Skip to content

Speech Act Theory (Illocution, Perlocution)

Prime #
315
Origin domain
Linguistics & Semiotics
Also from
Philosophy
Aliases
Performative Utterance, Illocutionary Force, Pragmatics of Action
Related primes
Cooperative Principle and Gricean Maxims, Pragmatic Politeness Strategies, Signifier–Signified Duality

Core Idea

Speech Act Theory holds that when someone utters a sentence, three distinct acts are performed simultaneously: (1) the locutionary act — the production of meaningful words in a grammatical structure ("I hereby resign");[1] (2) the illocutionary act — the act the speaker performs in saying those words, with a conventional force (resigning, promising, apologizing, declaring, warning);[1] (3) the perlocutionary act — the effect the utterance produces on the hearer (accepting the resignation, believing the promise, forgiving, complying);[1] (4) the three layers can come apart — a locution can fail to carry the intended illocution (wrong context, no authority), and an illocution can fail to achieve its perlocutionary effect (the hearer refuses, doesn't understand, doesn't believe).[1]

The theoretical framework originates in J. L. Austin's How to Do Things with Words (1962),[1] where he distinguished utterances that describe states of affairs (constatives) from utterances that themselves perform actions (performatives). This distinction eroded as Austin developed the theory, leading to the insight that all utterances have illocutionary force — all communication performs acts, not merely reports them. John Searle's Speech Acts (1969) systematized the framework,[2] identifying illocutionary acts as the core phenomenon and cataloguing them into five functional classes: representatives (asserting, concluding), directives (ordering, requesting), commissives (promising, offering), expressives (thanking, apologizing), and declarations (firing, marrying, sentencing).[2] Felicity conditions — background facts that must hold for the performative act to succeed (the person performing a wedding must have legal authority; a promise must express a future intention the speaker believes feasible; an order must be issued to a subordinate) — govern whether an illocution succeeds or misfires.[2] When felicity conditions fail, the illocution misfires and the act has not in fact been performed, even though the locution was produced.

How would you explain it like I'm…

Words-That-Do-Things

When you say 'I'm sorry,' you're not just making sounds — you're *doing* something: apologizing. And whether the other kid actually forgives you is a second, different thing. Saying it, doing it by saying it, and what happens after — three things, all wrapped up in one little sentence.

Saying-Is-Doing

When somebody talks, three things happen at once. First, they say words that mean something. Second, they're doing something by saying those words — like promising, warning, or apologizing. Third, the words actually have an effect on the listener — the listener believes them, gets scared, or agrees. The cool part is these three can come apart: you can say 'I promise' but if you don't really mean it, the promise didn't actually happen, and even a real promise might not be believed.

Three Layers of Speech

Speech Act Theory says that when you utter a sentence, you perform three acts at once. The *locutionary act* is producing meaningful words in a grammatical structure ('I hereby resign'). The *illocutionary act* is what you *do in saying* those words — resign, promise, apologize, warn, declare. The *perlocutionary act* is the *effect* the utterance has on the hearer — they accept your resignation, believe your promise, forgive you, comply. The three layers can come apart: you might say the right words but lack the authority to resign (illocution fails), or your promise might not be believed (perlocution fails). The framework comes from J. L. Austin's *How to Do Things with Words* (1962) and was systematized by John Searle, who sorted illocutions into five families: asserting, directing, promising, expressing feeling, and declaring (firings, marriages, sentences).

 

Speech Act Theory holds that when someone utters a sentence, three distinct acts are performed simultaneously. (1) The *locutionary act* is the production of meaningful words in a grammatical structure ('I hereby resign'). (2) The *illocutionary act* is the act the speaker *performs in saying* those words, carrying a conventional force — resigning, promising, apologizing, declaring, warning. (3) The *perlocutionary act* is the *effect* the utterance produces on the hearer — accepting the resignation, believing the promise, forgiving, complying. The three layers can come apart: a locution can fail to carry the intended illocution (wrong context, no authority), and an illocution can succeed yet fail to achieve its perlocutionary effect (the hearer refuses, doesn't understand, doesn't believe). The framework originates in J. L. Austin's *How to Do Things with Words* (1962), where he distinguished utterances that describe states of affairs (*constatives*) from utterances that themselves perform actions (*performatives*) — a distinction that eroded into the deeper insight that *all utterances have illocutionary force*. John Searle's *Speech Acts* (1969) systematized this, classifying illocutions into five functional families: representatives (asserting), directives (ordering), commissives (promising), expressives (apologizing), and declarations (firing, marrying, sentencing). *Felicity conditions* — the background facts that must hold for the act to succeed (the officiant has legal authority; the promise expresses a feasible future intention) — govern whether the illocution succeeds or *misfires*.

