Austin's How to Do Things with Words: Speech Act Theory

Michael BrenndoerferUpdated May 13, 202513 min read

In lectures published in 1962, J. L. Austin examines what people do in speaking and the conditions that make an utterance succeed.

PhilosophyAll chapters
Reading tools Notes & comments
Notes & comments
Part 8 of 12Logic, Language, and Interpretation
Chapter 102Doing things by speaking › Making a promise
Reading the full chapter · 9 sections · 13 min9 sections · 13 min

All 9 sections · 13 min read

Doing things by speaking › Making a promiseFull chapter · Sections in this chapter1 / 9 ↓

Making a promise

A friend says, “I promise to pay you back on Friday.” You can immediately ask what the friend has undertaken. You can also ask later whether they kept their word or whether they ever intended to pay. Those questions have different answers. J. L. Austin starts with utterances like this because it is easy to see that the words can be part of performing an action. One of his examples is the appropriate formula for naming a ship at its launching. Someone needs to make the utterance at the right point in the procedure, in the right capacity, for there to be a naming. Austin, lecture I, pp. 5–9.

Austin delivered the William James Lectures at Harvard in 1955. The material appeared posthumously as How to Do Things with Words in 1962, edited by J. O. Urmson. The sequence of lectures matters because Austin changes the framework along the way. His initial examples separate cases that philosophers often handled together, while the later lectures ask what the apparently different utterances have in common. Austin, editor’s preface and lectures I–II and XI–XII.

Consider the difference between “The meeting begins at nine” and “I declare the meeting open.” A participant who recites the second sentence while waiting for the chair has done something different from chairing the meeting. A chair who utters the first sentence can be reporting a schedule or reminding those who are still talking. These examples suggest why a complete account needs the utterance and the occasion together. The same words can do different work, and two different expressions can do much the same work.

When truth is only one question

Austin opens with a habit he calls the descriptive fallacy. A sentence that looks like a statement is easily assumed to describe something and to matter only as a bearer of truth or falsity. Looking at cases of apologizing, betting, and bequeathing reveals other questions we routinely ask. The difference between apologizing and reporting that an apology took place is readily available in ordinary practice. The lectures use it to test what sorts of evaluation each case invites. Austin, lectures I, V, and VIII.

When someone says “I apologize” to the person they wronged, the words, given the circumstances, count toward making the apology. A bystander who later says “She apologized” is reporting that the apology was made. These uses have different relations to an existing event. The report can be mistaken, while the speaker issuing the apology might fail to apologize in circumstances where the intended recipient cannot tell it was addressed to them.

The Core Insight: Words That Act

Austin introduces the term performative through cases where uttering the words can be the performance of an action, as with naming the ship. The verb in an explicit form such as “I promise” helps state its own force. Promises can also be made without using that verb, and every occurrence of the formula need not be a promise. If I recite it as a line in a play, no commitment to repay you on Friday follows from my reading it. Austin, lecture I, pp. 5–7, and lecture VI.

The performative as a starting point

Austin initially introduces performatives as utterances that are evaluated for how well their action is performed. Constatives are introduced by their role in stating how things stand. The later lectures ask how much these ways of evaluating speech overlap.

His first move was to draw a careful distinction between two kinds of utterances, a distinction he later abandoned as too simple but which remains a useful entry point. On one side we have constatives: statements that describe the world and can be true or false. On the other side we have performatives: utterances that perform an action and that, rather than being true or false, are subject to a different kind of evaluation, whether they succeed or fail, are appropriate or inappropriate, genuine or hollow.

Consider saying the familiar marriage formula as part of a ceremony. Austin stipulates that an accepted procedure is in place and that circumstances have to be right for the formula to do its work. An unauthorized bystander might imitate the words perfectly without producing the ceremony’s intended effect. A participant might also give the required answer insincerely. The difference matters: where standing is absent the act can misfire, while some defects concern the way a successfully invoked procedure was used. Austin, lectures I–II.

The first lectures ask whether looking at these failures gives us a cleaner test than grammatical form. Imagine hearing “I bet you that the horse will win” after the race is over. The timing can prevent it from counting as a bet of the intended kind. Compare a bet made in time and in a way the other person accepts, but made on a horse that loses. What first looked like the same result, the speaker was wrong, includes importantly different ways for things to go badly. Austin, lecture II, pp. 13–18.

How Performatives Succeed: Felicity Conditions

To make this rigorous, Austin catalogued the conditions under which a performative succeeds. These are the felicity conditions, the requirements Austin uses to assess whether an utterance is fully felicitous. Failures of these requirements can have different effects.

Austin sets out six conditions in lecture II. The letters group related failures, and he says he is making a preliminary scheme, one that admits overlap between cases. The point is to make an unfamiliar kind of diagnosis possible. A listener can distinguish whether people were in a position to act, whether they went through the steps, and whether the act they performed was later abused. Austin, lecture II, pp. 14–18, and lecture III.

