An act of speech (or act of speech) is a means implemented by a speaker to act on his environment by his words: he seeks to inform, incite, ask, convince, promise, etc. his or her interlocutors by this means.
This theory, related to the philosophy of ordinary language, was developed by John L. Austin in Quand dire c’est faire (1962), then by John Searle. She insists that besides the semantic content of an assertion (its logical meaning, independent of the real context), an individual can address another in the idea of doing something, namely to transform representations of things and goals of others, rather than simply saying something: then we speak of a performative utterance, in contrast to a constative statement. Unlike the latter, it is neither true nor false.
We can then model the act of speech like any other type of act: it has a goal (also called communicative intention), a pre-requisite, à body (that is to say a realization) and an effect.
There are different types of speech acts, which are generally categorized according to their purpose: to quote, inform, conclude, give an example, decree, deplore, object, refute, concede, advise, distinguish, move, exaggerate, irony, minimize, mock, reassure, rectify … The identification of the speech act largely conditions the interpretation of the delivered message, beyond the understanding of its semantic content. For example, the motivation for the statement “I learned that you graduated” may be to congratulate the recipient, to apologize for doubting their success, to joke about a late success or simply inform him of the reported fact.scholastic philosophers (in the context of sacramental theology), as well as Thomas Reid and CS Peirce. Adolf Reinach developed a rather complete theory of “social acts” as performative expressions, although his work had little influence, perhaps because of his premature death. Roman Jakobson had similar ideas in the 1960s , in the form of what he calls the conative function of language .
Linguistic pragmatics has largely developed on the basis of the theory of speech acts, which has historically constituted the crucible. The theory of speech acts is based on the idea that the function of language, even in declarative sentences, is less to describe the world than to facilitate actions (“Knowing, but knowing to act”, according to the formula of Roger Bacon). The question is obviously little debatable concerning orders, promises, advice or institutional acts (baptism, marriage, etc.). Its development by Searle, following Austin as pioneer, has influenced the recent development of linguistic pragmatics . The cognitive pragmatics that emerged with the work of Dan Sperberand Deirdre Wilson in the 1990s, takes a different approach. According to Sperber and Wilson, who have proposed a greatly simplified categorization of speech acts, performative statements are infrequent, or at least the interpretation of many of these utterances is not based on their eventual performative character (ie that these statements make sense regardless of the intention to communicate). In other words, one should not exaggerate the scope of speech acts.
Pragmatics can be considered to have been born in Harvard in 1955, when John Austin gave the William James lectures and introduced the new notion of “speech acts”. Thus, contrary to what one might think, pragmatics takes root in the work of a philosopher who protests against the tradition in which he was educated and according to which language is used mainly to describe reality. Austin, in opposition to this “vericonditionalist” conception of the function of language, which he pejoratively calls the descriptive illusion , defends a much more “operationalist” vision according to which language serves to accomplish acts. He bases his theory of language and its use on the examination of certain statements of form affirmative, in the first person of the singular of the present indicative, active voice, utterances which, according to Austin, in spite of their grammatical form, would describe nothing (and would therefore be neither true nor false) but rather correspond to the execution of an action.
The theory of acts of speech is therefore based on an opposition to the “descriptivist illusion” which states that language has the primary function of describing reality and that affirmative statements are always true or false. According to the theory of acts of language, on the contrary, the function of language is as much to act on reality and to allow the one who produces a statement to accomplish, in doing so, a reaction. From this perspective, the statements are neither true nor false.
Austin’s thesis, in its first version at least, is based on a distinction among the affirmative statements between those who describe the world and those who perform an action .
· (1) The cat is on the doormat.
· (2) I promise you I’ll take you to the movies tomorrow.
The former are called constatives , while the latter are performative . The first can receive a truth value: thus (1) is true if and only if the cat is on the doormat. The latter can not receive a truth value. However, they may be happy or unhappy , the act may succeed or fail and, just as the truth values attributed to the constative statements depend on the truth conditions attached to them, the happiness of a performative utterance depends on its conditions. of happiness .
The conditions for happiness depend on the existence of conventional (sometimes institutional: marriage, baptism, etc.) procedures and their correct and complete application, appropriate or inappropriate mental states of the speaker, the fact that the subsequent conduct of the speaker and the interlocutor complies with the requirements of the spoken word act. More generally, there are two primitive conditions of success :
· the speaker must address someone,
· his interlocutor must have understood what he was told in the statement corresponding to the act of speech.
The performative / constative distinction, based as it is on the distinction between the condition of happiness and the conditions of truth, has not withstood the severe scrutiny to which Austin has subjected it. In particular, he noticed that besides explicit performatives like (2), there are implicit performatives like (3), which can also correspond to a promise, but where the verb “promise” is not explicitly used:
· (3) I’ll take you to the movies tomorrow.