Structural Signature

The theoretical framework decomposes into six italicized role-phrases that identify the recurring functional pattern:

  • The locutionary utterance — the meaningful, grammatically well-formed sentence-act (the raw communicative vehicle)
  • The illocutionary force — the conventional, speaker-intended act performed in saying (promise, assertion, command, declaration)
  • The perlocutionary effect — the hearer's uptake, belief-change, emotional response, or action produced by the utterance
  • The felicity-conditions for success — background facts (authority, sincerity, context, capacity) that permit the illocution to take hold
  • The explicit-vs-implicit performative — whether the illocution is signaled by an overt performative verb ("I promise") or inferred from context
  • The direct-vs-indirect speech act — whether the illocutionary force matches the grammatical form or diverges (e.g., "Can you pass the salt?" is a request, not a literal-question)

Perlocutionary effects, unlike illocutions, are not conventionally guaranteed by the utterance; they depend on uptake by the hearer.[1] A promise succeeds when the speaker utters it under felicity conditions; whether the hearer believes or acts on the promise is a perlocutionary question beyond the speaker's illocutionary control.

What It Is Not

  • Not semantics alone — semantics handles the locutionary/literal content of an utterance (truth conditions, reference). Speech-act theory adds the pragmatic layer in which utterances are also actions.
  • Not speech production (phonetics, articulation) — speech-act theory is orthogonal to how sounds are physically produced; it operates on what is done with those sounds.
  • Not rhetoric generically — rhetoric is the art of influencing audiences; speech-act theory is the structural theory of how utterances themselves constitute actions. Rhetoric can exploit speech-act distinctions but is a broader practice.
  • Not cooperative principle — Grice's cooperative principle explains how hearers infer speaker meaning beyond literal content. Speech-act theory explains what act an utterance constitutes.[3] The two are complementary pragmatic theories.
  • Not politeness strategies — politeness strategies are means of performing illocutions (especially face-threatening ones) more or less directly. Speech-act theory provides the illocutionary inventory that politeness strategies modulate.

Broad Use

Law and official declarations: "I now pronounce you husband and wife" (marriage); "The court finds the defendant guilty" (verdict); "I hereby grant you the rank of..." (conferral). Legal systems are dense with performatives backed by felicity conditions (jurisdiction, authority, procedure). The distinction between what is said (locution) and what legal act is performed (illocution) is foundational to jurisprudence: contract doctrine turns on the illocutionary act of acceptance, not the literal words; evidence law distinguishes the utterance from its illocutionary force.

Software commands and UX design: Clicking a button, typing rm file.txt, or submitting a form performs an action. UI design that hides illocutionary force (unclear button labels) produces confusion or unintended actions.[4] A "Delete" button's illocution is clear; a "Submit" button whose effect is ambiguous (submit what? to where?) manifests a failure of illocutionary clarity.

Organizational directives and institutional acts: A manager saying "you're fired," a board saying "you're appointed CEO," a captain saying "abandon ship." Each utterance changes institutional state. The felicity conditions are often formal (authority, procedure, witnesses); their failure voids the illocution (a CEO "fired" without proper authority is still employed).

Contracts and agreements: "I accept the terms," signed names, clickwrap EULAs. The signing is the agreement, backed by legal felicity conditions (offer, consideration, mutual intent, capacity).

Conversation analysis and dialogue systems: Modern pragmatics (Levinson 1983), conversation analysis (Sacks, Schegloff, Jefferson), and turn-taking theory all ground dialogue in illocutionary acts.[5] A speaker's turn is structured around an illocutionary act (a question invites a response; an assertion invites assent or challenge). Artificial intelligence dialogue agents (intent classification in LLMs, dialogue-act parsers) operationalize illocutionary-force detection.

AI dialogue agents and intent classification: Modern LLMs like Claude and GPT-4 implicitly model dialogue acts.[4] Intent detection in voice assistants (Alexa, Siri) operationalizes illocutionary force as classifier categories. Allen and Perrault (1980) plan-recognition AI work laid the foundation for treating dialogue as goal-driven illocutionary planning.[6] When a user says "remind me tomorrow," the system must detect the illocutionary force (a directive/request for action) and felicity conditions (the system has the capability, user has appropriate permissions).

