Conversational implicature is meaning that a speaker conveys through an utterance without stating it explicitly, and that a hearer can infer by considering the utterance’s linguistic meaning, its context, and expectations about purposeful communication. A central concept in pragmatics, it explains how speakers can communicate more than their words literally express. If someone asks, “Are you coming to dinner?” and receives the answer “I have to work tonight,” the answer may implicate that the respondent will not attend, although having to work does not logically exclude attending dinner.
The concept was developed by the philosopher H. P. Grice as part of a broader account of speaker meaning. Conversational implicature is a subtype of implicature, distinguished by its dependence on reasoning about communication rather than on a conventional meaning attached to a particular expression.
Linguistic meaning and communicated meaning
An utterance can communicate several kinds of information at once. Its words and grammatical structure contribute linguistic meaning; contextual interpretation helps determine what is said; further reasoning can yield what is conversationally implicated.
For example:
A: Has the report been approved?
B: The director has not read it yet.
B explicitly states that the director has not yet read the report. If both participants know that the director’s review is required for approval, B also conveys that the report has not been approved. That additional message depends on background assumptions, not simply on the sentence’s vocabulary and syntax.
Conversational implicature therefore belongs to the study of language use, although its analysis closely interacts with semantics. It is not equivalent to any conclusion a listener happens to draw. A hearer might infer from a speaker’s trembling voice that the speaker is nervous, but nervousness is not thereby a conversational implicature: the speaker need not have intended to communicate it. In the Gricean account, implicature concerns meaning that the speaker makes available for recognition through communicative reasoning.
The cooperative principle
Grice explained conversational implicature through the cooperative principle: participants normally interpret contributions as appropriate to the accepted purpose or direction of an exchange. “Cooperation” here does not mean friendliness, agreement, or complete honesty. It concerns the expectations that make contributions interpretable as parts of a communicative activity.
Grice organized these expectations into four categories of conversational maxims:
| Category | Central expectation |
|---|---|
| Quantity | Provide as much information as the exchange requires, but not more than is required. |
| Quality | Do not assert what is believed false or unsupported by adequate evidence. |
| Relation | Be relevant. |
| Manner | Be clear, orderly, and appropriately concise; avoid obscurity and ambiguity. |
These are analytical principles rather than rules of etiquette or prescriptions for good writing. Speakers often depart from them, and some departures themselves help convey meaning.
Inference from an apparently unhelpful answer
Consider:
A: Where can I buy some milk?
B: There is a shop around the corner.
B’s answer commonly implicates that the shop sells milk and may be open. Otherwise, mentioning it would not obviously help answer A’s question. The hearer interprets the contribution as relevant and supplies a connection between the shop and the immediate purpose of the exchange.
The inference is defeasible. If B adds, “But it closed an hour ago,” the interpretation changes. Its dependence on circumstances is precisely what distinguishes it from a conclusion guaranteed by sentence meaning.
Flouting and covert violation
A speaker may flout a maxim by departing from it conspicuously enough to invite an interpretation beyond the literal statement. Saying “What wonderful weather” during a downpour may communicate a negative assessment through verbal irony, provided the context makes that intention recognizable.
Flouting differs from covert violation. Someone who lies while intending to be believed may secretly violate the maxim of quality rather than invite recognition of an ironic message. An apparent failure to follow a maxim therefore does not automatically generate implicature. It may instead reflect deception, misunderstanding, lack of knowledge, or refusal to participate.
Implicatures can also arise without an obvious breach. An appropriately informative answer may acquire additional meaning because a more informative alternative was available but not chosen.
Main types
Particularized conversational implicature
A particularized conversational implicature requires a relatively specific context.
A: Shall we sit outside?
B: The mosquitoes are already out.
In a setting where mosquitoes are an unwelcome nuisance, B may implicate a preference to sit indoors. In a discussion among researchers observing mosquito activity, the same sentence may support a different interpretation.
Its interpretation depends on such factors as the current question, the participants’ goals, and their common ground: information they treat as shared for the purposes of the exchange.
Generalized conversational implicature
A generalized conversational implicature is commonly associated with an expression or construction without requiring an unusual context. It remains defeasible rather than being guaranteed by linguistic meaning.
For example:
Some of the applicants passed.
This often suggests that not all of the applicants passed. Under a standard inclusive interpretation, however, “some” is compatible with “all.” The stronger conclusion can be suspended:
Some of the applicants passed—in fact, all of them did.
The distinction between generalized and particularized implicatures concerns their dependence on special contextual assumptions. It does not establish that generalized implicatures are necessarily automatic or processed before context is considered.
Scalar implicature
Scalar implicature is an important family of cases involving alternatives ordered by informational strength. Typical examples include “some” and “all,” or “possible” and “certain.”
When a speaker chooses a weaker expression, a hearer may reason that the speaker was not in a position to assert a stronger one. Given further assumptions about the speaker’s knowledge, this can support the conclusion that the stronger alternative is false.
These steps are not equivalent:
- The speaker does not know that all applicants passed.
- The speaker knows that not all applicants passed.
Which interpretation is justified depends partly on whether the speaker is assumed to have complete information. Scalar inferences are also sensitive to the conversational question, the relevant alternatives, and the grammatical environment.
Characteristic properties
Grice identified several properties that help distinguish conversational implicature from other kinds of meaning. They are best treated as interconnected diagnostics rather than an infallible checklist.
