At the beginning of How to Do Things with Words (1962), John Langshaw Austin challenged the common assumption that "the business of [a declarative sentence] can only be to 'describe' some state of affairs, or to 'state some fact'" (p. 1). Obviously, that is not the business of interrogative and imperative sentences, but Austin argued that even certain declarative sentences are typically used to do something other than make statements. For example, an employer can fire someone by saying "You're fired," and an employee can quit by saying "I quit." In uttering such a sentence, one is not merely saying what one is doing, one is actually doing it. Such a sentence has a remarkable property: To utter it is (typically) to perform an act of the very sort named by its main verb.
It does seem remarkable that you can do something just by saying what you are doing. Most types of acts are not like that. You cannot stand on your head by saying that you are standing on your head, and you cannot convince someone that you love them by saying that you are convincing them that you love them. Yet in the right circumstances you can fire someone or quit a job just by uttering the right sort of sentence. How is this possible, and what sorts of acts can be performed in this way? Does this phenomenon of performativity require a special explanation, perhaps involving some kind of convention, or it is just a special case of something more general?
Explicit Performative Utterances
Austin (1961) dubbed performative such verbs as "promise," "apologize," "request," "fire," and "quit." Performative sentences are generally in the first-person singular with their main, performative verb in the simple present tense, active voice. So, for example, you can promise to attend by saying "I promise to attend" (but not by saying "I promised to attend" or "She promises to attend"), and you can apologize by saying "I apologize" (but not by saying "I apologized" or "She apologizes"). The word "hereby" may be inserted before the performative verb, thereby indicating that this utterance is the vehicle of the performance of the act named by the verb. Some performative sentences are in the first-person plural ("We guarantee your safety"), the second-person singular or plural ("You are advised to get vaccinated"), or the impersonal passive ("Smoking is prohibited)." Occasionally the performative verb is in the present progressive, as in "I'm warning you to stay away" and "I'm asking you for the last time to clean up your room." Because utterances of performative sentences are characteristically performances of acts of the very sort named by their main verbs, Austin called them "explicit performative utterances," or simply "performatives."
Notice that such acts as promising, apologizing, and requesting, which Austin called "illocutionary acts," can be performed without using a performative sentence, hence without making explicit what one is doing. For example, one can promise by saying "You can count on me to … ," apologize by saying "I'm sorry," and request by saying "I'd like you to …" This raises the question whether performativity, although involving the use of a special sort of sentence, requires a special explanation. In this regard note also that performative sentences do not have to be used performatively and obviously are not so used when they are embedded in larger linguistic contexts. For example, saying "If I promise to take you to the play, will you quit nagging me?" is not to make a promise, and saying "I apologize only if I feel guilty" is not to apologize.
Performatives and Conventions
It is generally accepted that linguistic meaning is a matter of convention. So to that extent every utterance is conventional, insofar it is made with linguistic means. However, it might seem, as it did to Austin, that performatives are conventional in a more specific way and that this explains their performativity. If so, then, for example, an utterance of "I promise to …" amounts to a promise because, and only because, there is a convention, or what John Searle (1969) called a "constitutive rule," to the effect that an utterance of such a sentence counts as a promise. That is, roughly, it counts as such only because it is generally recognized to count as such. This view seems plausible as regards certain institution-bound performatives, where a specific form of words is designated, and often required, for the performance of an act of a certain sort. For instance, uttering the words "I pronounce you husband and wife" counts (in the requisite circumstances) as the act of marrying a couple; uttering "The jury finds the defendant guilty" counts as finding the defendant guilty (convicting the defendant); and uttering "(I) double" counts as doubling in bridge. Indeed, in institutional contexts there are often designated expressions that, though not performative in form, have the same effect, such as an umpire's "Out," a legislator's "Nay," or a judge's "Overruled." Of course, these specialized performatives and other designated forms of words have to be uttered by the appropriate person in the appropriate circumstances, but the relevant convention provides for this. Not just anyone can adjourn a meeting, sentence a convicted criminal, or christen a ship, and not just under any circumstances (with his "doctrine of the Infelicities" Austin classified the various ways in which such utterances can go wrong as "flaws," "hitches," and "abuses" [1962, pp. 12–38]). So it does seem that in institutional cases performativity is a matter of convention: A certain person's uttering a certain form of words in a certain context plays a certain official role because, and only because, it is generally recognized as so doing.