The A conditions concern an accepted procedure and an occasion to which it can apply. A.1 asks whether the procedure exists and has the relevant conventional effect. Suppose a club recognizes no motion by which one member can take over as its chair. Asserting such a takeover at a meeting cannot by itself supply that missing procedure. A.2 asks whether the actual people and circumstances suit a procedure that does exist. A club that permits appointments by its committee does not thereby permit any visiting observer to make one. Austin, lecture II, pp. 14–15.

The B conditions distinguish doing the procedure correctly from doing it completely. Suppose a committee can appoint a chair by putting a proposal to a vote and announcing the result. Announcing that the wrong member was elected is an example of doing a required step incorrectly. Proposing the person and never taking the vote would leave a necessary step undone. Austin treats failures of invocation or execution as misfires. The attempted act is void in the intended sense even if the attempt has consequences of its own. Austin, lecture II, pp. 15–18.

The remaining conditions, Γ.1 and Γ.2, apply when a procedure calls for particular thoughts, feelings, or intentions, or requires relevant behavior afterward. A person who undertakes to repay you with no intention of doing so can make an insincere promise. Someone can intend to repay you when they speak and then fail to do what they promised. Both differ from a case in which there was never a promise. Austin groups these kinds of infelicity as abuses. His labels help track which aspect of the performance is defective. Austin, lecture II, pp. 15–18 and lecture IV.

From Performatives to the General Theory of Speech Acts

The division between performatives and constatives becomes harder to sustain in the later lectures. Consider someone who says there is a cat on the mat. It is fair to ask whether the statement is true. It also matters whether the speaker is repeating a rumor, testifying as a witness, or joking. The practice of asserting has circumstances and standards of its own, just as promising does. The original distinction remains a useful way into the topic, but Austin treats it as a special case of a more general theory of speech acts. Austin, lectures XI–XII, especially pp. 133–148.

The later lectures distinguish three questions about a particular occasion of speaking. What was said? What did the speaker do in saying it? What did the speaker do by saying it? The answers can describe a single event from different angles. They do not require three independent performances.

The locutionary act concerns saying something with a certain sense and reference. Austin distinguishes the making of sounds, the speaking of words with a construction, and their use with a particular sense and reference. Someone can mimic the sounds of a sentence in a language they have never learned. When a person tells you the door is open and means the door to your left, the act includes the relevant use of words and their reference on that occasion. Austin, lecture VIII, pp. 94–99.

The illocutionary act is what the speaker does in saying it. A person could tell you the door is open to warn you about an escaped pet, to answer your question about whether an office is ready for visitors, or to ask you to close it. The sentence’s understandable content makes those uses possible, and the force comes from what is happening on the occasion. Austin gives a role to conventions, and considers how context and such features as mood and tone can help make the force clear. Austin, lectures VI and VIII, pp. 98–101.

The perlocutionary act concerns what someone brings about by speaking, such as persuading the listener to go back and close the door, or alarming them. An effect can exceed what the speaker intended. The listener can also recognize a warning and ignore it. That distinguishes recognizing what action the speaker was performing from being brought to act as the speaker hoped. Austin, lectures VIII–IX, pp. 101–103 and 116–119.

Locutionary, Illocutionary, Perlocutionary

The locutionary act is saying something with a certain sense and reference. The illocutionary act is what you do in saying it, such as warning someone. Its force depends on the conventions that make the warning recognizable. The perlocutionary act is what you bring about by speaking, such as frightening someone who hears the warning.

Austin calls the hearer’s understanding of the meaning and force uptake. Hearing and understanding a warning belongs to the warning’s successful delivery. Being frightened by it is a further possible effect. He also distinguishes the conventional effects an illocution may have, such as the name a ship now has, and the responses it may invite. Those are reasons to avoid a shortcut in which everything that happens after the utterance simply counts as perlocution. Austin, lecture IX, pp. 116–119.

Lecture XII tries out five families of illocutionary acts. Verdictives deliver an assessment, as when a judge issues a finding. Exercitives exercise authority or a right, as in appointing or voting. Commissives undertake commitments, as in promising. Behabitives include responses to someone’s conduct, such as apologizing or congratulating. Expositives help make clear what one is doing with an argument or conversation, such as conceding or giving an example. Austin treats the grouping as preliminary. A concession may acknowledge evidence and simultaneously commit you to a stance toward what has been argued. Austin, lecture XII, pp. 148–164.

Why Felicity Fails: Concrete Cases

Austin’s case of an unauthorized person trying to name a ship gives a concrete test of the categories. Someone could rush up to the launching party and use the right formula and even smash the bottle. The attempt may confuse the people there, interrupt the ceremony, or lead them to remove the intruder. Those consequences do not give the ship the name the intruder pronounced. Compare the person who has been asked to do the naming, goes through the appropriate procedure, and privately wishes the ship had received a different name. Austin’s distinction lets us ask separately about appointment, execution, and any attitudes the practice requires. Austin, lectures I–II, especially pp. 8, 15–18, and 23–24.

