Page 484 - Encyclopedia of Philosophy of Language

Pragmatics and Speech Act Theory
a constituent of Lb, then ek will be another; (d) if there is no counterevidence, then S is speaking Lb (this procedure is not experimentally tested). Having established that U is made in L, H must use his/her knowledge of L when categorizing what s/he perceives in U as particular lexiconic, syntactic, and prosodic constituents of L. For instance, H needs a knowledge of L's lexicon and how to use it, in order to match the appropriate set of lexicon entries with the lexicon items which s/he perceives S to use in U. H must use knowledge of the syntactic properties of lexicon items in establishing their scope relations and the syntactic structures which combine them in U. Once these are recognized, H can determine the meanings of e's con- stituents, and the meaning ofe itself; H will then take into account the effect of the prosody n on the meaning of e in order to determine the sense or senses of the locution <TT,e>.
The next step for H is to recognize what S is refer- ring to in the world spoken of. To do this, H will match the sense(s) of e's constituents with entities in the world spoken of, using knowledge of thecor- relation between sense and denotation/reference. Having determined the reference of U, H uses knowl- edge of the definitions of illocutionary acts to deter- mine S's illocutionary intention in U. Once again, it is a matter for H's categorizing ability: this time, H perceives what S is denoting by means of the locution <TT, e>, and s/he must determine what kind of message S intends to communicate using this proposition. The set of illocutionary intentions is located among the definitions of illocutionary acts which constitute a part of H's linguistic knowledge (and are presumably located along with other information about mean- ingful properties and meaning relations). Recognition of the illocutionary intention will lead to the identi- fication of the illocutionary act which it helps to define. There are only five or six primary illocutionary acts, each being determined by the form of the locution. To determine the illocutionary point of U, H invokes the reasonableness condition and seeks some reason for S's primary illocution in the context, C; the
conversational maxim of relation (or relevance) will often be invoked too, and so may the other maxims of the cooperative principle—though these are more likely to be called upon at later stages. H will need to keep tabs on the meaningful properties of U and its constituents at this as in all subsequent steps leading to the decision on the illocutionary point of U; s/he must also constantly monitor the semantic relation- ships of U and its constituents to their textual environ- ment; the semantic properties and relations of U may well be significant to its proper interpretation. In seek- ing a reason for the primary illocution, H looks to the context C, and will also check background knowledge of many kinds, including knowledge of the kinds of things that people might say in C, and the kinds of reasons that other speakers might have had when
employing a similar primary illocution; in other words, H uses knowledge of L and the use of L to infer from the primary locution and the circumstances of utterance what the illocutionary point might be. This may lead through a number of indirect illo- cutions. The illocutionary point is only recognized when H can infer no further illocutions in U. When H recognizes the illocutionary point, s/he has finally determined the meaning of U.
Although this description offers a rational model for H's understanding of U, it is misleading because it pretends that each step is completed before the next is begun. This cannot be true because, in reality, peo- ple interpret parts of utterances as they are presented. The inferential theory of speech acts needs to be reworked to allow for this.
12. Performative Analysis
Katz and Postal (1964) recognized the contribution that clause-type makes to meaning, and postulated dummy morphemes 'Q' for interrogatives, interpreted 'I request that you answer'; T for imperatives, inter- preted 'I request'; and no dummy for declaratives. Searle (1969:64) commented '[if] we can reduce all illocutionary acts to some very small number of illo- cutionary types it would then seem likely that the deep structure of a sentence would have a simple rep- resentation of its illocutionary type.' R. Lakoff (1968) proposed that one of the pair of abstract performative verbs which she dubbed *imper 'command' and *hort 'exhort' underlie all imperative sentences in Latin. Lakoff was motivated by the principle of economy: constraints stated on these two abstract verbs apply to all the surface verbs of commanding and exhorting, and do not need to be repeated for each verb separ- ately. Her proposal legitimized abstract verbs as theoretical constructs. Here were the seeds of the per- formative analysis theory of speech acts.
In its original form, the so-called performative analysis in transformational grammar postulates that 'every deep structure contains one and only one per- formative as its highest clause' (Ross 1970:261). Illo- cutionary force was thought to be a property of sentences rather than utterances, and to capture the illocutionary point of a sentence Ross proposed that the highest clause of the deepest phrase marker under- lying every sentence is performative, whether or not the surface sentence contains a lexical performative.
[s[N P /] [ypPERFORMATTVE VERB (you)
[sNONPERFORMATTVE SENTENCESs]vp]s]
This was wrong, and so was the claim that a sentential phrase marker can contain one, and only one, per- formative: / say (that) I promise to be home by eight confutes both claims. Performative 'I say' is the high- est clause, but the embedded performative 'I promise' indicates the illocutionary point. Furthermore, most surface sentences do not contain a performative verb,
462