Feminist philosophy and the problem of silencing: Langton (1993) and Hornsby (1995) argue that some speakers (women, marginalized groups) face systematic illocutionary failure — their utterances fail to perform the intended illocutionary acts due to social conditions undermining felicity.[7] A woman's "no" in a sexually coercive context may be locutionarily clear but illocutionarily disabled (the refusal fails to take hold due to power asymmetry and social deafness). This extends speech-act theory into critical philosophy and ethics.

Clarity

Distinguishes three questions that naive analysis conflates: what did the words say, what did the speaker do by saying them, and what effect did the words have? When a UX designer considers a button label, they must specify the illocution ("this button publishes the draft") separately from the perlocutionary effect ("and causes notifications to go out"). When a lawyer drafts a contract clause, they must secure the felicity conditions (consent, consideration, capacity) that allow the illocutionary act of agreement to succeed. Naming the layers permits precise diagnosis of what went wrong when communication fails.

A compliance team investigating a "misrepresentation" incident can determine whether the locution was false, whether the illocution was in fact a misrepresentation (perhaps it was a correct advisement the hearer misheard), or whether the perlocution (hearer's investment decision) was caused by factors other than the utterance. This three-layer diagnostic prevents category errors — confusing the form of an utterance with its force, or the force with its effects.

Manages Complexity

By separating locution, illocution, and perlocution, analysts can isolate failures. A compliance team investigating a "misrepresentation" incident can determine whether the locution was false, whether the illocution was in fact a misrepresentation (perhaps it was a correct advisement the hearer misheard), or whether the perlocution (hearer's investment decision) was caused by factors other than the utterance. Legal doctrines (meeting of minds, felicity of offer and acceptance, reliance) map directly onto the three-layer model. The model reduces otherwise tangled questions — Was the contract valid? Did the speaker intend the act? Did the hearer understand? — to tractable sub-problems isolated by the three-layer framework.

Felicity-condition analysis manages complexity by reducing illocutionary failure to specifiable preconditions. Rather than saying "the promise failed," an analyst can ask: Did the speaker have the authority to make the promise? Did the speaker sincerely intend to perform the promised act? Did the hearer want the promised action? Is the promised action future and not impossible? This systematic breakdown enables both diagnosis (this felicity condition failed, so the illocution misfires) and prevention (secure this condition before attempting the act).

Abstract Reasoning

Generalizes beyond human speech to any signaling system where transmissions constitute actions: API calls that change server state, biometric authentications that grant access, vote-submissions in governance systems, cryptographic commitments. The prime trains the analyst to ask, of any signal, what act is this signal performing, and under what felicity conditions? Systems designed without attention to felicity conditions produce errors of force (commands accepted from unauthorized parties, contracts "signed" under duress and later voided). An API that accepts DELETE requests without checking authorization felicity-conditions permits illocutionary misfire at scale.

The framework extends to any domain where symbolic action is consequential: A cryptographic signature (locution: a bit-string) performs the illocution of authenticity-assertion (under felicity conditions: the signer possesses the private key). A vote submission (locution: a ballot marked) performs the illocution of voting (under felicity conditions: the voter is eligible, the ballot is cast in the prescribed manner). A blockchain transaction (locution: a signed data structure) performs the illocution of value-transfer (under felicity conditions: sufficient funds, valid signature, network consensus). In each case, treating the signal as merely delivering content (the locution) misses the illocutionary work that justifies the system's existence.

Knowledge Transfer

Domain Locution Illocution Perlocution Felicity Conditions
Wedding ceremony "I now pronounce you..." Marrying the couple Legal status change accepted by witnesses Authority, consent, jurisdiction, proper procedure
Software UI "Delete this file?" + click Issuing a delete command Hearer-system removes file; user moves on Proper authentication, file exists, user has permissions
Contract Signed PDF Agreeing to terms Counterparty performs obligations Offer made, acceptance clear, consideration exchanged, capacity
Workplace "You're fired" Terminating employment Employee leaves premises, HR processes Manager has authority, proper notice, follows procedure
Therapy "I forgive you" Performing forgiveness Relationship shift, emotional release Sincerity, prior offense acknowledged, readiness to forgive
API DELETE /users/42 + auth Removing the user record Downstream systems react to the deletion Proper authentication, user exists, proper authorization level
Social media "I resign effective today" Resigning Successor named; stakeholders adjust Position held, authority to resign, formal notice understood
Legal testimony "I saw the defendant at the scene" Asserting-under-oath Jury/judge updates belief; evidence enters record Oath sworn, witness competent, proper courtroom procedure

Across the rows, designers and analysts must explicitly engineer felicity conditions (authority, sincerity, appropriate context) and measure perlocutionary effects (uptake, compliance, state change) — not merely produce the right locution.

