Implicature
On this page
What a speaker means or suggests beyond what the words literally say, as when 'I have to work' conveys 'I am not coming to the party'. Paul Grice distinguished conversational implicature, which depends on context, from conventional implicature, which is fixed by particular words such as 'but' or 'therefore'.
Also: implicatures
The argument at a glance
A guide to the entry below, quoting its own words. Skip to the full entry and its sources.
- Question
- What licenses a hearer to take a speaker to have conveyed something she did not say, and how can that inference be told apart from what her words mean? The second half matters to philosophy of language directly: Richard Grandy and Richard Warner's Stanford entry on Grice reports his use of the distinction to defend the view that the material conditional gives the truth conditions of 'if', while what makes a conditional appropriate to assert is explained separately, by a conditional probability.
- Reasons , COT reconstruction: the numbering is ours, not the philosopher's
-
(COT's reconstruction, following Davis's Stanford entry and the Internet Encyclopedia article by Emma Borg, Antonio Scarafone and Marat Shardimgaliev)
- Conversation is a rational, cooperative activity whose participants are presumed to observe the Cooperative Principle and maxims, which Davis describes as 'instances of general rules governing rational, cooperative behavior'.
- A speaker has said that p.
- Taken at face value, saying only that p would fall short of a maxim in this context.
- The shortfall disappears, and the presumption of cooperation is preserved, on the hypothesis that the speaker thinks that q.
- The speaker can expect the hearer to reach this hypothesis and has done nothing to block it.
- Therefore the speaker has conversationally implicated that q. Borg and colleagues note that the step to (6) is an inference to the best explanation, not a deduction. (the entry continues below)
- Intended conclusion
- A large class of apparent meanings can be explained by general principles of rational communication rather than by enlarging the semantics of the words concerned, with what is said held fixed as the input to the explanation. Davis states the calculability condition in the words 'Conversational implicatures must be capable of being worked out.'
Objections and replies
- Objections , Raised by other thinkers and critics, as reported
-
Davis's own entry adds several criticisms to those just listed. Calculability is, he argues, an empirical hypothesis rather than a definition, and overgeneration shows the working-out schema to be invalid: the derivation that yields 'not all' from 'some' would equally yield unobserved implicatures such as 'not several'. Maxims clash, as in irony, where Quality and Manner pull apart, and principles of style and politeness compete with them. Grice characterised generalized implicatures as those a form of words 'would normally (in the absence of special circumstances)' carry; Davis finds this too strong, since the default reading does not always arise, and he, like Lepore and Stone, treats such implicatures as conventional in a broad, customary sense, which threatens the claim that they are worked out from the maxims at all. The objection known as Grice's Circle is that determining what is said already requires the same inferences that are supposed to start from it. Relevance theory raises a philosophical objection to the framework itself: Borg, Scarafone and Shardimgaliev report Sperber and Wilson's argument that the mutual knowledge Grice's account presupposes would require infinitely many knowledge states, beyond limited cognitive resources, and their rejection of the presumption that communication is cooperative.
- Replies
-
Davis answers Grice's Circle by noting that Grice's process is serial: what is said serves as a premise for determining what is implicated. Neo-Griceans answer overgeneration by reorganising the maxims, Horn into a Q Principle ('Say as much as you can') and an R Principle ('Say no more than you must'), Levinson by adding an M Principle and ordering the principles Q > M > R; Davis replies that each ordering misclassifies some cases. Relevance theorists replace the maxims with a single principle relating positive cognitive effects to processing effort; Davis objects that relevance cannot be measured and that the theory struggles to explain why a speaker chooses irony over literal speech, and Borg and colleagues report the charge that its principles are too vague to be falsified, with Wilson's (2017) reply that they would be falsified if interpretation were driven by informativeness rather than relevance. They also suggest that the mutual-knowledge objection may mistake inferential steps for representational states. Davis's own alternative explains speaker implicature by ordinary ways of inferring intentions, such as abduction, analogy and testimony, and sentence implicature by second-order conventions that are not part of a language's lexical and syntactic rules.