Moreover, the constatives correspond to implicit speech acts, so assertion acts are subject to conditions of happiness, as are the performatives. Finally, they can be compared to their explicit performative correspondent, like (4), which definitely ruins the performative / constative distinction:
· (4) I say that the cat is on the doormat.
The opposition between conditions of happiness and conditions of truth is therefore not complete (they can be combined on the same statement), and consequently, the opposition between performatives and constatives is not as decided as there is appeared on a first examination. Austin abandons the opposition of constants and performative statements and builds a new classification of speech acts into 3 categories:
· Locutory acts that one accomplishes when one says something and independently of the meaning that one communicates;
· The illocutionary acts that one accomplishes by saying something and because of the meaning of what one says;
· The perlocutionary acts that one accomplishes by having said something and who are consequences of what one has said.
If we return to the example (2), the simple fact of having pronounced the corresponding sentence, even in the absence of a recipient, is enough to perform a locutionary act. On the other hand, the statement of (2) has accomplished an illocutionary act of promise if and only if it has been pronounced (2) by addressing a recipient who might understand the meaning of (2) and that act illocutionary will be happy only if the conditions of happiness which are attached to it are fulfilled. Finally, we will have by the enunciation of (2) performed a perlocutory act only if the understanding of the meaning of (2) by the recipient results in a changé in his beliefs: for example, he can be persuaded, thanks to the enunciation of (2), that the speaker has a certain benevolence towards him.
In accordance with his doubts as to the constative / performative distinction, Austin admits that any utterance of a complete grammatical sentence under normal conditions is therefore consistent with the performance of an illocutionary act. This act may take different values depending on the type of act performed and Austin distinguishes five major classes of illocutionary acts:
· The verdictifs or legal acts (discharge, convict, decree …);
· The exercitifs (degrade, order, order, forgive, leave …);
· The promissifs (promise, vow, secure, bet, swear …);
· The comportatifs (apologize, thank, lament and criticize …);
· The expository (affirm, deny, apply, notice …).
Austin’s death prevented him from continuing his work, and the development of speech act theory was continued later by John Searle.
The centrality of the notions of intention and convention does not really constitute a break with the Austinian theory of speech acts. Rather, Searle is content to explicitly indicate concepts that had remained more implicit in Austin. Searle’s main innovation is to distinguish two parts in a statement: thepropositional content marker and the illocutionary force marker. If we go back to example (2), we see that it is easy to distinguish, as in most explicit performatives, the propositional content marker: “I’ll take you to the cinematomorrow”, and the illocutionary force marker: “I promise you”. While this distinction is easier to apply to explicit performatives such as (2), the expressability principle assumes that implicit performatives, like (3), are equivalent to explicit performatives and that, to this extent, the distinction between illocutionary force and propositional content marker can apply to it.
Searle also gave his version of the rules applying to different types of speech acts and his own taxonomy of these different types of speech acts. This taxonomy is based on a number of criteria:
· the purpose of the illocutionary act;
· the direction of adjustment between words and the world – the words “adjust” to the world, as in an assertion, or the world “adjusts” to words, as in a promise;
· the differences in propositional content that are determined by mechanisms related to the illocutionary force: a promise , p. eg, will determine the propositional content of the statement in such a way that this content will relate to the future, and to something in my power; an excuse will determine the content so that it relates to a past event, and that was under my control;
· the force with which the illocutionary purpose is represented, which depends on the degree of explanation of the act;
· the respective status of the speaker and interlocutor and their influence on the illocutionary force of the statement;
· the relations of the statement with the interests of the speaker and the interlocutor;
· relationships to the rest of the speech;
· the differences between the acts which necessarily pass through the language (to take an oath) and those which can be accomplished with or without the language (to decide);
· the difference between institutional acts;
· the existence or not of a performative verb corresponding to the illocutionary act;
· the style of accomplishment of the act.
This somewhat heterogeneous set of criteria allows Searle to identify five major classes of speech acts, based primarily on the first four criteria:
· the assertive [representative] (statement, statement …); words fit the world;
· the directives (order, request, advice …); the world adjusts to words;
· the promissifs (promise, offer, invitation …); the world adjusts to words;
· the expressive (congratulations, thanks …); no adjustment direction;
· the declaratives (declaration of war, appointment, baptism …); double adjustment direction (words – world / world – words).
To end the impact of Searle’s theory of speech acts, current attempts to formalize speech act theory are based on Searle’s work.
On the basis of the principle of expressiveness, a linguist in the generative semantics movement , John R. Ross , proposed in 1970 a hypothesis which has met with a certain fortune under the name of a performative hypothesis . The performative hypothesis consists in treating the implicit performatives, like (3), as equivalent to explicit performatives, like (2)
· (2) I promise you I’ll take you to the movies tomorrow.
· (3) I’ll take you to the movies tomorrow.