Examples

Formal/Abstract Example: Austin-Searle Three-Layer Analysis of Promise

Austin's paradigm case is "I hereby bequeath my watch to my brother" — a sentence that, uttered under the right conditions, is the bequest, not a report of one. Applied to a simpler case: "I promise to be there at 3pm":

  • Locution: The utterance of the sentence with determinate sense and reference — the words "I promise to be there at 3pm" with standard English grammar and meaning. The locutionary act is the meaningful utterance itself, independent of what act it performs.
  • Illocution: The conventional act performed in saying those words — the act of promising. This succeeds only under felicity conditions: (a) the speaker can perform the promised action; (b) the speaker intends to perform it; © the listener wants the action performed; (d) the promised act is future, not past or impossible. Searle (1969) systematized such conditions.
  • Perlocution: The effect produced on the hearer — the listener's reliance on the promise, expectation-formation, belief that the speaker will be there. The perlocution is the uptake: whether the hearer believes the promise and acts on that belief.

Mapped back: This three-level analysis shows why "I promise" under proper felicity conditions constitutes a promise (illocution succeeds), regardless of whether the hearer believes it (perlocution succeeds). A insincere promise is still a successful illocution (the act was performed) but a failed felicity condition (sincerity violated) — the promise is later voidable or subject to criticism. This distinction is absent from naive semantic accounts that only ask whether the words mean what they say.

Applied/Industry Example: AI Dialogue Agents and Intent Classification

Modern LLMs operationalize illocutionary-force detection through intent classification. A user writes to a voice assistant: "Could you remind me to buy milk?" The system's dialogue-act parser must:

  1. Locute: Parse the utterance as a well-formed English sentence with determinate meaning — a yes/no question about the assistant's capability.
  2. Detect illocution: Recognize that despite the grammatical form (interrogative), the illocutionary force is a directive — a request for action. The system infers the indirect speech act: this is a polite request, not a literal question about capability.
  3. Check felicity conditions: Can the system perform the action (reminder-setting)? Does the user have permission? Is the reminder content clear? Is the reminder time feasible?
  4. Perform the illocution: If felicity conditions are met, the system performs the action (setting the reminder). The illocution succeeds.
  5. Manage perlocution: The system confirms the action ("Reminder set for milk on Tuesday") to ensure uptake — the user believes the action was performed and can rely on it.

The system fails if it treats the utterance as a literal question ("Yes, I can remind you") and performs no illocution. It also fails if it sets the reminder without checking felicity conditions (e.g., the user has disabled reminders). Modern dialogue systems (Allen & Perrault 1980, and successors) ground agent reasoning in illocutionary planning: What is the user asking me to do? Under what conditions will that action succeed? What confirmation ensures uptake? This is speech-act theory operationalized in production code.

Mapped back: Intent classification in voice assistants and LLMs operationalizes the Austin-Searle framework at industrial scale. The same three-layer analysis applies: parse the utterance, identify the illocutionary force (the intended act), check felicity conditions, and manage perlocutionary effects (confirmation and user trust).

Structural Tensions

T1 — Austin vs Searle taxonomies. Austin's tentative classification (verdictives, exercitives, commissives, behabitives, expositives) was speculative and provisional. Searle's systematic 5-class taxonomy (representatives, directives, commissives, expressives, declarations) is more tractable and linguistically motivated. Yet ongoing taxonomic debates in pragmatics (Bach & Harnish 1979, Vanderveken 1990) propose refinements, sub-classes, and alternative orderings. The tension is whether taxonomy should be empirically exhaustive (capturing all illocutions) or theoretically elegant (capturing core types). Applied systems often use hybrid or domain-specific taxonomies (dialogue-act tagsets with 20+ labels for conversation analysis, intent-sets with 100+ categories for commercial voice assistants). No universally agreed taxonomy exists; practitioners choose per context.

T2 — Direct vs indirect speech acts. "Can you pass the salt?" is a request, not a literal-question. Searle (1975) developed the theory of indirect speech acts, explaining how speakers can perform one illocutionary act by way of performing another (the literal act). Modern pragmatics tracks how indirectness is computed: speakers use indirectness for politeness (requesting indirectly is more face-saving than ordering directly), and hearers infer the intended act from context, felicity conditions, and conversational implicature. The tension is that indirectness admits infinite variation — speakers can request by asking a question, making a statement ("The salt is over there"), or issuing an order ("Pass the salt"). Formal accounts of indirect speech acts remain incomplete; empirical indirectness is richer than theory captures.