Coordinate 6.10Coordinate 6.10.2Paul GriceSources and primary textsRead the full entry
What it means
Implicature is the part of what a speaker communicates that goes beyond what is strictly said. The term was introduced by Paul Grice, who separated saying from implicating. In Wayne Davis's Stanford Encyclopedia example, Alan asks Barb whether she is going to the party and she replies 'I have to work': she says one thing and implicates another, namely that she is not going. Irony, understatement, hyperbole and metaphor are among the familiar figurative forms.
The question. What licenses a hearer to take a speaker to have conveyed something she did not say, and how can that inference be told apart from what her words mean? The second half matters to philosophy of language directly: Richard Grandy and Richard Warner's Stanford entry on Grice reports his use of the distinction to defend the view that the material conditional gives the truth conditions of 'if', while what makes a conditional appropriate to assert is explained separately, by a conditional probability.
Provenance. The term and the theory are Grice's own (originated). 'Logic and Conversation' was published in 1975 in Cole and Morgan's collection Speech Acts, pp. 41-58, as Davis cites it and as the publisher's DOI record confirms, and is also listed by Grandy and Warner in D. Davidson and G. Harman (eds), The Logic of Grammar (Encino, CA: Dickenson, 1975), pp. 64-75; it is reprinted in Studies in the Way of Words (Harvard University Press, 1989), pp. 22-40. 'Neo-Gricean' is a later label for the Horn–Levinson reworking of the maxims, and 'Grice's Circle' a later name for the objection that what is said cannot be fixed before implicatures are worked out.
Grice's two kinds are treated here in one entry because the contrast between them is the point of the notion. Conversational implicatures depend on the context of utterance and not only on sentence meaning. Conventional implicatures (which Davis calls semantic implicatures) are fixed by the meaning of particular words: 'but' signals a contrast and 'therefore' an inference, though neither contrast nor inference is part of what is said. The SEP lists the tests usually used to tell them apart: conversational implicatures can be cancelled ('some athletes smoke; indeed all do'), are not detachable (another way of saying the same thing carries the same suggestion) and are calculable, that is, a hearer can work them out. Conventional implicatures, by contrast, cannot be cancelled in this way. The SEP's 'some athletes smoke' case is a generalised conversational implicature, one carried by a form of words in most contexts ('some' tends to suggest 'not all').
Grice explained conversational implicature by a Cooperative Principle and four maxims: Quality (say what you believe true), Quantity (be as informative as required), Relation (be relevant) and Manner (be perspicuous). A hearer who assumes the speaker is cooperating infers whatever is needed to make the utterance fit the maxims, and a speaker can exploit a maxim by openly flouting it. 'Grice's Razor' recommends explaining an extra meaning as implicature rather than positing a new sense of the word.
The text's own reasoning. Grice presents the derivation of an implicature as a pattern a hearer could follow, not as a formal proof. Davis gives it in paraphrase: the speaker has said something; there is no reason to suppose she is not observing the maxims, or at least the Cooperative Principle, which in Grice's words, as Grandy and Warner print them, begins 'Make your conversational contribution such as is required, at the stage at which it occurs'; she could not be doing so unless she believed some further proposition; she knows the hearer can see this and has done nothing to stop him thinking it; so she has implicated it. The hearer's data are the conventional meaning of the words, the Cooperative Principle and maxims, the context, background knowledge, and the mutual availability of all of these. In Grandy and Warner's example, someone asked where petrol can be had replies 'There is a station around the corner', and the hearer infers that the speaker believes it is open, since otherwise the remark would not be relevant.