More specifically, in the framework of the generativist distinction between surface structure and deep structure, the performative hypothesis consists in supposing that an utterance that has (3) as a surface structure nevertheless shares the same deep structure as a statement that has (2) as surface structure. In other words, every utterance has in its deep structure a performative preface (I promise that, I order that, I assume that …), that this performative preface is explicitly expressed (that it belongs to its surface structure) or that it is not. This hypothesis had the advantage, which explains the repercussion it had had, to give a more reliable basis for the Searlian distinction between illocutionary force marker and propositional content marker: the performative preface, present in all hypothetical statements. ,
However the performative hypothesis encountered a fundamental objection, perfomadox. It consists in pointing out that insofar as the deep structure of a sentence corresponds to its semantic analysis (its logical form) and to the truth conditions of the sentence in question, the performative hypothesis leads to a paradox for most assertions. According to the performative hypothesis indeed, a sentence like (1) has the same deep structure as a sentence like (4), so they would be identical. In other words, we could no longer say that (1) is true if and only if the cat is on the mat, but we should say that (1) is true if and only if I affirm that the cat is on the mat. However, it goes without saying that the truth of (1) does not depend on whether the speaker asserts anything, but depends on whether the cat is on the mat.
· or should not be interpreted semantically (to avoid generating incorrect truth conditions), but hence the sentence is not interpretable;
· or must be interpreted semantically (so that the sentence is interpretable), but from then on the sentence is attributed to incorrect truth conditions.
These two consequences being unacceptable, the author of the performadoxe, William G. Lycan, concludes that the performative hypothesis must be abandoned.
Linguistic pragmatics developed on the basis of Austin’s rejection of “the descriptive illusion,” the common-sense thesis that language is used to describe reality. Austin, and Searle after him, substituted another thesis, that the main function of language is to act on the world rather than to describe it. The idea that every utterance of a complete grammatical sentence corresponds to an illocutionary act has the corollary of the need to identify, for each statement, its illocutionary force, which is an indispensable part of its interpretation. Linguistic pragmatics, based on the theory of speech acts, has tended to emphasize “the conventional and codic aspect of language” and to ignore its “under-determination”. Faced with a statement, indeed, the theory of acts of speech, especially because of the principle of expressability, admits that the interpretation is essentially conventional. The last ten to fifteen years, however, have seen the emergence of a pragmatic, cognitivist current, which, following the generative school, sees language as a means of describing reality (and only incidentally a means of action), which emphasizes language under-determination and the importance of inferential processes in the interpretation of utterances. This new approach to pragmatic problems, the following the generative school, language is first and foremost a means of describing reality (and only incidentally a means of action), and which emphasizes language under-determination and the importance of processes. inferential in the interpretation of statements. This new approach to pragmatic problems, the following the generative school, language is first and foremost a means of describing reality (and only incidentally a means of action), and which emphasizes language under-determination and the importance of processes. inferential in the interpretation of statements. This new approach to pragmatic problems, thetheory of relevance , is now widely accepted and its authors, Dan Sperber and Deirdre Wilson , have questioned a number of the principles underlying speech act theory.
In addition to the principle of expressability, the difficulties encountered by the performative hypothesis indicate the limits, Sperber and Wilson question the very relevance of the classifications of speech acts proposed by Austin and Searle. They note, rightly, that if the determination of a precise illocutionary force is both possible (and necessary) in some cases, in many other cases it is very difficult, if not impossible, and not does not seem essential to the interpretation of a statement. Thus, in example (5), it is unclear whether this statement corresponds to an act of promise, preaching, or threat, and it does not actually seem to determine whether it is one or on the other hand is indispensable to the interpretation of (5):
· (5) It will rain tomorrow
Sperber and Wilson’s suggestion is to drastically reduce classes of speech acts to three classes that can be linguistically identified (via the lexicon or syntax), namely the acts of ” saying ”, ” telling ” and to ‘ask if’ “:
· the acts of ‘saying that’ correspond roughly to the declarative sentences and especially to the assertions, the promises, the predictions …;
· the acts of ‘say de’ roughly correspond to imperative sentences, orders, councils …;
· the acts of ‘asking if’ correspond to interrogative sentences and more generally to questions and requests for information.
One might object to this approach that it does not take into account institutional acts. Sperber and Wilson anticipate this objection and respond in advance by pointing out that the rules governing institutional acts (baptism, marriage, conviction, opening of session, announcements bridge …) are not linguistic rules, nor cognitive rules, but more of a sociological study and institutional acts can enter the first great class, that of ‘saying that’. There remains one possible objection to this approach: one might think that it implies a new version of the performadoxe. If, indeed, we admit that a statement like (1) is equivalent to ‘I say that the cat is on the mat’, it is quite obvious that this interpretation will have the same consequences as the performative hypothesis, that of producing incorrect conditions of truth for (1). To avoid this objection, it must be admitted that the attribution of the illocutionary force goes through a pragmatic process which delivers, not the complete semantic interpretation of the statement, but an enrichment of this interpretation which does not intervene in the determination of the truth conditions of the utterance.