However, as P. F. Strawson (1964) contended, Austin was overly impressed with institution-bound cases. In such cases there are specific, socially recognized circumstances in which a person with specific, socially recognized authority may perform an act of a certain sort by uttering words of a certain form in order to effect, or officially affect, institutional states of affairs (see Bach and Harnish 1979, ch. 6). Ordinary performative utterances, on the other hand, are not bound to particular institutional contexts. Like most illocutionary acts, Strawson argued, they involve an intention not to conform to an institutional convention but to communicate something to an audience. An utterance counts as a promise, an apology, or a request because, and only because, the speaker intends it to count as such and the audience, recognizing that intention, regards it as such. To be sure, it is only under certain circumstances that a speaker will make such an utterance with such an intention and his audience will so regard it, but this is not in virtue of any convention.
It might be suggested, as it was by Jerrold Katz (1977), that performativity is explained not by social conventions but by linguistic ones. Perhaps there is some distinctive feature of the meaning of performative verbs that explains how one can perform an act of the very sort named by the verb by uttering a performative sentence containing that verb. However, this suggestion loses its plausibility when one takes into account a range of linguistic data beyond the simple performatives considered so far. In particular, there are what Bruce Fraser (1975) called "hedged performatives," which philosophers have largely overlooked, such as "I can promise you … ," "I must ask you … ," and "I would like to invite you …" Utterances of such sentences standardly have performative effect, but the meanings of the sentences themselves are not inherently performative. This is clear because without contradicting myself I could say "I can promise you, but I won't," "I must ask you, but if I did, my wife would never forgive me," or "I would like to invite you, but I can't." In each of these cases I would not be performing an act of the type in question but would merely be telling you that I am able to promise, that I am required to ask you, or that I would like to invite you. In addition, there are other sorts of sentences that, unlike hedged performatives, do not even contain performative verbs but which are standardly used in the same kind of way: "It would be nice if you …" to request, "Why don't you … ?" to advise, "Do you know … ?" to ask for information, "I'm sorry" to apologize, and "I wouldn't do that" to warn. Clearly these standard uses are not predictable from their linguistic meanings alone.
The variety of forms of sentences that are standardly used to perform acts of the same types as those accomplished by explicit performative utterances suggests that performativity is not a matter of convention, whether social or linguistic. Performativity requires no special explanation. Rather, its explanation belongs to the general theory of speech acts (see Searle 1989 and Bach and Harnish 1992 for two contrasting accounts). Performative sentences are just one kind among various kinds of sentences that are standardly used to perform types of illocutionary acts not predictable from their meanings alone (see Bach and Harnish 1979, ch. 10). Performativity is a pragmatic phenomenon not a semantic one, a matter of language use rather than linguistic meaning. The standardization of performative and other forms of sentences for uses not predictable from their meanings does not show that they are governed by special conventions but merely that there is a practice of using sentences of certain forms in certain ways. The claim that they are conventional falsely entails that an utterance of a certain form of words would not have the force it has unless it is generally recognized to count as such. The claim that they are merely standardized for these special uses requires something less. Standardization merely streamlines the inference the hearer must make to identify the speech act being performed; it creates the illusion of conventionality where there is really but a pragmatic regularity. (For further discussion of these issues, see Reimer 1995, Bach 1995, and Harnish 1997).
Performatives and Statements
When introducing the notion of performatives, Austin contrasted them with utterances like "I state that … ," "I claim that … ," and "I predict that …" These explicit constatives are like utterances of ordinary declarative sentences in that they "describe some state of affairs, or to state some fact," which Austin denied that performatives do. Yet he came to realize that explicit constatives are relevantly similar to explicit performatives: Their main verbs also make explicit the type of act being performed. After all, an assertion or a prediction is made with "I assert …" or "I predict …" in just the same way that a promise or a request is made with "I promise…" or "I request …" Accordingly, what makes explicit performatives distinctive is not what the speaker does but that the speaker makes explicit what he or she is doing.
Austin also came to realize that what can be done explicitly without a performative can also be done without making explicit the type of act being performed. In the latter part of How to Do Things with Words he developed the distinction between locutionary and illocutionary acts, which effectively superseded the distinction between constative and performative utterances. Locutionary acts are acts of saying something, and illocutionary acts are performed in the act of saying something. This distinction applies not only to promises, requests, and apologies, but also to statements and the like (Austin retained the term "constative" for them).
For example, in uttering "I promise to be there" and thereby explicitly saying that one promises to be there or in uttering merely "I will be there" and thereby just saying that one will be there, one can promise to be there. Similarly, in uttering "I state that Mars has two moons" and thereby explicitly saying that one states that Mars has two moons or in uttering merely "Mars has two moons" and thereby just saying that Mars has two moons, one can state that Mars has two moons. Note that stating is distinct from saying. In the right circumstances, one might say that Mars has two moons but state, albeit figuratively, that a certain belligerent person has two obsequious functionaries. In general, a speaker need not make explicit what he or she is doing in order to do it. Explicit performatives do have a distinctive self-referential character, but that does not mean that their illocutionary force requires special explanation. Indeed, if the successful "performance of an illocutionary act involves the securing of uptake" (Austin 1962, p. 116), then if anything it should be easier for an explicit performative to succeed, precisely because the speaker is saying what he or she is doing.