Structure. (COT's reconstruction, following Davis's Stanford entry and the Internet Encyclopedia article by Emma Borg, Antonio Scarafone and Marat Shardimgaliev) (1) Conversation is a rational, cooperative activity whose participants are presumed to observe the Cooperative Principle and maxims, which Davis describes as 'instances of general rules governing rational, cooperative behavior'. (2) A speaker has said that p. (3) Taken at face value, saying only that p would fall short of a maxim in this context. (4) The shortfall disappears, and the presumption of cooperation is preserved, on the hypothesis that the speaker thinks that q. (5) The speaker can expect the hearer to reach this hypothesis and has done nothing to block it. (6) Therefore the speaker has conversationally implicated that q. Borg and colleagues note that the step to (6) is an inference to the best explanation, not a deduction. It is reasoning about what a rational agent is doing, which is why conversational implicatures are meant to be calculable, whereas conventional implicatures, fixed by word meaning, need no such derivation.
Intended conclusion. A large class of apparent meanings can be explained by general principles of rational communication rather than by enlarging the semantics of the words concerned, with what is said held fixed as the input to the explanation. Davis states the calculability condition in the words 'Conversational implicatures must be capable of being worked out.'
The SEP surveys the main objections: implicature occurs where cooperation is absent (among adversaries, in private writing); the Gricean reasoning overgenerates, predicting implicatures that do not arise; and figurative implicatures are often indeterminate. Neo-Gricean theorists such as Laurence Horn and Stephen Levinson reduced the maxims to a few principles, while Dan Sperber and Deirdre Wilson replaced them with a single principle of relevance. Robyn Carston's 'explicature' and Kent Bach's 'impliciture' mark contested territory between what is said and what is implicated. Implicature is a central topic of pragmatics and matters even to truth-conditional semantics, since it explains why literally true statements can mislead.
Objections. Davis's own entry adds several criticisms to those just listed. Calculability is, he argues, an empirical hypothesis rather than a definition, and overgeneration shows the working-out schema to be invalid: the derivation that yields 'not all' from 'some' would equally yield unobserved implicatures such as 'not several'. Maxims clash, as in irony, where Quality and Manner pull apart, and principles of style and politeness compete with them. Grice characterised generalized implicatures as those a form of words 'would normally (in the absence of special circumstances)' carry; Davis finds this too strong, since the default reading does not always arise, and he, like Lepore and Stone, treats such implicatures as conventional in a broad, customary sense, which threatens the claim that they are worked out from the maxims at all. The objection known as Grice's Circle is that determining what is said already requires the same inferences that are supposed to start from it. Relevance theory raises a philosophical objection to the framework itself: Borg, Scarafone and Shardimgaliev report Sperber and Wilson's argument that the mutual knowledge Grice's account presupposes would require infinitely many knowledge states, beyond limited cognitive resources, and their rejection of the presumption that communication is cooperative.
Replies. Davis answers Grice's Circle by noting that Grice's process is serial: what is said serves as a premise for determining what is implicated. Neo-Griceans answer overgeneration by reorganising the maxims, Horn into a Q Principle ('Say as much as you can') and an R Principle ('Say no more than you must'), Levinson by adding an M Principle and ordering the principles Q > M > R; Davis replies that each ordering misclassifies some cases. Relevance theorists replace the maxims with a single principle relating positive cognitive effects to processing effort; Davis objects that relevance cannot be measured and that the theory struggles to explain why a speaker chooses irony over literal speech, and Borg and colleagues report the charge that its principles are too vague to be falsified, with Wilson's (2017) reply that they would be falsified if interpretation were driven by informativeness rather than relevance. They also suggest that the mutual-knowledge objection may mistake inferential steps for representational states. Davis's own alternative explains speaker implicature by ordinary ways of inferring intentions, such as abduction, analogy and testimony, and sentence implicature by second-order conventions that are not part of a language's lexical and syntactic rules.