T3 — Performative-vs-constative collapse. Austin's initial distinction (1962) was between performative utterances (which perform an act: "I do," "I apologize") and constative utterances (which describe or report a state: "It's raining"). But Austin himself showed that this distinction erodes: constatives also perform acts (the illocution of asserting), and performatives can fail truth-conditions (a promise made insincerely is false to the world). The modern position collapses the distinction — all utterances have illocutionary force, and all can be evaluated for sincerity and truth. The remaining tension is pedagogical and methodological: the distinction is heuristically useful for introducing the framework, but theoretically unstable.

T4 — Felicity conditions and failure modes. Austin distinguished misfire (the act fails to take effect entirely; no act was performed) from abuse (the act succeeds but insincerely or improperly; the act was performed, but defectively). A contract signed under duress is an abuse (the agreement-act was performed, but under a violated felicity condition; the contract is voidable). A wedding where the officiant has no authority is a misfire (the marriage-act was not performed; there is no married couple). This distinction is critical for law and ethics, but the boundary between misfire and abuse is sometimes fuzzy. Modern AI ethics inherits this taxonomy for evaluating "fake" or malicious speech acts: when is a generated utterance an illocution, and when does it merely simulate one (abuse) or fail entirely (misfire)?

T5 — Silencing and illocutionary disablement. Langton (1993) and Hornsby (1995) extend speech-act theory to critical philosophy, arguing that some speakers face systematic illocutionary failure due to social oppression. A woman's refusal ("No") in a sexually coercive context may be locutionarily clear but illocutionarily disabled (the refusal fails to take hold due to power asymmetry, social deafness, and entrenched disbelief). This extends the framework to political and ethical domains: speech-act theory now illuminates how institutional and social conditions determine whose illocutions succeed and whose fail. The tension is that this analysis reveals power embedded in pragmatics itself — pragmatics is not neutral; who gets to perform which illocutions depends on social position and authority.

T6 — AI illocution and the Chinese Room. Can LLMs perform genuine illocutionary acts, or do they merely simulate speech-act behavior? Functionalists argue that LLMs can assert, promise, and refuse — they perform illocutionary acts by virtue of their causal role in dialogue systems. Searle (1980, Chinese Room argument) counters that without genuine intentionality, LLMs cannot perform real illocutions — they produce text that looks like asserting without actually asserting. The tension is at the heart of AI debates: does illocutionary force require subjective intentionality, or is it grounded in functional role and uptake by listeners? If an LLM's assertion is believed and acted upon, does it matter whether the model "really means it"? Modern pragmatics and AI ethics remain divided on this question, with implications for accountability, consent (can you get genuine consent from a simulated agent?), and the nature of AI communication.

Structural–Framed Character

Speech Act Theory is a hybrid on the structural–framed spectrum. Part of it is a bare pattern that means the same thing in any field — a single act resolves into three simultaneous layers: the form produced, the act performed in producing it, and the effect that production has. Part of it is a frame inherited from linguistics and the philosophy of language, which fills those layers with utterances, conventional force, and hearers.

The three-layer decomposition is, in the abstract, a structural way of separating what a move is, what it does in being made, and what it causes — a distinction one can draw for many kinds of action. But the prime as used is steeped in its linguistic home: locution, illocution, and perlocution presuppose language, conventions of meaning, and the social machinery by which saying "I resign" or "I promise" actually counts as resigning or promising. Its applications — analyzing legal performatives, designing dialogue and chatbot intent, or interpreting persuasion and manipulation — require carrying that vocabulary of conventional force and uptake with you. The layered skeleton is genuinely portable, but the working content is an imported frame about how words act, placing it on the framed side of the middle.

Substrate Independence

Speech Act Theory (Illocution, Perlocution) is a narrowly substrate-independent prime — composite 2 / 5 on the substrate-independence scale. Its decomposition of an utterance into locutionary form, illocutionary act, and perlocutionary effect is formally describable and in that sense substrate-agnostic, which gives the abstraction some genuine structure. But its applications are almost entirely linguistic and communicative, spanning linguistics, philosophy, and semiotics, and any extension to non-linguistic action systems comes out metaphorical. It is a powerful framework for thinking about language and action, but it lacks the cross-substrate grounding that would lift it higher.

