Interpersonal context: speakers intention and hearers inference
As we have seen, Austin did not believe that illocutionary acts were accompanied by any predictable grammatical or lexical markers. Both he and Searle did, however, believe that the making and understanding of speech acts is governed by social conventions like the ones for statements and requests discussed above. However, the idea that conventions underlie the illocutionary force of utterances has been much criticized. The central problem with such a theory is that it proves to be exceedingly difficult to state what the convention behind any given speech act might be. Thus, the conventions governing statements mentioned above seem inadequate: we often state things for which we do not have evidence (e.g. You’re not going to go bald), which it is obvious that the hearer knows already (You have lost a bit of hair, though), and which we do not believe anyway (But it’s nothing to worry about). Similarly, the putative conventions governing the making of requests may also be violated, without detracting from the nature of the utterance as a request. For example, imagine that S feels obliged to invite H to dinner, but does not want her to come. S may thus invite H to come at a time at which they know H is unavailable. In a case like this, the request Do come and have dinner with us tomorrow is made with S not wanting H to come (condition (iv) in (1) above), and knowing that H is unable to do so (condition (ii)). The utterance is none the less, however, a request (see Strawson 1971: 153–154 for further examples).
The general problem with convention-based approaches to illocutionary force is that they ignore the role of the appreciation of speakers’ intentions in our understanding of meaning. The importance of intention in meaning was first emphasized by the British philosopher H.P. Grice, a col laborator of Austin’s in the 1940s and 1950s. For Grice, ‘the meaning (in general) of a sign needs to be explained in terms of what the users of the sign do (or should) mean by it on particular occasions’ (1989: 217). If I understand that a certain utterance is a statement, a request, or a warning, on Grice’s theory, it is because I attribute to the speaker a certain type of intention: the intention to state, to request, or to warn. It is because I attribute these intentions to the speaker that I am able to interpret the utterance in the right way; if I had credited the speaker with a different intention, I would have taken the utterance differently.
The importance of speaker’s intention applies to both the illocutionary and the locutionary aspects of utterances. On the illocutionary side, the hearer’s interpretation of the speech act performed by the speaker will depend, as we have just seen, on their interpretation of S’s intentions. S’s utterance of the words It’s easy to fall over in the dark may function as a request for H to turn on the lights, a warning to H to be careful, or a metaphorical observation about the dangers of ignorance, uttered without the intention of provoking any particular immediate action on the part of H. In reacting to the utterance, H has to infer which of these possibilities was the one S intended. This is not to say, of course, that S intended only a single one of them: it is quite possible that S had several intentions in uttering those words. Perhaps, indeed, S didn’t even know what their intention was; they just uttered the words. Nevertheless the hearer is obliged to make inferences about S’s overall intentions in order to respond appropriately.
On the locutionary side, it is by making inferences about the speaker’s intentions that the hearer selects the relevant aspects of the encyclopaedic knowledge called up by a linguistic expression: the encyclopaedic information relevant to the interpretation of an utterance is the information which the speaker intended to convey, and the hearer must decide which of the potentially infinite elements of encyclopaedic knowledge the speaker had in mind. Thus, if I use the word frog in reference to a French person in the phrase He may be a Frog, but no princess is kissing him (see (19) in the previous chapter), it is because I am considering certain facts and not others as relevant in this context: the fact that there is a fairy story in which a princess kisses a frog, and the fact that French people may be referred to as frogs. In order to understand (19) correctly, any hearer will have to appreciate my intent to convey this information. But the role of intentions is not limited to the selection of the appropriate encyclopaedic facts about a word. We also need to understand the speaker’s intention in order to disambiguate words and assign referents, both basic aspects of the determination of the locutionary act of what is actually said. If I hear the sentence There was a mouse here this morning, my choice between the interpretations there was a small rodent in the house this morning and there was a computer accessory on this table this morning will be made on the basis of my beliefs about the speaker’s intentions: did the speaker intend me to understand her to be making a comment about the presence of wildlife somewhere in the house or about a computer part that should have been on the table?
So inferring the speaker’s intention is, on this view, a fundamental aspect of the process of meaning-creation and understanding in language. Linguistic communication is an intentional-inferential process, in which hearers try to infer speakers’ intentions on the basis of the ‘clues’ provided by language. It is, as described by Sperber and Wilson (2002: 3), ‘essentially an exercise in metapsychology, in which the hearer infers the speaker’s intended meaning from evidence she has provided for this purpose’. The viability of an analysis of meaning in terms of intentions has not infrequently been called into question by philosophers of language, and it does indeed seem, for reasons that there is not space to go into here, as though the details of this analysis are rather problematic (see e.g. Schiffer 1987: 242–249). Nevertheless, Grice’s programme of intentional inferential semantics is assumed by many linguists and has proven to be a fruitful way of understanding language use.
Grice called the type of intention-dependent meaning characteristic of human language non-natural meaning (meaningNN). The label ‘non-natural’ is intended to contrast with natural types of meaningfulness which are not mediated by a speaker’s intentions, such as when we say those spots mean measles: here, the link between the spots and their ‘meaning’ (measles) is causal, direct and independent of any human agency, whereas the meaning of an utterance in human language depends on the intention of the utterer. In general, for Grice, the notion of what a word means is only explicable in terms of what speakers mean by using the word. What is important in communicating is thus what speakers intend by their use of language, what speakers use words to mean, and it is only derivatively, in light of these intentions, that we may speak of words themselves meaning anything (Grice 1989: 214–221).
QUESTION Consider involuntary exclamations of pain like ouch or ow. Are these instances of meaningNN? If so, why? If not, why not?
