J.L. Austin and John Searle — Speech Acts

J.L. Austin and John Searle — Speech Acts

§SELF — thinker-austin-searle

What this page is: An explanation of speech act theory and its two principal architects. What it explains: How saying something is a form of doing something, and why this distinction matters for designing systems that process language. Why read it: To understand the difference between describing an action and performing one — a distinction with direct consequences for protocol design.

What Speech Act Theory Is

Speech act theory is the branch of philosophy of language that studies how utterances perform actions, not just describe the world.

J.L. Austin (1911–1960), a British philosopher, introduced the theory in a series of lectures at Harvard in 1955, published as How to Do Things with Words (1962). His core claim: language is not only for stating facts. Some sentences do not describe anything — they perform an action by being spoken. Example: "I promise to pay you" does not describe a promise; it makes one.

Why It Matters

If saying something is doing something, then any system that processes language must distinguish between descriptions (statements about the world) and performances (actions carried out through words). A protocol that treats a capability description as a command has a design flaw: it confuses what a user can do with what they are doing.

The Key Idea

A speech act has three components:

1. Locution — the act of saying something. This is the words themselves, with their literal meaning. Example: the string of sounds or text "I promise to pay you."

1. Illocution — what the speaker is doing in saying it. This is the force of the utterance: promising, ordering, warning, asking, naming. Example: the act of making a promise.

1. Perlocution — the effect the utterance has on the listener. This is the consequence: persuading, frightening, convincing, reassuring. Example: the listener now believes they will be paid.

Example in full: "I promise to pay you $50."

• Locution: the words "I promise to pay you $50"
• Illocution: the act of promising
• Perlocution: your confidence that you will receive $50

John Searle (born 1932) systematized Austin's work in Speech Acts: An Essay in the Philosophy of Language (1969). Searle added formal rules for what makes a speech act valid (felicity conditions) and classified illocutionary acts into five types: assertives, directives, commissives, expressives, and declarations.

What They Got Right

• Identified that meaning is not only in the words but in what the speaker intends to do with them.
• Created a framework that applies to any language use, including machine-processed language.
• Showed that context determines meaning: the same sentence can be a warning, a threat, or a promise depending on who says it and when.
• Searle's classification system allows systematic analysis of any utterance's function.

What They Got Wrong or Left Unfinished

• Austin died before fully developing the theory; his published work is reconstructed from lecture notes.
• Searle's classification is disputed: some philosophers argue the five categories overlap and cannot be cleanly separated.
• Neither addressed how speech acts function in non-human or machine-mediated communication. The theory was built for face-to-face conversation.
• The perlocutionary effect is inherently unpredictable: you cannot know what effect your words will have, which makes it hard to model formally.

How It Connects to Other Ideas

Protocol design. An OIP (Object Invocation Protocol) invocation is a speech act. The locution is the HTTP request (the bytes sent over the network). The illocution is the object invocation — what the request is doing (e.g., "I invoke the NOW object"). The perlocution is the receipt — the proof that the action was performed.

Capability description vs. invocation. Austin's key insight applies directly to the token drop problem. A capability description ("you can access object X") is a description — an assertive speech act. It states what is possible. An invocation request ("I invoke object X now") is a directive — it performs an action. The model must not confuse the two. A token that says what you can do is not the same as a request that says what you are doing.

Receipt as proof of illocution. In speech act terms, a receipt is proof that the illocution happened. It documents that the performance was executed, not just described.

What OIP should take from this. Explicit separation of capability description (what you CAN do — an assertive) from invocation request (what you ARE doing — a directive or commissive). The receipt proves the illocution occurred, not that it was possible.

Sources

• Austin, J.L. How to Do Things with Words. Harvard University Press, 1962.
• Searle, John R. Speech Acts: An Essay in the Philosophy of Language. Cambridge University Press, 1969.
• Searle, John R. Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge University Press, 1979.

---

Up the tree

• OIP root — protocol root, zero-context entry
• Thinker Reference hub — full hierarchy map
• Thinkers shelf — siblings on this shelf
• Voxel graph article — how pages link as voxels
• Self-describing protocol

Related on this shelf

• Alan Kay — The Big Idea Is Messaging
• Alfred North Whitehead — Process and Reality
• Barbara Liskov — Abstract Data Types and Distributed Consensus
• Bram Cohen — BitTorrent and Content-Addressed Protocol Design
• Butler Lampson — Protection and Access Control
• Carl Hewitt — The Actor Model
• Charles Sanders Peirce — Signs, Abduction, and Pragmatism
• Doug Engelbart — Augmenting Human Intellect

Machine surfaces

• Public page: https://miscsubjects.com/a/thinker-austin-searle
• JSON article: https://miscsubjects.com/api/articles/thinker-austin-searle
• OIP ask: https://miscsubjects.com/api/dispatch?ask=J.L.%20Austin%20and%20John%20Searle%20%E2%80%94%20Speech%20Acts