- Length
- 192 pagesUrmson and Sbisà, How to Do Things with Words, Harvard, 2nd ed. 1975
- Reading
- OpenWritten for a general reader and readable without preparation
- Availability
- In printA current standard edition is available
How to Do Things with Words
Twelve lectures noticing that saying something is often doing something, and then discovering that the distinction they were built on will not hold. Austin abandons his own framework halfway through and rebuilds, in public, which is why the book is both a founding text and a demonstration of how to think.
- Date
- 1962
- Attributed to
- J. L. Austin
- Region
- Oxford and Harvard
- Language
- English
Where it stands in the argument
- Position
- Reply — Answers a named earlier work
- Form
- Lectures
- Method
- Analytic — Clarification of language and logical form
- Transmission
- Posthumous — Published after the author died
- Standing
- Live — Still argued in current philosophical work
- Years back to what it answers
- 26 years
Reconstructed from lecture notes after its author's early death, and the origin of speech act theory, which is now used in linguistics, law and computer science.
What the book says
Twelve lectures that begin with a small observation and end by dismantling the framework they started in. Austin points out that philosophers have assumed the business of a sentence is to state something true or false, and that a great many perfectly ordinary sentences do nothing of the kind. Saying I do at a wedding, I name this ship, I bet you sixpence, I promise - these are not reports of an act, they are the act. He calls them performatives and spends the first half working out what can go wrong with them: not falsity but various kinds of failure, which he catalogues with evident enjoyment - the ship named by the wrong person, the promise made without intent, the bet nobody accepts. Then the argument turns. Trying to find a grammatical or lexical test that separates performatives from ordinary statements, he fails repeatedly, and concludes that the distinction cannot be drawn because stating is itself a thing done. The framework is replaced: every utterance has a locutionary aspect, what is said; an illocutionary one, what is done in saying it; and a perlocutionary one, what is brought about by saying it.
What it argues
That every utterance is an act, and that describing is one act among many rather than the standard case. The triple analysis is the result: in saying anything I perform a locutionary act, produce an utterance with sense and reference; I thereby perform an illocutionary act, which is what I am doing in saying it - warning, promising, ordering, stating; and I may bring about a perlocutionary effect, what I achieve by saying it, such as persuading or alarming. The middle term is the innovation. Illocutionary force is conventional, tied to circumstances and uptake, and it is what makes a form of words a warning rather than a prediction. Truth and falsity then become one dimension of assessment among several, appropriate to some acts and irrelevant to others.
What it answered
The descriptive fallacy in general and logical positivism in particular, and behind them the philosophical habit of treating an artificially regimented language as the model of a real one. Austin's whole approach - ordinary language philosophy as practised at Oxford - rests on the view that the distinctions embedded in everyday speech have survived because they mark differences that matter, and that a philosopher should look at them before inventing new ones.
How it argues
By example and by self-correction. Austin builds a distinction, tests it against dozens of ordinary sentences, finds it will not hold, and says so - the fifth and sixth lectures are a sustained record of a promising idea failing. This is rare in philosophy and it is why the book reads as thinking rather than as reporting. The tone is dry and comic, and the examples are so numerous that the argument sometimes disappears into them, which some readers regard as the method and others as its weakness.
The moment
Oxford in the 1950s, where Austin ran a Saturday morning discussion group that dominated the philosophical generation, and Harvard in 1955 for the James Lectures. He had spent the war in military intelligence, working on the invasion of Normandy, and was said to be brilliant at it. He died of cancer in 1960 at forty-eight, having published very little; almost everything under his name was assembled by others afterwards.
How it travelled
Out of philosophy into several disciplines. Searle systematised speech acts into a theory in 1969 and made them a standard part of philosophy of language; linguistics took them into pragmatics; legal theory uses them to analyse what a statute or a verdict does; computer science uses them in protocols for communicating agents. Derrida's reading in the 1970s, and the argument with Searle that followed, made Austin a reference point in continental theory as well. Feminist and critical race theory took up the notion of an utterance that does something in work on hate speech and on how a person is constituted by what is said to and about them.
The strongest objection
That the central distinction is never made precise enough to do the work asked of it. Austin himself could not find a criterion separating illocutionary from perlocutionary acts that survives his own examples, and the lectures end with a provisional list of five classes of illocutionary force which he calls a mess and offers with apologies. Searle's systematisation supplies the precision and does so by introducing intentions and rules Austin was careful to avoid, which some readers regard as a completion and others as a different theory wearing his vocabulary. The deeper complaint is that a method proceeding entirely by example cannot tell when the examples have run out.
Why read it
Because it is short, funny and the clearest demonstration in modern philosophy of a thinker changing his mind in front of the reader. The first two lectures set up the performative and the eighth introduces the triple analysis; the middle lectures, where the distinction collapses, are the ones worth reading slowly. No background needed.
- First printed
- 1962
- First in English
- 1962
- Digitised
- No
Assembled by J. O. Urmson from Austin's manuscript notes and the notes of those who heard the lectures; the second edition of 1975 revises the reconstruction. What survives is a record of a spoken argument rather than a book.
Written in English. In copyright.
Where it stands
Live and widely used. Speech act theory is standard equipment across several fields, and the specifically Austinian questions - whether illocutionary force is conventional, what uptake requires, how the categories divide - are still argued in philosophy of language.
On quoting this book
In copyright. Nothing is quoted.
Contested
What Austin would have published is unknowable, since the text is a reconstruction from notes by an editor and the lectures were spoken. Whether the final classification of illocutionary forces is meant seriously or as a placeholder is disputed, and he calls it unsatisfactory himself. Whether Searle's theory develops or replaces the position is argued. And the Derrida-Searle exchange of the 1970s, over whether Austin's exclusion of jokes and stage utterances is a legitimate simplification or a revealing evasion, generated a great deal of heat and is still cited by both traditions with opposite conclusions.