Grice distinguished what is said (literal, truth-conditional content) from what is conveyed through implicature. "It's cold in here" literally states a fact; said to someone near a window, it implicates they should close it. This distinction grounds pragmatics as a distinct discipline and explains how communication succeeds despite gaps between literal meaning and speaker intent.
Use clear examples of implicature: "Do you have a pen?" said to a locked store implicates it's closed. Analyze what is literally said (question about pen-possession) versus what is communicated (store status). Study Grice's tests for what-is-said: consistency with other linguistic elements, relation to linguistic meaning, contrast with implicature. Practice distinguishing saying from implicating in dialogue examples.
From your study of conversational implicature, you know that much of what gets communicated in everyday speech goes beyond what the words literally say — that "Can you pass the salt?" is not primarily a question about physical capability, and that "It's a bit cold in here" said to someone near an open window conveys a request, not a meteorological report. From your study of speaker meaning, you know that communication involves the speaker intending the hearer to recognize an intention. Grice's concept of what is said draws a precise boundary around the literal, truth-conditional core of an utterance — the part that must be true for the utterance to be true — and separates it cleanly from everything else that gets communicated.
The distinction matters because it grounds pragmatics as a discipline. Without a clear notion of what is said, there is no way to define what is *merely* implicated. Grice's picture is: the semantic content of an utterance — what its words literally mean plus their grammatical composition — determines a proposition. That proposition, once it is appropriately contextualized (reference resolved, ambiguity disambiguated), is what is said. Everything communicated beyond that is pragmatic: it arises from the interaction of what is said with the cooperative norms of conversation (the maxims of quantity, quality, relation, and manner). Implicature is calculable: a hearer can reason from what was said, the assumption of cooperation, and the context to what was implicated.
To use Grice's own tools: if you ask "How is your essay going?" and I say "I haven't started yet; at least the weather's been nice," I have said that I haven't started and that the weather has been pleasant. I have implicated — but not said — that my essay is going badly, perhaps irreparably. The implicature is defeasible: I could add "and I work brilliantly under last-minute pressure" and cancel the negative implication without contradicting myself. This defeasibility test is one of the key markers that distinguishes implicature from what is said: you can't cancel what is said without contradiction, but you can cancel implicature.
The boundary is less stable than it first appears. Consider "She took the keys and left." What is said? Grammatically, there is no ordering or causal connective — just a conjunction. Yet hearers strongly interpret the utterance as meaning she took the keys and *then* left, and perhaps that she left *because* she had the keys. Is that ordering part of what is said, or merely implicated? Grice's original account suggests it is implicated (the maxim of manner and the expectation of temporal order in narrative generate the inference), but later theorists have argued that some pragmatic processes enrich what is said rather than merely adding to it. This debate — between minimalist views that keep what is said close to semantics and contextualist views that allow pragmatic processes to shape the truth conditions themselves — is one of the central live disputes in philosophy of language and linguistics.
Why does it matter? The what-is-said/implicature distinction has practical consequences in legal interpretation, in understanding political speech, and in the theory of communication itself. A politician who says "I never took money illegally" may implicate that they took no money, while insulating the literal claim from falsity. A contract that is technically satisfied by the literal words may violate the spirit and implication of what was agreed. Grice's framework gives us the conceptual vocabulary to make these distinctions explicit — to ask precisely: what did the utterance say, and what did it communicate beyond that?
Topics in reflective domains aren't scored by quiz answers. Read, reflect, and mark when you've thought it through.