Traditional speech act theory

Doing thing with words: J. L. Austin

The language philosopher J. L. Austin ([1962]1975: 5) observed, at the beginning of his published lectures, that some utterances “do not ‘describe’ or ‘report’ ” something that is “ true or false”, that is to say, they are not truth conditional. In fact, they are not a matter of “ ‘just’ saying something” but of “doing an action” (ibid.). This essential observation, expounded in his works, gave raise to speech act theory, a theory that underpins much of pragmatics.

Austin initially focused on a special class of verbs, performative verbs, whose purpose is not to “describe” things but, “in the appropriate circumstances”, to “do” them (ibid.: 6). Some examples are:

A cautionary note is in order here. The performativity of performative verbs does not rest entirely in their semantics but is sensitive to both linguistic and non-linguistic context. For example, if I said I apologize all the time, that usage of apologize does not seem to be performing an act of apology but describing what I do all the time. Performative verbs name the action they perform (e.g. apologize names the apology which the utterance performs); they have a metapragmatic aspect. If this is the case, there is a sense – contrary to Austin – in which such verbs are not simply engaged in performance (see Bach and Harnish 1992). At one level they name or describe the act being performed (e.g. whether somebody is pronouncing, sentencing, etc.), and at this level are truth conditional (e.g. whether someone is truly doing it). At another more indirect level they actually perform the action in the context of a wedding, and so on.

The utterances in which performative verbs occur are called performative utterances or simply performatives (Austin 1975: 6). Subsequent researchers (e.g. Thomas 1995) have pointed to different types of performative utterances, including:

Ritual performatives e.g. I baptize you, I name this ship [These are associated with highly institutionalized contexts.]

Self-referential performatives e.g. I apologize, I plead not guilty [These are somewhat ritualistic too, or at least formal, but note that because they refer to what the speaker is doing, they are difficult to falsify (literally speaking, one could not accuse someone of not apologizing, pleading guilt, etc., when they have just labelled what they have done).]

Collaborative performatives e.g. I bet you, I challenge you [These require the collaboration of another participant, e.g. if you propose a bet, it only works if the other participant accepts the bet.]

A test of whether one’s utterance contains a performative verb, thus making it a performative utterance, is the hereby test (Austin 1975: 57–61; other tests are elaborated in Austin 1975: 79–80). If you can insert the word hereby between the subject and the verb, and it sounds acceptable, you probably have a performative verb on your hands. All of the above examples could accept the word hereby before the performative verb. In these contexts, hereby has the sense of “by these words or behaviors” I perform the action named by the verb. This is why examples such as I hereby am tired sound very odd. However, the test is not perfect. One problem is that the formality of the word hereby makes it sound odd for “everyday” apologies, promises, etc.. This formality, and specifically its association with institutional contexts, can be seen in the words with which it collocates most frequently. The top 25 most frequent collocates of hereby in the two-billion word Oxford English Corpus (OEC) are as follows:

declare, agree, give, grant, order, find, sentence, acknowledge, consent, authorize, certify, undertake, invite, confirm, present, dismiss, call, announce, proclaim, request, covenant, appoint, offer, release, pledge

The actions of granting, sentencing, consenting, authorizing and so on are all clearly associated with power in institutional contexts, and indeed scrutiny of the contexts of these collocates reveals governmental, legislative, bureaucratic and so on contexts. This association is also true of some of the other more innocuous items. For example, find is associated with the expression “find you (not) guilty” issued in the courtroom. A further problem is that hereby can sound fi ne in an utterance, yet need not have anything to do with performative verbs. In the following example, it refers to the means, spelt out in the first sentence, by which a fi lm reappeared:

[6.3]

The fetish value of films like Human Lanterns is proportional to the inflated wistfulness with which we desire them, and surpluses of pristine DVDs make us wistful no longer, deflating our infantile fantasies of cult-film deviance and transgression into a hesitant sigh of belated self-examination. Two such “deflationary films” by Shaw Brothers workhorse Ho Meng-hua have hereby resurfaced. (OEC)

However, we should also note that performative utterances often have other formal characteristics, including the following grammatical features (Austin 1975: 55–66) (the asterisk means that the utterance is infelicitous as far as being a performative promise is concerned):

• past tense (*I promised to cook you a meal);

• simple or progressive aspect (I am promising to cook you a meal. *I am promised to cook you a meal); • declarative and non-hypothetical (*Shall I promise to cook you a meal? *Promise that you will cook me a meal. *I might promise to cook you a meal); and

• subject is usually a first person pronoun, but second or third person is possible if the speaker is acting as a “mouthpiece” (You are hereby promised [ ... ]. Lancaster University hereby promises you [ ... ]).

To these we could add other features, such as prosody (e.g. a rising intonation indicating an enquiry or question), or non-verbal actions (e.g. gestures) (cf. Austin 1975: 73ff.).

Beyond formal characteristics, performative utterances are crucially dependent not on truth conditions but on a set of contextual factors, the “appropriate circumstances” mentioned earlier. When something “goes wrong” with the act of the performative utterance, including its performance in inappropriate circumstances, it is not “false but in general unhappy” (ibid.: 14). Things that can be or go wrong are termed “infelicities”, and are listed as a set of “necessary conditions” (ibid.) for a performative to function happily – hence felicity conditions. Austin (1975: 12ff.) devoted some space to these felicity conditions, because they are important. For example, if a judge sentenced somebody in a courtroom according to “an accepted conventional procedure having a certain conventional effect” (1975: 14) (e.g. using the conventional expression I sentence you to X), one could not dismiss

the sentencing as false (untrue), as the judge would have acted in this respect felicitously in performing the act of sentencing. Compare this with a judge saying, for example, you’re going down for a long time. If it turned out that the judge was in fact the courtroom cleaner, the act would be infelicitous with respect to “the particular persons or circumstances” being “appropriate for the invocation of the particular procedure invoked” (1975: 15). If the judge had sentenced the defendant before the verdict was reached, as the Queen of Cards tries to do in Lewis Carroll’s fictional story Alice in Wonderland, or had fainted after saying just I sentence you, the act would be infelicitous with respect to the procedure being “executed by all participants both correctly and completely” (ibid.). Some performatives require participants to have certain “thoughts, feelings and intentions” (ibid.). Apologies are a case in point. Here, the apologizer must (appear to) be contrite about what they have done. Not only this, but with such acts Austin adds that the participant “must actually so conduct themselves subsequently” (1975: 15). Somebody who apologizes for tripping somebody up and then does it again and again, would be acting infelicitously.