized in terms of the more detailed account of the process of uttering sentences which is being outlined in this chapter.

But it was also asserted earlier that many languages, including English, do in fact grammaticalize distinctions of sentence-type and mood: and furthermore that there is an essential connexion between sentence-type and mood, on the one hand, and what we are now calling illocutionary force, on the other. What is the nature of this connexion? And how do statements, questions and directives differ from one another semantically? I will do no more than provide a partial answer from a particular point of view.

To make a statement is to express a proposition and simulta­neously to express a particular attitude towards it. I will call

254 Speech acts and illocutionary force

this attitude, for reasons which will be clearer when we look at the notion of modality, epistemic commitment. (The term 'epistemic', which comes from a Greek word meaning "knowl­edge", is used by logicians to refer to that branch of modal logic that deals with knowledge and related matters.) Anyone who states a certain proposition is committed to it, not in the sense that they must in fact know or believe it to be true, but in the sense that their subsequent statements - and anything that can be legitimately inferred from their accompanying and subse­quent behaviour - must be consistent with the belief that it is true. Hence the unacceptability or paradoxical character of

(36) It is raining but I don't believe it

(construed as a statement). In making any such statement the speaker is guilty of a breach of epistemic commitment.

When one asks a neutral (i.e., epistemically unbiased) ques­
tion, one expresses a proposition and simultaneously expresses
one's attitude of non-commitment with respect to its truth-
value. But there is more to it than this. As was noted in Chapter
6, Is the door open? '— that is a question which I refuse to ask is a perfectly
acceptable utterance. In this case a question is posed, but not
asked. To ask a question then is not merely to express the prop-
ositional attitude of non-commitment - that is, to pose the prop­
osition as a question - but also, in so doing, to indicate to one's
addressee — prosodically, paralinguistically or otherwise — that
one desires them to resolve one's uncertainty by assigning a
truth-value to the proposition in question. It follows, for this
and other reasons, that questions are not, of their nature, a sub­
class of directives (as several authors have suggested).

What then of commands, requests and other kinds of direc­tives? These differ from statements and ordinary (i.e., neutral) questions in that they involve a different kind of commitment on the part of the speaker: deontic commitment. (The term 'deontic' comes from a Greek work relating to the imposition of obligations. Like 'epistemic', it is borrowed from modal logic.), In issuing a directive speakers commit themselves not to the; truth, or factuality, of some proposition, but to the necessity' of some course of action. To make the same point in more

 

8.4 Statements, questions and directives 255

traditional terms: they express, not their belief that something is so, but their will that something be so.

In making a request (rather than issuing a command or an order), speakers express their will that something should be so, but they also explicitly concede to the addressee the right of non-compliance. Requests are in this respect like non-neutral, so-called leading or conducive, questions - questions such as

(37) The door is open, isn't it?,

in the utterance of which speakers express their own tentative or provisional commitment to the truth-value of the proposition "The door is open" but simultaneously concede to the addressee the right to reject it (see 6.7). Another way of making the point is to say that in conducive questions and requests the speakers express their commitment to the "it-is-so" or "so-be-it" compo­nent of the utterance and invite the addressee to do the same.

The analysis of statements, questions, commands and requests that has been presented in outline here suggests that their illocu­tionary force can be factorized, in each case, into two compo­nents: a component of commitment ("I say so") or non-commitment, on the one hand, and what might be referred to as a modal component of factuality ("it is so") versus desirabil­ity ("so be it"), on the other. I have used the term 'modal' in this connexion (instead of introducing some more specialized terminology) for two reasons. First, the distinction between fac­tuality and various kinds of non-factuality falls within the scope of what logicians have traditionally referred to as modality: I have prepared the way for the treatment of modality that will be given later by deliberately introducing the terms 'epistemic' and 'deontic' (10.5). Second, such distinctions are commonly grammaticalized in languages in the category of mood. It is important to realize, however, that mood in natural languages may also grammaticalize different kinds or degrees of commit­ment.

The grammatical category of mood must not be confused with what some logicians refer to as the mood of a proposition, which rests upon the objectification of the essentially subjective component of commitment. This is only part of what is covered

256 Speech acts and iilocutionary force

by the grammatical category of mood, which, as we shall see in Chapter 10, always expresses subjectivity. If a language has a grammatical mood which is used distinctively and characteristi­cally for the purpose of expressing the speaker's unqualified epis-temic commitment, that mood is traditionally described as the indicative. Similarly, if a language has a grammatical mood which is used distinctively and characteristically for the purpose of imposing one's will on others for the purpose of issuing direc­tives, that mood is traditionally described as the imperative.