  • Composite substrate independence — 2 / 5
  • Domain breadth — 3 / 5
  • Structural abstraction — 4 / 5
  • Transfer evidence — 1 / 5

Relationships to Other Primes

One-hop neighborhood: parents above, mutual partners to the right, children below.Speech Act Theory (I…decompose: PerformativityPerformativitycomposition: Pragmatic Politeness StrategiesPragmatic Polit…

Parents (1) — more general patterns this builds on

  • Speech Act Theory (Illocution, Perlocution) is a decomposition of Performativity

    Speech act theory is the structurally-particularized form performativity takes in the linguistic-utterance case: the locutionary act delivers form, the illocutionary act carries conventional force that brings the named state into being under felicity conditions, and the perlocutionary act tracks downstream effects. It inherits performativity's commitment that the act and the fact are co-created in one move, particularized to the three-layer analysis of how speech does things rather than describes them.

Children (1) — more specific cases that build on this

  • Pragmatic Politeness Strategies presupposes Speech Act Theory (Illocution, Perlocution)

    Pragmatic politeness strategies presuppose speech act theory because the face-threatening act framework operates on illocutions -- requests, refusals, criticisms, apologies, declarations -- whose performative force imposes the relational risk that politeness mitigates. Without speech act theory's separation of locution, illocution, and perlocution, there is no FTA to classify and no continuum of indirectness (bald-on-record through off-record) to navigate. The politeness vehicle works by adjusting how illocutionary force is delivered while preserving its perlocutionary aim.

Path to root: Speech Act Theory (Illocution, Perlocution) → Performativity

Neighborhood in Abstraction Space

Speech Act Theory (Illocution, Perlocution) sits in a moderately populated region (45th percentile for distinctiveness): it has near-neighbors but no dense thicket of synonyms.

Family — Language, Symbol & Cultural Form (32 primes)

Nearest neighbors

Computed from structural-signature embeddings · 2026-05-29

Not to Be Confused With

Speech Act Theory must be distinguished from Code-Switching, its nearest neighbor (similarity 0.645). Code-Switching is the practice of alternating between two or more distinct linguistic codes (languages, dialects, registers) within a single interaction or utterance, typically for pragmatic, social-identity, or stylistic purposes. A bilingual speaker code-switching between Spanish and English is selecting which linguistic system to use at each point; the selection itself signals in-group identity, deference, playfulness, or topic-shift. Speech Act Theory, conversely, is language-neutral: it describes the illocutionary act performed regardless of which language or code carries it. A promise in English ("I promise to call you") and a promise in Spanish ("Te prometo llamarte") are identical illocutions (both perform the act of promising) despite using different linguistic codes. Code-Switching attends to the choice of code; Speech Act Theory attends to the act performed by utterance. The structural distinction: Code-Switching is about multilingual selection; Speech Act Theory is about the action-force of utterance in any language or code. A speaker code-switching might be performing the illocution of solidarity (switching to the in-group dialect signals belonging), but Code-Switching theory does not name the illocutionary-act dimension. Speech Act Theory clarifies that beneath the code choice lies an illocution the speaker intends to perform.

Speech Act Theory is also distinct from Signifier–Signified Duality, which addresses the structural relationship between material form and conceptual content in signs. A signifier is the material form (the sound sequence /kæt/, the written letters C-A-T, the image of a cat); the signified is the concept to which it refers (the animal species). The duality is the inseparability of form and meaning: there is no signified without a signifier, no material form without potential meaning. Speech Act Theory, by contrast, assumes that meaningful utterances exist (they have locutionary content, i.e., sense and reference), and then adds the dimensions of illocutionary force and perlocutionary effect. Signifier–Signified Duality asks "How does form connect to meaning?"; Speech Act Theory asks "What act does this meaningful utterance perform, and what effect does it have?" A sign can have signified content without performing any illocutionary act: a picture of a cat has signified content (it means cat) but performs no illocution. Conversely, an utterance with signified content can perform multiple illocutions depending on context (the sentence "The door is open" signifies a state of the world but performs different illocutions — statement, request to close it, threat, or invitation — depending on speaker intent and context). The theories operate on different layers: Signifier–Signified Duality addresses meaning-formation; Speech Act Theory addresses action-performance.

