Performative Utterances

It is an earlier and shorter treatment of themes that Austin worked out at greater length in his William James Lectures, published after his death as How to Do Things with Words. In the paper he still treats the performative as a distinct class contrasted with ordinary statements. In the lectures he came to doubt that contrast and moved to a general theory of speech acts. The paper is therefore useful for seeing the starting point of the argument, but readers should not take its sharper formulations as his final view.

A performative utterance is one whose utterance is, in the right circumstances, the doing of an action: saying 'I promise' is promising, and saying 'I do' in a marriage ceremony is marrying. It does not report an action performed elsewhere. Austin contrasts this with constative utterances, which describe or state and are assessed as true or false. He notes that the performative is typically first-person and present tense, though he admits that the grammatical marks are not decisive, a worry that later led him to revise the classification.

Austin argues that a performative can go wrong in ways other than being false. Some fail because no accepted procedure exists or the persons or circumstances are not appropriate, as when someone without authority tries to christen a ship. Others fail because the procedure is executed incorrectly or the speaker lacks the required thoughts and intentions, as in an insincere promise. These failures show that the success of speech depends on conventions and circumstances as well as meaning, a point that attracted attention in legal and social theory.

The criticism begins with Austin himself, who found that statements also depend on conditions of felicity and that performatives can be assessed against the facts. Later, John Searle reworked the theory into a rule-based account of illocutionary acts, and others, such as Stanley Cavell, explored the philosophical consequences of treating utterances as acts. Jacques Derrida criticized Austin's reliance on ideal conditions and speaker intention, and his exchange with Searle is a well-known dispute. The paper is a common starting point for these debates.

Austin's method here is typical of his work in that he proceeds by attending closely to how we actually speak, using everyday examples such as naming, betting, and bequeathing, instead of building an abstract theory first. The paper shows how an ordinary linguistic phenomenon, long overlooked by philosophers who assumed language mainly describes, can reveal a structural feature of meaning. Readers interested in the method should compare it with his essays on excuses and on knowledge in the same collection.

A frequent misreading assumes performatives are a small set of ceremonial utterances such as wedding vows, so that the idea is of limited interest. Austin's point is broader: that language is used to perform many acts, and that the descriptive view of meaning is only part of the story. Another misreading takes the theory to concern the speaker's private intentions. Austin gives weight to conventions and circumstances, and treats intentions as one condition among several, not as the whole of what makes an utterance succeed.