Semantics and pragmatics
The study of pragmatics has only arisen fairly recently in linguistics. Investigation of meaning, by contrast, without which the study of gram mar is impossible, has always been considered as a central part of the study of language, even before its constitution as a separate subdiscipline round the time of Bréal (see Chapter 1). The focus in pragmatics on language as it is actually used in context poses a significant challenge to linguistic semantics. If, as an empirical discipline with ‘scientific’ aspirations, linguistics doesn’t set out to study language as it is actually used, then what is it supposed to study? If semantics focuses on the abstracted, idealized, context-free meanings of linguistic expressions, what genuine evidence is there for these meanings other than the ways words are used in actual discourse? And if the very hypothesis of meaning is supposed to explain how words are used, couldn’t the kind of gaps between meaning and use studied in pragmatics suggest that there’s something wrong with the very postulation of abstracted, idealized, context-free meanings in the first place?
Considerations like these lead many pragmatists to see pragmatic questions, not semantic ones, as the central ones in any study of meaning. One of the first to realize the importance of pragmatic questions was the philosopher Rudolph Carnap. For Carnap, pragmatic facts are especially central since they are discovered before any possible semantic ones in empirical investigation of unknown languages:
Suppose we wish to study the semantical and syntactical properties of a certain Eskimo language not previously investigated. Obviously, there is no other way than first to observe the speaking habits of the people who use it. Only after finding by observation the pragmatical fact that those people have the habit of using the word ‘igloo’ when they intend to refer to a house are we in a position to make the semantical statement ‘“igloo” means (designates) house’ and the syntactical statement ‘“igloo” is a predicate.’ In this way all knowledge in the fi eld of descriptive semantics and descriptive syntax is based upon previous knowledge in pragmatics. . . . pragmatics is the basis for all of linguistics. (Carnap 1942: 21; italics original)
As we have observed at several points, pragmatic considerations like reference assignment, scope interpretation and implicatures like the temporal subsequency reading of and may enter into the truth-conditions of an utterance. Even on the strictest truth-conditional approach to meaning, then, the boundary between semantics and pragmatics is porous: acts considered as prototypically part of the domain of pragmatics are necessary to the very calculation of truth-conditional meaning. The interpenetration of semantics and pragmatics also applies in the domain of speech acts. As noted by Strawson (1971: 150), in performative utterances like I apologize, meaning and illocutionary force are co-extensive: the meaning of I apologize is that the speaker is performing the speech act of apologizing. In lexical semantics, also, rejection of the dictionary encyclopaedia distinction amounts to a rejection of any split between semantics and pragmatics. If any piece of encyclopaedic knowledge may become linguistically relevant, there would seem to be little reason to view some as part of dictionary knowledge of meaning, governed by semantic processes, and others as part of encyclopaedic knowledge, governed by pragmatic ones.
This interpenetration between semantics and pragmatics has led some scholars, such as Sperber and Wilson, to reconceive of the nature of linguistic communication in general. On the traditional view, language consists in the communication of a definite content: a certain proposition either is or is not communicated by a given utterance; other propositions, in turn, may or may not be implied by it, but, if they are implied by it, this means that they are not specifically communicated. Grice had already questioned this view: it is probably not possible, according to him, to devise a decisive test to settle the question whether a conversational implicature is present or not – a test, that is to say, to decide whether a given proposition p, which is normally part of the total signification of the utterance of a certain sentence, is on such occasions a conversational (or more generally a nonconventional) implicatum of that utterance or is, rather, an element in the conventional meaning of the sentence in question. (Grice 1989: 42–43)
While we seem to be able to react to implications in the course of normal discourse, it does not seem to be possible for us to formulate any absolute test to distinguish between what an expression means and what it merely implicates: the boundary between semantics and pragmatics, therefore, is entirely fluid.
Sperber and Wilson share Grice’s view. For them, communication is a matter of degree. Each utterance draws a certain set of propositions to the attention of the hearer, weakly activating a vast number of pieces of encyclopaedic knowledge, and the hearer will apply the comprehension procedure in order to determine which is the one the speaker probably intends. It does not make sense, on this picture, to speak of a single proposition which is uniquely conveyed by an utterance: an utterance’s meaning is always the product of a choice from among a set of options, and the fac tors leading to the choice of one interpretation can only be understood against the background of the other interpretations which the hearer considered and rejected. Adopting this picture allows us to avoid the fallacy that communication is always what Sperber and Wilson call strong communication (Sperber and Wilson 1995: 58–60): communication, that is, in which a single, definite content is uniquely conveyed. There is much more vagueness and indeterminacy in real language than this idealized picture suggests, and to suggest otherwise is to endow communication with spurious precision.