Cancellability
A conversational implicature can often be explicitly withdrawn without contradicting what was said:
There is a shop around the corner, but it does not sell milk.
It can also disappear in a context that blocks the usual inference. “Some students passed” need not imply “not all” if the speaker is reporting only the portion of the results available so far.
Cancellation may nevertheless sound awkward, misleading, or socially inappropriate. Such awkwardness does not by itself establish a logical contradiction.
Calculability
An implicature should be explainable through a reasoning process involving what was said, the context, conversational expectations, and assumptions about the speaker.
This does not mean that hearers consciously reconstruct an argument every time they understand an indirect message. Calculability is primarily an explanatory requirement: an analysis should show why that interpretation is available.
Non-detachability
Many conversational implicatures persist when an utterance is replaced by another expression conveying the same relevant content. If “I must work this evening” implicates declining an invitation, a suitable paraphrase may do so as well.
There are important exceptions. Implicatures involving manner can depend on a particular formulation. An unnecessarily elaborate description may suggest hesitation or unusual circumstances precisely because a simpler expression was available.
Context dependence and indeterminacy
A speaker may convey a range of related suggestions rather than one uniquely identifiable proposition. A conspicuously lukewarm recommendation can suggest limited enthusiasm without specifying exactly which stronger judgment the writer holds.
Consequently, an analysis should not assume that every implicature has one fully determinate verbal paraphrase.
Distinctions from related concepts
Entailment
Entailment is a relation in which the truth of one proposition guarantees another. “Every applicant passed” entails that no applicant failed. Conversational implicature is normally defeasible: its conclusion can be withdrawn while the asserted content remains true.
The contrast is not simply between explicit and implicit information. Entailed information may also go unstated; what matters is whether the conclusion follows necessarily from the relevant semantic content.
Presupposition
A presupposition is information presented as background or taken for granted. “Maya stopped cycling to work” ordinarily presupposes that Maya previously cycled to work. That background inference often survives negation: “Maya did not stop cycling to work.”
Conversational implicatures arise differently and commonly change or disappear under alterations to the utterance or context. Nevertheless, presuppositions can also be context-sensitive, so persistence under negation is not a complete test by itself.
Conventional implicature
A conventional implicature is traditionally understood as a non-at-issue contribution associated with an expression’s conventional meaning, rather than derived through the conversational maxims. In Grice’s classic treatment, “but” contributes a contrast in a sentence such as “She is young but experienced.”
The category and its boundaries remain debated. The relevant distinction is between conventionally encoded meaning and meaning inferred through the circumstances of communication.
Indirect speech acts
An indirect speech act occurs when a speaker performs an act through an utterance whose direct form suggests another act. “Can you close the window?” is grammatically a question about ability but commonly functions as a request.
Conversational inference can help explain such uses, but implicature and indirect speech acts are not identical. The former concerns indirectly conveyed meaning; the latter concerns the communicative action performed, such as requesting, promising, or refusing.
Historical development and theoretical approaches
Grice presented his account in the 1967 William James Lectures at Harvard University. “Logic and Conversation,” published in 1975, became its best-known formulation; related work was collected in Studies in the Way of Words in 1989. The account helped establish a systematic distinction between conventional linguistic meaning and meaning communicated through rational interpretation.
Subsequent approaches differ over how much explanatory work should be assigned to general communicative reasoning, default expectations, or grammatical mechanisms.
- Neo-Gricean approaches reorganize or reduce the original maxims and develop systematic accounts of informational alternatives, stereotypical interpretations, and marked wording.
- Relevance theory explains interpretation through expectations of relevance, balancing cognitive effects against processing effort rather than deriving interpretation from Grice’s four categories.
- Grammatical approaches to scalar inference propose that some strengthening effects are produced by semantic or grammatical mechanisms, often involving an unpronounced exhaustivity operator, rather than solely by reasoning about conversational conduct.
The boundary between what is said and what is implicated is itself disputed. Some context-dependent material appears necessary to obtain a complete proposition or determine truth conditions. Competing accounts classify such material as contextual enrichment, part of explicitly communicated content, or implicature, with consequences for truth-conditional semantics and theories of linguistic interpretation.
Evidence, variation, and applications
Research in psycholinguistics examines when implicatures arise, how quickly they are processed, and how interpretation changes with context, cognitive demands, and access to alternatives. Studies of language acquisition investigate how children learn to recognize messages that are not directly encoded. Performance on an implicature task depends not only on linguistic knowledge but also on the task’s wording, available alternatives, and assumptions about the speaker.
Interpretation also interacts with sociolinguistics and linguistic politeness. An indirect answer may soften a refusal, avoid an explicit criticism, or preserve uncertainty. Such functions are not universal properties of particular sentences: social relationships, local conventions, and communicative settings affect what is inferred.
In translation and interpreting, retaining literal wording may fail to preserve an implicature when the audience lacks the relevant background or the target language organizes alternatives differently. Conversely, making an implication explicit changes how the message is presented and may remove its original ambiguity or indirectness.
Computational treatments, including work on artificial intelligence, use implicature to study context-sensitive language understanding. A central difficulty is distinguishing an inference the speaker intended from one that is merely plausible. Generating a likely unstated conclusion does not by itself establish that the conclusion was communicated.