Key thinkers and texts. Paul Grice, 'Logic and Conversation' (1975) and Studies in the Way of Words (1989); Laurence Horn, 'Towards a New Taxonomy for Pragmatic Inference: Q-based and R-based Implicature' (1984); Dan Sperber and Deirdre Wilson, Relevance: Communication and Cognition (Harvard University Press, 1986; second edition, Blackwell, 1995); Kent Bach, 'Conversational Impliciture', Mind and Language 9 (1994); Wayne Davis, Implicature: Intention, Convention, and Principle in the Failure of Gricean Theory (Cambridge University Press, 1998); Stephen Levinson, Presumptive Meanings: The Theory of Generalized Conversational Implicature (MIT Press, 2000); Robyn Carston, Thoughts and Utterances (Blackwell, 2002).
Current status. Grice's distinction between saying and implicating remains the common starting point, but whether the Cooperative Principle and maxims explain implicature is disputed. Davis, himself a party to the debate, concludes that the Gricean hypotheses 'are empirical hypotheses that may turn out to be false' and provide neither 'the only nor the best explanations of the data'. Neo-Gricean, relevance-theoretic and conventionalist accounts all remain in contention, and no consensus has formed.
Related entries on this site. Speaker meaning sets out the intention-based analysis on which Grice's notion of implicating depends; Truth-conditional semantics is the semantic theory against which what is said is usually measured; Speech act and Use theory of meaning treat neighbouring accounts of what speakers do with words.
On this site it belongs to Coordinate 6.10 (Philosophy of Language) and Coordinate 6.10.2 (Speech Acts).
See also the glossary entries Pragmatics, Presupposition, Speech act, Semantics and Philosophy of language.
The questions it belongs to
- 6.10 Philosophy of LanguageHow do noises come to be about things?
- 6.10.2 Speech ActsHow does saying "I promise" make a promise?
Thinkers
Traditions
To read
Sources and further reading
- Wayne Davis, 'Implicature', Stanford Encyclopedia of Philosophy (substantive revision 10 January 2024)
- Kepa Korta and John Perry, 'Pragmatics', Stanford Encyclopedia of Philosophy (substantive revision 28 May 2024)
- Richard E. Grandy and Richard Warner, 'Paul Grice', Stanford Encyclopedia of Philosophy (substantive revision 19 June 2026)
- Emma Borg, Antonio Scarafone and Marat Shardimgaliev, 'Meaning and Communication', Internet Encyclopedia of Philosophy
- H. P. Grice, 'Logic and Conversation', in Speech Acts (Cole and Morgan, eds, 1975), pp. 41-58, DOI 10.1163/9789004368811_003 (publisher metadata checked via Crossref; text not opened)
Cite this page
Responsibility: Coordinates of Thought Editors. Each statement was drafted from the sources listed above and checked against them in a separate pass before publishing; this page is not peer-reviewed by outside specialists. Source-checked 15 November 2026. How to cite this site · How we write and check.
APA (7th edition)
Coordinates of Thought Editors. (2026, November 15). Implicature (Arguments). Coordinates of Thought. Retrieved [date accessed], from https://coordinatesofthought.com/arguments/implicature/
MLA (9th edition)
“Implicature (Arguments).” Coordinates of Thought, 15 Nov. 2026, coordinatesofthought.com/arguments/implicature/. Accessed [date accessed].
Chicago (17th edition)
Coordinates of Thought Editors. “Implicature (Arguments).” Coordinates of Thought. Source-checked November 15, 2026. https://coordinatesofthought.com/arguments/implicature/.
BibTeX
@misc{cot-glossary-implicature-2026,
author = {{Coordinates of Thought Editors}},
title = {Implicature (Arguments)},
howpublished = {Coordinates of Thought},
year = {2026},
month = nov,
url = {https://coordinatesofthought.com/arguments/implicature/},
note = {Source-checked 15 November 2026}
}
RIS
TY - ELEC AU - Coordinates of Thought Editors TI - Implicature (Arguments) T2 - Coordinates of Thought PB - Coordinates of Thought PY - 2026 DA - 2026/11/15 UR - https://coordinatesofthought.com/arguments/implicature/ LA - en ER -
For reference managers (Zotero, Mendeley, EndNote): BibTeX file · RIS file