As we shall also see in Chapter 10, there are various ways in which the speakers can qualify their epistemic or deontic com­mitment. All natural languages provide their users with the pro-sodic and paralinguistic resources which enable them to do this in speech. Some, but by no means all, natural languages gram-maticalize different kinds and different degrees of commitment in the category of mood; and some languages lexicalize or semi-lexicalize them by means of modal adverbs and particles.

All this will be taken up later in connexion with the notion of subjectivity. I have mentioned it here, without detailed develop­ment or exemplification, in order to show how a fairly tradi­tional view of mood can be reformulated within the framework of the theory of speech acts developed by Austin and his fol­lowers. As we have seen in this chapter, Austin began by identi­fying explicit performatives as a rather special class of utterances, in the production of which the speaker is doing some­thing, rather than saying something, by means of language. He later came to realize that all saying is doing and that all kinds of saying — including the production of statements, questions and directives - are regulated by the central concepts of authority and commitment.

Austin himself recognized the social basis of these concepts, even though he did not go into this aspect of the matter in detail; and at this point he makes contact, though not explicitly, on the one hand with the later Wittgenstein and, on the other, as we: shall see in the next chapter, with Grice. He might just as well have emphasized the personal, or expressive, character of the concepts of authority and commitment. This is what is done in the traditional grammarian's accounts of mood, couched in

8.4 Statements, questions and directives 257

terms of the speaker's judgement and will. Here, as elsewhere, not only in the use of language, but in all communicative beha­viour, the expressive merges with the social and is ultimately indistinguishable from it.

Indeed, as some philosophers, anthropologists and social psy­chologists have argued, there is an important distinction to be drawn, in this connexion, between persons and individuals, in that it is persons rather than individuals that one is, or should be, concerned with in the discussion of communication. It has also been argued that the person (or the self) is a social prod­uct - the product of socialization - and that socialization is a process in which the acquisition and use of a particular language in a particular culture plays a vital part. This point also will be picked up in connexion with the notion of subjectivity (10.6).

One final point must now be made. The theory of speech acts is sometimes advocated, or criticized, as if it were an alternative to truth-conditional semantics. It should be clear from earlier chapters of this book that the two theories are, in principle, com­plementary. Truth-conditional semantics, as it is currently applied to natural languages, is a theory of the prepositional content of sentences; speech-act theory - if we grant that it is or aspires to be a theory - deals with the iilocutionary force of utterances. There has been much discussion by linguists and phi­losophers of the question whether Austin was right or wrong when he said that utterances such as I promise to pay you Ј5, when used to make a promise, are neither true nor false, but either effi­cacious (or felicitous) or not. I have said nothing about this con­troversy here, because in my view it is of no consequence whether we resolve the issue one way or the other; and how we resolve it depends not so much on the facts of the matter as on the theoretical and philosophical framework within which we are operating. The most important point, for our purposes, is that the iilocutionary force of ordinary descriptive statements, such as It is raining, cannot be accounted for satisfactorily within the framework of truth-conditional semantics. Austin is at one with Frege in making this point. I judge it to be incontrovertible.

CHAP T E R 9

Text and discourse; context and co-text

9.0 INTRODUCTION

We have been operating with the assumption that utterance-meaning is crucially dependent on context. So far, however, I have made no attempt to say what context is or how it deter­mines the meaning of utterances and controls our understanding of them. Nor have I said anything in detail about spoken and written text: I have, however, made it clear in previous chapters that speech must be distinguished from writing (and the prod­ucts of speech from the products of writing), even though, in the technical metalanguage of semantics that we have been building up throughout the book, 'utterance' and^;'text' are being applied to the products of both speech and writing.

In this chapter, we shall be dealing with both text (and dis­course) and context (and co-text). As we shall see, text and con­text are complementary: each presupposes the other. Texts are constituents of the contexts in which they are produced; and con­texts are created, and continually transformed and refashioned, by the texts that speakers and writers produce in particular situations. It is clear that even sentence-sized utterances, of the kind we considered in the preceding chapter, are interpreted on the basis of a good deal of contextual information, most of which is implicit.

