Consequences of the illocutionary perspective
Focusing on the illocutionary aspects of utterances has two important consequences for linguistic theory. The first concerns the centrality of truth and falsity to meaning. We saw in 3.2.1 that Frege had made truth the central notion of his semantic theory. However, considerations of truth and falsity are simply irrelevant for many types of illocutionary act. This is particularly so for performatives, such as the sentences in (2):
As Austin points out, it does not make sense to ask whether it is true that ‘I apologize for the mess I’ve made’: the very act of saying the words I apologize constitutes the apology. Instead of being assessed as true or false, the sentences in (2) must conform to certain conditions, just like the conditions governing the act of requesting described above. Austin called these conditions felicity conditions.
QUESTION What might the felicity conditions be for each of the speech acts mentioned above? What problems are there in deciding?
The second break with traditional theories of language brought about by a focus on speech acts concerns the question of the basic object of semantic analysis. Austin was struck by the fact that it seems impossible to specify any list of criteria which might distinguish expressions which can function as performatives from those which cannot. Any expression, for Austin, can carry any illocutionary force. There is no single way, in any language, of performing a given illocutionary act: the illocutionary force of an utterance is in principle unpredictable from its overt syntactic or lexical form. Speakers often perform speech acts whose communicative purpose (utterance meaning) does not correspond to their obvious sentence meaning. For instance, to get someone to close the door, I may well not choose an imperatival construction (close the door, please), but may opt instead for either a question (could I get you to close the door?) or a statement (it’s suddenly got draughty in here). Speech acts like this, whose illocutionary force does not correspond to the most obvious illocutionary force of their sentence type, are known as indirect speech acts.
Because of the frequency of indirect speech acts, any proposed convention linking a given communicative purpose with a given illocutionary form will thus have to reckon with the fact that the same form may also be used to achieve quite different purposes. French, for example, uses a variety of linguistic mechanisms to express commands. As well as the expected imperative form (tais-toi ‘be quiet’), these include verbs in the infinitive mood (ne pas ouvrir ‘do not open’), and in the indicative (vous dresserez une liste des consonnes sourdes, ‘write a list of voiceless consonants’) – even, in some conventional cases, noun phrases (ta gueule ‘shut up!’, literally ‘your mouth’). Furthermore, even explicit performative expressions may be used in ways which do not correspond to their conventional meanings. The phrase ‘I guarantee there are no slackers in this company’, for example, has the apparent illocutionary force of a guarantee, but could function as a warning, a threat, a promise, and so on. As Davidson (1979: 73) says, the fact that a single linguistic structure may serve an unlimited number of contextual communicative ends points up a fundamental feature of human language that he calls the autonomy of linguistic meaning:
Once a feature of language has been given conventional expression, it can be used to serve many extra-linguistic ends; symbolic representation necessarily breaks any close tie with extra-linguistic purpose. Applied to the present case, this means that there cannot be a form of speech which, solely by dint of its conventional meaning, can only be used for a given purpose, such as making an assertion or asking a question.
The impossibility of discerning fixed characteristics of illocutionary force bearing expressions imposes a major reorientation on one’s perspective on meaning. If, given the right context, any expression can be used to create any contextual effect, then
[t]he unit of linguistic communication is not, as has generally been supposed, the symbol, word or sentence, or even the token of the symbol, word or sentence, but rather the production or issuance of the symbol or word or sentence in the performance of the speech act. (Searle 1969: 16)
Under this perspective, the meaningful nature of human communication cannot simply be attributed to the semantic properties of words: meaning fulness is not a property of language on its own, but of language use as part of an interpersonal context, as part of a network of shared social practices. This insight extends even to the assertive, referential use of language. As noted by Recanati (1987: 128), ‘It is not the sentence “It will rain,” but rather the fact of its being uttered by Jules, that “expresses” or conveys pragmatically Jules’ belief that it will rain.’ As Austin himself pointed out, the use of language to describe aspects of reality needs itself to be seen as just one other speech act among many – it is, in other words, an activity subject to a set of conventions entirely equivalent to the conventions governing obvious speech acts like promising, thanking or congratulating.
The conventions governing constative utterances (statements) – let us say the statement that p – include the following (Searle 1969: 66; cf. Austin 1962: 136–147):
QUESTION Do all statements conform to these conventions? If not, can you reformulate them to remove the problems?
The utterance of a statement, just like the performance of an illocutionary act like promising, thus involves both speaker and hearer in a network of commitments and consequences which imply certain things about their current states and beliefs, and commit them to certain future actions. If, for example, I utter the words Rat soup and leather belts were eaten on the Long March, then, in many types of situation, I will be understood to have made a factual statement which I have some evidence or justification for making, and to which I can be expected to be held. The uttering of this statement commits me, as a matter of convention, to certain other propositions, the statement’s entailments and presuppositions (see Chapter 6), such as the proposition that the Long March took place, that leather belts were available, and that there were rats in China at the time. If I subsequently denied any one of these propositions, or the original statement itself, I could incur sanctions from the other members of the exchange: I could be accused, for example, of inconsistency. These sorts of conditions and commitments can be seen as essentially of the same order as those involved in other types of utterance, such as promises. Just as the act of making a statement suggests I have evidence or justification for what I say, the act of making a promise suggests I have the intention to fulfil what I say I will do. And just as the statement commits me to assent to its entailments and presuppositions, a promise commits me to follow through on the promised act. In both cases, then, the utterance of a linguistic expression can be seen as occupying a place in a structure of past and future actions and mental states, maintained and enforced by social convention.
The use of language is thus not the disembodied exercise of human reason asserting neutral facts about the world. It is a situated, contextual act in a network of social roles and responsibilities. This is always the case, regardless of whether the utterance in question seems essentially constative (factual) or performative. As Austin says (1962: 145–147), the difference between constatives and performatives can be seen principally as one of emphasis. Thus, constative utterances (statements, descriptions) abstract away from the illocutionary aspects of the utterance act in order to focus on its locutionary aspects, and the extent to which the utterance corresponds with facts. Performative utterances, contrastingly, focus on the illocutionary aspects, abstracting away from the locutionary dimension.
