University of Virginia Library

The following remarks arise from a reading of J. L. Austin's How to do Things with Words. The most frequent references by textual critics to Austin's theories are to the terms illocution, perlocution, and locution. These are thought to correspond to intention and execution, for illocution (the way an utterance is used—as warning, advice, etc.) and perlocution (the effect aimed at by the utterance—as persuading one to respond appropriately) are thought of as the intended force and meaning of the locution (the whole act of uttering something).

There are several problems with this appropriation of Austin's ideas. Austin writes of "constative utterances" which are statements that can be tested as logically or factually true or false, and he distinguishes them from "performative utterances" which are not true or false but by which or in which the speaker does something such as warn, advise, threaten, marry, or contract—that is, the speaker does something besides just "make an utterance" (which could be said of constative statements also). Just as constative statements can be true or false, the effect of performative utterances might be "happy" or "unhappy" depending on a number of factors including the sincerity of the speaker, his right to do the performative act, and the success of the utterance within the circumstances attending it.

One should note that most of Austin's remarks are particularly apropos to spoken, rather than written, utterances—which he says are "not tethered to their origin" as speech is. He admits written utterances if they are signed. This would seem to admit books which bear the author's name; indeed they nearly always bear the publisher's name, and sometimes that of the printer, book binder, and occasionally, on a sticker inside the cover, that of the bookseller. But it is difficult to apply Austin's categories to fiction because, for one

thing, as an utterance fiction is usually multi-voiced; not only are the agents mentioned above in a sense the utterers of the work as a whole, there are the voices of the narrator (or narrators) and of each character in the book. Austin's categories seem more apropos when applied to the "speech acts" of a character than when applied to those of the author.

In addition, a book, a poem, a novel, is in a way not a single utterance but a whole string of utterances. Austin's categories apply to sentences, not to whole paragraphs. Second, though Austin develops the idea of performative utterances very interestingly, he sees them as but one kind of utterance in a range of possible kinds of utterances—"constative utterances" being one of them. Third, in developing the idea of performative utterances, Austin explains their occasional failure to succeed by listing a variety of "infelicities" which cause the utterance to "misfire" or to be "abused." Among the "abuses" of performative utterances are "insincerities" such as jokes and lying. He considers a speaker's authority and sincerity as essential elements for "happy" performative statements. That is to say, if the performative utterance is given with any other than its "sincere" surface "intention" it fails. Irony and parody, therefore, have no place in his scheme. Obviously, if Austin's scheme is to be applied to fiction or poetry, a whole range of new types of "felicities" would have to be elaborated, for there is no room in Austin's theory for an ironic utterance that succeeds in spite of itself by "intentionally" subverting its surface meaning, not in order to lie, but to convey "happily" a subsurface meaning that is expected to be understood. Austin himself considered fiction and poetry an "etiolated use of language."

Austin seems therefore an oddly weak reed to lean on in explaining the problems of textual criticism. His theories apply to single sentences uttered in spoken form. And he restricts his interests to certain types of sentences which exclude the majority of uses we associate with literary texts. Nevertheless, Austin's categories have been picked up and elaborated with some success for use with literature, in particular by Richard Ohmann, who modifies Austin's categories for analysis of jokes, irony and fiction as "happy" acts, by John Searle, who considerably extends Austin's notion of varieties of illocutionary modes and who emphasizes the importance of context in understanding the "functional" referentialness of words when that deviates from the "normal or expected" references, and by Quentin Skinner, who develops the idea of locutions as functional within circumstances and through conventions in such a way that a theory of "write acts" could be developed, except that he tends to treat the medium (sound waves and written text) as transparent vehicles.[59] Paul Hernadi drops Austin's divisions into constative and performative

utterances, adds the terms prelocutionary (why the speaker decides to make the utterance) and postlocutionary (what reaction or outcome eventuates in the listener and reader), and divides perlocution and illocution into originating and receiving counterparts. Thus the text points back to authorial prelocutionary input, perlocutionary intention and illocutionary force and at the same time provides the basis for the reader's illocutionary uptake, perlocutionary impact and postlocutionary outcome.[60] This use of Austin's terms is somewhat loose but it applies better to literary utterances than Austin's original schema which was both more precise and more rigid.

The most frequent references to Hancher are to "Three Kinds of Intention," Modern Language Notes, 87 (1972), 827-851. Hancher's three kinds of intention correspond roughly to Austin's analysis of performative utterances, except that Hancher applies his theories directly to literary texts. His terms are "active," "programmatic," "final."

By active intention, Hancher means "the act of meaning-something-by-the-finished-text." Here he subsumes Austin's idea of locution which corresponds roughly with "the-finished-text" and the idea of content or subject, "meaning-something." But it is not entirely clear whether Hancher is defining "meaning" in any other than the ordinary general sense of the word, which Austin has done a much better job of subdividing into types of meaning.

By programmatic intention, Hancher refers to the attempt to have a certain effect or elicit a certain response. Here Hancher incorporates what Austin calls perlocution and which Austin considers "happy" if it has the desired or intended effect. Hancher does not measure the "happiness" of programmatic intentions by their success, but he suggests that the motive power of "active intentions" might run counter to the motive power of "programmatic" intentions—as when the active intention to describe a scene with honesty and detail might run counter to a programmatic intention to satisfy the censors.

By final intention, Hancher means the author's intention, by the finished product, to cause a reaction or to make something happen.