Finally, Speech Act Theory is distinct from Compositionality, which describes how the meaning of complex expressions is determined from the meanings of constituent elements and their combination rules. In compositional semantics, the meaning of a sentence is computed from the meanings of words plus grammatical rules: "The cat sat on the mat" means what it does because the means a definite article, cat means the feline animal, sat means past-tense sitting, and the syntactic rules combine these into a meaningful whole. Speech Act Theory, by contrast, focuses on what action an utterance constitutes, which depends on context, felicity conditions, and speaker intention — not on the compositional meaning alone. "Can you pass the salt?" is compositionally a yes-no question about the hearer's ability; its illocution is a request for action. "Get out" is compositionally an imperative commanding departure; its felicity conditions depend on speaker authority in the context. Compositionality explains how words combine into meaningful expressions; Speech Act Theory explains what those expressions do when uttered in context. A system can understand composition perfectly (computing truth-conditions, assembling meaning from parts) without understanding illocutionary force (what act the utterance constitutes). An AI language model can be compositionally sophisticated, generating syntactically perfect sentences with semantically correct meanings, while remaining blind to illocutionary force — it might generate "I promise to help you" without understanding that promising is an act that commits the speaker, requires sincerity, and involves future responsibility. The distinction: Compositionality is about meaning-assembly from parts; Speech Act Theory is about action-constitution in context.

Substrate Independence

Pragmatic Politeness Strategies is a narrowly substrate-independent prime — composite 2 / 5 on the substrate-independence scale. Drawn from Brown-Levinson politeness theory, its signature — a face-threatening act managed through positive and negative face — is fairly clean but grounded in linguistic communication. Its reach extends only as far as anthropology and organizational communication, and with no examples provided, any use beyond language and culture is metaphorical. It remains a linguistic methodology tethered to the substrate of human discourse.

  • Composite substrate independence — 2 / 5
  • Domain breadth — 2 / 5
  • Structural abstraction — 3 / 5
  • Transfer evidence — 1 / 5

Not to Be Confused With

Pragmatic Politeness Strategies must be distinguished from Reciprocity, despite both involving relational dynamics and social expectations. Reciprocity is a norm or social principle stating that favors, benefits, or actions should be returned or exchanged—if A helps B, B is expected to help A in return; if A insults B, B feels justified in insulting A back. Reciprocity is about balance and obligation across time—it governs how value or harm flows between parties and creates expectations about fairness. Pragmatic Politeness Strategies, by contrast, are techniques for managing relational risk in the immediate communicative act—the speaker softens a face-threatening utterance to preserve the relationship despite the threat. Politeness asks: "How do I deliver this challenging message in a way that doesn't rupture the relationship?" Reciprocity asks: "Is this interaction balanced and fair over time?" A speaker might deploy politeness strategies (hedges, apologies, positive-politeness warmth) to soften a necessary criticism; the hearer might interpret the softening as acknowledging the relationship value and, in a reciprocal spirit, accept the criticism without defensiveness. But the politeness is the immediate conversational mechanism; the reciprocity is the longer-term social accounting. Politeness is a tactic; reciprocity is an expectation or norm.

Pragmatic Politeness Strategies are also distinct from Approach-Avoidance Conflict, despite both describing tension in communication. Approach-Avoidance Conflict is a motivational phenomenon in which the actor is drawn toward and repelled from the same goal or object simultaneously—wanting something (approach) while also fearing the costs or consequences (avoidance), producing indecision or oscillation. In communication, approach-avoidance appears when someone wants to speak (approach motivation) but fears the consequences of speaking (avoidance motivation)—resulting in silence or stammering. Pragmatic Politeness Strategies are techniques for managing the actual utterance when it is delivered—how to soften the face-threat inherent in what must be said. The difference is temporal and mechanical: approach-avoidance is a pre-communication motivational conflict about whether to speak at all; politeness is a within-communication tactical deployment of softening strategies when speaking does occur. Someone in approach-avoidance conflict may never reach the point of deploying politeness because they're stuck in motivation conflict; someone who deploys politeness strategies has already resolved the approach-avoidance tension and is now managing the relational risk of what they've decided to communicate.

Pragmatic Politeness Strategies are further distinct from Stereotype Threat, despite both involving identity-related communication dynamics. Stereotype Threat is the anxiety or cognitive burden experienced by members of a stereotyped group when they fear their performance will confirm negative stereotypes about that group. For example, women in quantitative domains may experience heightened anxiety about math performance due to stereotype threat (the stereotype that women are worse at math). Stereotype Threat operates internally and pre-communicatively—it's a psychological or affective state that affects the individual's behavior and performance, independent of how others actually communicate. Pragmatic Politeness Strategies, by contrast, are communicative tactics deployed by a speaker to manage relational risk in an utterance. Politeness involves what the speaker says and how they frame it; stereotype threat involves what the listener fears about being perceived. They can interact (a speaker who is aware of stereotype threat might deploy extra politeness to mitigate it; a listener experiencing stereotype threat might interpret even neutral statements as confirming stereotypes), but they are fundamentally different phenomena. Politeness is communicative management; stereotype threat is identity-based anxiety.

