Speech act theory, developed by Austin and Searle, recognizes that utterances do not merely describe the world but perform actions. Austin distinguished locutionary acts (what is literally said), illocutionary acts (what is done in saying it — promising, warning, asserting), and perlocutionary acts (the effect produced in the hearer). Searle classified illocutionary acts into five types: assertives, directives, commissives, expressives, and declarations. Indirect speech acts occur when the literal and intended illocutionary force diverge (e.g., 'Can you pass the salt?' as a request, not a question about ability).
Collect examples of each illocutionary type and analyze the felicity conditions required for each to succeed. Work through indirect speech acts by asking what contextual reasoning licenses the non-literal interpretation.
You already know from pragmatics that meaning in context goes beyond what sentences literally say. Speech act theory sharpens this insight into a formal account of the different things language does. When Austin observed that some utterances — "I promise," "I now pronounce you married," "I hereby sentence you" — are not descriptions of actions but the performance of actions, he identified a puzzle: the standard truth-conditional view of meaning cannot capture what these utterances do, because asking whether "I promise to call you" is true or false misses the point entirely. The utterance is a promise, not a claim.
Austin's solution was to distinguish three layers in every utterance. The locutionary act is the act of saying something with literal sense and reference — the phonological and semantic content. The illocutionary act is what is done *in* saying it: promising, warning, asserting, asking, apologizing. The perlocutionary act is the effect produced *by* saying it in the hearer: being persuaded, being frightened, being informed. These three levels are analytically distinct. A single utterance ("There's a loose wire behind you") may be a locutionary act with a propositional content, an illocutionary act of warning, and a perlocutionary act of alarming the hearer — but whether the alarm is felt depends on the hearer, not the speaker.
Searle systematized illocutionary acts into five categories based on their direction of fit and sincerity conditions. Assertives commit the speaker to the truth of a proposition (claiming, concluding). Directives attempt to get the hearer to act (requesting, commanding, asking). Commissives commit the speaker to a future course of action (promising, offering). Expressives express the speaker's psychological state (thanking, apologizing, congratulating). Declarations bring about the state of affairs they name, contingent on the speaker's institutional authority (sentencing, firing, marrying). This last category captures Austin's original insight about performatives: declarations work not because they describe a pre-existing reality but because the institutional context gives the utterance the power to create one.
Indirect speech acts arise when the illocutionary force of an utterance diverges from its literal form. "Can you pass the salt?" has the syntactic form of a yes/no question (a directive for information), but virtually no one hears it as a sincere inquiry into motor ability. Contextual reasoning — including Gricean maxims you studied in pragmatics — licenses the move from the literal question to the intended request. Indirect speech acts are not evasive or confusing; they are pervasive and efficient, allowing speakers to make requests politely, reduce face threat, and leave the hearer room to decline gracefully.
A key practical concept is felicity conditions — the requirements that must be met for an illocutionary act to succeed. A promise requires that the speaker sincerely intends to do the thing, that the thing is possible, that the hearer would prefer it to be done, and that it is not already going to happen regardless. If any condition is violated, the speech act is infelicitous — not false, but defective. Understanding felicity conditions is what lets you see why "I promise I'll be rude to you" sounds odd (it violates the preference condition) and why authority matters so much for declarations.