One remaining question concerns whether performatives are statements too (see Bach 1975), contrary to Austin's insistence that making explicit "is not the same as stating or describing" (1962, p. 61). When he introduced the category of explicit performative utterances, he claimed that even though they are utterances of declarative sentences, they are not cases of making statements and are not descriptive. However, this does not seem right, for the simple reason that the verbs in performative sentences can be modified, as in "I gladly promise … ," "I sincerely apologize … ," and "I reluctantly request …" This strongly suggests that a speaker of such a sentence would be making a statement. The speaker would be describing himself or herself, as promising gladly, apologizing sincerely, or requesting reluctantly.
Performatives have even been described as "self-verifying" (originally by Lemmon 1962 and more recently by Johansson 2003). Clearly they are self-referential, in that if one utters a performative sentence and uses it performatively, one is making explicit what one is thereby doing. But to describe them as self-verifying is to claim that they make themselves true. This seems right, but notice that a performative is not self-verifying in the way that an utterance of, for example, "I am speaking" or "I am alive" is self-verifying. It is not the bare fact of the utterance that, given its content, makes it true. Suppose I utter "I hereby apologize" and thereby apologize. It is true that I am thereby apologizing, but what makes this true is that I am using the sentence to perform the illocutionary act of apologizing. In that way, it is self-verifying.
Does this self-referential, self-verifying character help explain performativity, is it just a curious feature of explicit performative utterances, or what? As Searle (1989) has argued, the performativity of performative utterances does not depend on their being self-verifying. That gets things backwards: they are self-verifying statements because of their performativity. However, as Kent Bach and Robert Harnish (1992) have argued, their character as statements plays a key role in the speaker's being able to communicate to his audience what he is doing, precisely because he is using a performative to make explicit what he is doing.
Austin, J. L. How to Do Things with Words. Cambridge, MA: Harvard University Press, 1962.
Austin, J. L. "Performative Utterances." In his Collected Papers, edited by J. O. Urmson and G. J. Warnock. Oxford: Oxford University Press, 1961.
Bach, Kent. "Conventionalization vs. Standardization." Linguistics and Philosophy 18 (1995): 677–686.
Bach, Kent. "Performatives Are Statements Too." Philosophical Studies 28 (1975): 229–236.
Bach, Kent, and Robert M. Harnish. "How Performatives Really Work." Linguistics and Philosophy 15 (1992): 93–110.
Bach, Kent, and Robert M. Harnish. Linguistic Communication and Speech Acts. Cambridge, MA: MIT Press, 1979.
Fraser, Bruce. "Hedged Performatives." In Syntax and Semantics. Vol. 3, edited by Peter Cole and Jerry Morgan. New York: Academic Press, 1975.
Ginet, Carl. "Performativity." Linguistics and Philosophy 3 (1979): 245–265.
Grice, H. P. "Meaning." Philosophical Review 66 (1957): 377–388.
Harnish, Robert. M. "Performatives and Standardization: A Progress Report." Linguistische Berichte 8 (1997): 161–175.
Johansson, Ingvar. "Performatives and Antiperformatives." Linguistics and Philosophy 26 (2003): 661–702.
Katz, Jerrold J. Propositional Structure and Illocutionary Force. New York: Crowell, 1977.
Lemmon, E. J. "Sentences Verifiable by their Use." Analysis 12 (1962): 86–89.
Recanati, François. Meaning and Force: The Pragmatics of Performative Utterances. Cambridge, U.K.: Cambridge University Press, 1987.
Reimer, Marga. "Performative Utterances: A Reply to Bach and Harnish." Linguistics and Philosophy 18 (1995): 655–675.
Searle, John R. "How Performatives Work." Linguistics and Philosophy 12 (1989): 535–558.
Searle, John R. Speech Acts: An Essay in the Philosophy of Language. Cambridge, U.K.: Cambridge University Press, 1969.
Searle, John R. "A Taxonomy of Illocutionary Acts." In Language, Mind, and Knowledge, edited by Keith Gunderson. Minneapolis: University of Minnesota Press, 1975.
Strawson, P. F. "Intention and Convention in Speech Acts." Philosophical Review 73 (1964): 439–460.
Kent Bach (2005)