We shall begin by recognizing explicitly that the term 'sen­tence' is commonly used by linguists (and also by non-linguists) in two senses, one of which is, to put it loosely, more abstract than the other. It is the more abstract sense of 'sentence' that is relevant when linguists talk about a grammar as generating the

9.1 Text-sentences 259

sentences of a language and when semanticists draw a distinc­tion, as I have been doing, between sentence-meaning and utter­ance-meaning. Relatively few linguists use the technical terms 'system-sentence' and 'text-sentence' that I introduce below. Most of those who draw a distinction between sentence-meaning and utterance-meaning do, however, acknowledge the impor­tance of distinguishing between the more abstract and the more concrete sense of 'sentence'. They would also recognize that the relation between these two senses has to be made explicit on the basis of a satisfactory theoretical account of the role of context in the production and interpretation of utterances. As we shall see, linguists who are engaged in the construction of such a theo­retical account (whether they call themselves semanticists or pragmaticists) have drawn heavily on Paul Grice's notion of implicatures.

9.1 TEXT-SENTENCES

Before we can talk sensibly about the relation between text and context, we must look again at the status of sentences.

It was pointed out, in the previous chapter, that many of our everyday utterances are grammatically incomplete or elliptical. Some of them are ready-made locutions of fixed form: Good heavens!, Least said, soonest mended, etc. I shall have no more to say about these. I mention them merely to indicate that, in all lan­guages, there are such expressions, finite in number and in some cases of more or less determinate grammatical structure, whose form and meaning cannot be accounted for, synchronically, in terms of the utterance of sentences. They must of course be accounted for in the description of the grammatical and seman­tic structure of particular languages. But they do not raise problems different in kind from those which arise in the analysis of the infinite set of potential utterance-inscriptions, any one of which can, in principle, result from the utterance of sentences. Only a finite, and relatively small, subset of this infinite set of potential utterances is ever actualized in the day-to-day use of a language. But, as generative grammarians have rightly insisted in recent years, linguistic theory cannot be restricted to the

 

 

260 Text and discourse; context and co-text

analysis of a finite set of actual utterance-inscriptions, however large and representative of (the products of the use of) a language that set might be. In insisting upon this point (which is generally accepted by formal semanticists whether or not they subscribe to the principles of generative grammar as these are currently formulated within linguistics), generative grammarians were merely reasserting something which had been taken for granted, over the centuries, by theorists and practitioners of traditional grammar. What is new and exciting in generative grammar (as also in formal semantics) is the attempt to give full effect to the principle of compositionality in accounting for the grammatical structure (and meaning) of the sentences of natural languages. However, there has been a good deal of confusion, both in generative grammar (and formal semantics) and in traditional grammar about the relation between sentences and utterances. Our first task in this section is to clear up this confusion. It is the infinite set of potential utterance-inscriptions with which we are mainly concerned here.

Paradoxical though it may appear, the product of the utter­ance of a sentence is not necessarily a sentence. The apparent paradox disappears immediately if we draw a distinction between a more abstract and a more concrete sense of 'sentence¹. Sentences in the more abstract sense are theoretical constructs, which are postulated by linguists, in order to account for the acknowledged grammaticality of certain potential utterances and the ungrammatically of others. They may or may not have some kind of psychological validity in the production and interpretation of utterances (i.e., utterance-inscriptions), but they certainly do not occur as the inscribed, and transcribable, products of utterances. I will refer to sentences in this more abstract sense of the term as system-sentences; they are what are generated by the grammatical rules in a generative gram­marian's model of some language-system (operating upon a vocabulary, or lexicon, which is part of the same language-system). But the term 'sentence' is also used both traditionally and in modern linguistics (as also in everyday non-technical discourse) in a more concrete sense.

9.1 Text-sentences 261

So, let me now introduce the term text-sentence for this more concrete sense of 'sentence' - the sense in which sentences are a subclass of utterance-inscriptions and, as such, may occur (in some languages at least) as whole texts or as segments of text. This will allow us to say that the utterance of a particular system-sentence, such as

(1) ' I have not seen Mary',

will result, in some contexts, in the production of a text-sentence, such as

(2) I have not seen Mary

(with or without the contraction of have not to haven't, and with some contextually appropriate prosodic structure). This may look like the multiplication of theoretical entities beyond neces­sity. But there is a very considerable pay-off.