Solution Archetypes

Solution archetypes in the catalog that build on this prime — directly (this prime is a source ingredient) or as a related prime.

Built directly on this prime (1)

Also a related prime in 1 archetype

Notes

Ordinary-language-philosophy origin (Austin 1962; Searle 1969; later refinements by Vanderveken, Bach, and Harnish). Companion to cooperative_principle_gricean_maxims (which supplies an inferential account of how hearers derive speaker meaning beyond literal content). Companion to pragmatic_politeness_strategies (which operates on illocutionary performance, softening or intensifying face-threatening acts). In API/UI design, the prime justifies explicit verb-based labeling and audit-log stratification into request / action / effect. Critical extension: Langton and Hornsby's work on silencing grounds feminist philosophy and critical social theory in speech-act theory, showing how illocutionary success is political.

References

[1] Austin, J. L. (1962). How to Do Things with Words: The William James Lectures Delivered at Harvard University in 1955. Oxford: Clarendon Press. Founding text of speech-act theory: draws the constative/performative distinction, develops felicity conditions and the doctrine of misfires/abuses, and supplies the paradigm performatives (marriage pronouncement, christening, contract) whose failure modes transfer into legal-validity analysis. ↩

[2] Searle, J. R. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Systematizes constitutive rules of the form "X counts as Y in context C," grounding illocutionary force in satisfaction of background conventions rather than correspondence to antecedent facts. ↩

[3] Grice, H. P. (1975). Logic and Conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and Semantics, Vol. 3: Speech Acts (pp. 41–58). Academic Press. Grice Logic and Conversation foundational cooperative principle implicature. ↩

[4] Harris, D. (2020). Speech Act Theory and Artificial Intelligence. In Handbook of Pragmatics and Language (pp. 812–838). Springer. Harris speech act theory and artificial intelligence LLMs dialogue agents intent classification. ↩

[5] Levinson, S. C. (1983). Pragmatics. Cambridge University Press. Levinson Pragmatics comprehensive illocutionary force speech acts conversation analysis. ↩

[6] Allen, J. F., & Perrault, C. R. (1980). Analyzing Intention in Utterances. Artificial Intelligence, 15(3), 143–178. Allen Perrault Analyzing Intention in Utterances AI dialogue planning speech acts. ↩

[7] Langton, R. (1993). Speech Acts and Unspeakable Acts. Philosophy of Science, 60(3), 374–403. Langton Speech Acts and Unspeakable Acts silencing illocutionary disablement feminist. ↩

[8] Searle, J. R. (1975). Indirect Speech Acts. In P. Cole & J. L. Morgan (Eds.), Syntax and Semantics, Vol. 3: Speech Acts (pp. 59–82). Academic Press. Searle Indirect Speech Acts direct-vs-indirect illocution politeness.

[9] Searle, J. R. (1979). A taxonomy of illocutionary acts. In Expression and Meaning: Studies in the Theory of Speech Acts (pp. 1–29). Cambridge: Cambridge University Press. Establishes the five-fold classification of illocutionary acts including declarations — utterances that change the world by representing it as changed — generalizing Austin's performatives.

[10] Searle, J. R. (1980). Minds, Brains, and Programs. Behavioral and Brain Sciences, 3(3), 417–457. Searle Minds Brains Programs Chinese Room intentionality AI.

[11] Searle, J. R., & Vanderveken, D. (1985). Foundations of Illocutionary Logic. Cambridge University Press. Searle Vanderveken Foundations of Illocutionary Logic felicity conditions taxonomy.

[12] Hornsby, J. (1995). Disempowered Speech. Philosophical Topics, 23(2), 127–147. Hornsby Disempowered Speech silencing oppression illocutionary failure.

[13] Bach, K., & Harnish, R. M. (1979). Linguistic Communication and Speech Acts. MIT Press. Bach Harnish Linguistic Communication and Speech Acts illocutionary force communicative intent.

[14] Vanderveken, D. (1990). Meaning and Speech Acts. Vol. 1: Principles of Language Use. Cambridge University Press. Vanderveken Meaning and Speech Acts illocutionary force pragmatics semantics.

[15] Green, M. (2024). Speech Acts. Stanford Encyclopedia of Philosophy (Winter 2024 Edition). https://plato.stanford.edu/entries/speech-acts/. Green Speech Acts Stanford Encyclopedia comprehensive overview pragmatics linguistics philosophy.