A teammate says, “We should think about moving the deadline.” Suppose the others leave believing the deadline has been changed. They may have mistaken a proposal for an announcement. The announcement and the proposal need different conditions. The proposal can invite a discussion among people who have no right to set the deadline. The announcement may require that the person already has that authority, or that an appropriate decision has taken place. Looking at both cases also separates the speaker’s force from the way the hearers were induced to plan their work.

Objections and Replies

The proposed families leave some cases in two places. An umpire declares a player out: the utterance assesses what happened and exercises a power within the game. Austin compares verdictives and exercitives because examples of this sort ask whether the speaker is giving a judgment, deciding what will be done, or both. A later classification can disagree with these divisions without giving up the value of studying force. The point of the exercise is as much to expose the mixed cases as to construct a list of clean ones. Austin, lecture XII, pp. 151–156.

Convention is harder to identify when the act has no ceremonial formula. Asking “Can you pass the salt?” can be a question about someone’s reach or an ordinary request to move the shaker. An analysis might appeal to what the hearer infers about the speaker’s intention. It might emphasize the familiar form of the request and the conditions under which that form is appropriate. The cases differ from naming a ship, where standing in a recognizable procedure is central. Austin gives a reason to examine both convention and the circumstances in which the hearer can find the force intelligible. Austin, lectures VI and IX.

Understanding the force of an utterance

Imagine finally that the warning about an open door reaches someone as a note slipped under a cup. The person can read the sentence successfully, recognize that the author is warning them, and decide to keep working. Under one description the communication worked, and under another the sender did not achieve the hoped-for result. A reply could ask what the danger is. It could challenge whether the warning applies to this particular door. Or it could acknowledge the risk and say it will be taken. The distinctions let us be precise about where a disagreement about the exchange lies.

The friend who promised repayment on Friday leaves us with several questions. Did it count as a promise on that occasion? Was it made in good faith? Did the friend do what they undertook? Some disputes about promises begin when a person answers one of those questions and the other person is concerned with a different one. Austin’s lectures show how much can be learned by finding out which action each person takes the words to have performed.

Further Reading

  • J. L. Austin, How to Do Things with Words. Read lectures I–II for the test cases and infelicities, VIII–IX for the three aspects of a speech act and the discussion of uptake, and XI–XII for the reassessment of the opening distinction and the tentative classification.
Join the discussionAsk a question or share what you learned.

Comments

Add to the conversation

A question, a correction, or an example from your own work.

No comments yet. Be the first to share your thoughts!

A different angle
All writing
Philosophy

Derrida's Of Grammatology: Writing, Trace, and the Supplement

Jacques Derrida's 1967 inquiry into the relation between speech and writing, with close readings of Saussure, Lévi-Strauss, and Rousseau.

13 min read
Philosophy

Derrida's Writing and Difference: Structure, Sign, and Play

Jacques Derrida's 1967 collection of essays, with a close reading of “Structure, Sign, and Play” and companion essays on Foucault, Freud, and Levinas.

13 min read
Philosophy

Kripke's Naming and Necessity: Rigid Designation and Essentialism

Saul Kripke’s lectures on how names refer, why some necessary truths are discovered empirically, and what people and things could have been.

13 min read

About the author

Reference

Citation details

Cite or share this article.

BIBTEXAcademic
@misc{brenndoerfer2025austins, author = {Michael Brenndoerfer}, title = {Austin's How to Do Things with Words: Speech Act Theory}, year = {2025}, url = {https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide}, organization = {mbrenndoerfer.com}, note = {Accessed: 2026-11-19} }
APAAcademic
Michael Brenndoerfer (2025). Austin's How to Do Things with Words: Speech Act Theory. Retrieved from https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide
MLAAcademic
Michael Brenndoerfer. "Austin's How to Do Things with Words: Speech Act Theory." 2026. Web. November 19, 2026. <https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide>.
CHICAGOAcademic
Michael Brenndoerfer. "Austin's How to Do Things with Words: Speech Act Theory." Accessed November 19, 2026. https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide.
HARVARDAcademic
Michael Brenndoerfer (2025) 'Austin's How to Do Things with Words: Speech Act Theory'. Available at: https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide (Accessed: November 19, 2026).
SimpleBasic
Michael Brenndoerfer (2025). Austin's How to Do Things with Words: Speech Act Theory. https://mbrenndoerfer.com/writing/austin-how-do-things-words-speech-act-theory-guide
Newsletter

Stay up to date

Get new articles, book updates, and notes on AI and technology.

Unsubscribe at any time. Privacy policy