THEORY-444: To mean something by an utterance, in the non-natural sense, is to intend it to produce an effect in an audience by means of the audience's recognition of that intention, where the recognition is meant to be a reason for the effect and not merely its cause; an intention to produce the effect, or to have it recognised, is not enough

Source

Grice (1957), LIT-1042 (The Philosophical Review 66(3):377–388), read in full in NOTE-735.

What was actually shown

Two senses of “mean” (pp. 377–378). Natural meaning (“Those spots mean measles”) entails its content and has no one who meant it. Non-natural meaning (“Those three rings on the bell mean that the bus is full”) reverses all five of Grice’s linguistic tests.

Against the causal theory (pp. 379–381). Stevenson’s tendency to produce an attitude, “dependent on an elaborate process of conditioning”, either lets the tail coat mean a dance or is circular. It also gives no account of what a speaker means on one occasion, which Grice takes as basic.

The analysis, reached by counterexample (pp. 381–384).

• An intention to induce a belief is not enough: the handkerchief left at the scene of a murder means nothing.
• An intention that the audience also recognise that intention is not enough: Herod showing Salome the head, the child letting its mother see how pale it is.
• The recognition must do work. Showing Mr. X a photograph means nothing, because it would convince him if left by accident. Drawing him a picture means something. A deliberate frown has its effect “only provided the audience takes it as intended to convey displeasure” (p. 383).
• Hence “A uttered x with the intention of inducing a belief by means of the recognition of this intention”, and generally (p. 385): “A meantNN something by x” is roughly “A intended the utterance of x to produce some effect in an audience by means of the recognition of this intention”. The effect can be an action: pointing to the door.

Reason, not cause (p. 385). The grunt meant to make someone blush shows the intended effect “must be something which in some sense is within the control of the audience”. The recognition is “a reason and not merely a cause”. Of cutting someone in the street: “one can refuse to be offended”.

What could have come out otherwise. A case with recognition doing no work could have counted as meaning, or one with recognition doing the work could have counted as not meaning. Grice’s cases sort the other way.

What this does not say

• What act was performed. The paper analyses meaning, not illocutionary force. It says nothing about what separates a warning, a request or a reproach when the same thing is meant. That step is Strawson’s (THEORY-511), and NOTE-735 says so.
• That the three intentions are sufficient. Strawson’s watcher case satisfies them without communication, and needs a fourth intention (LIT-1043, pp. 446–448).
• That sentence meaning reduces to speaker meaning. Grice is “sympathetic” to that and “aware that the point is controversial”. The reduction of timeless meaning is “a first shot”, and the red light, which means stop without anyone meaning it, is admitted as awkward (C5, weak).
• That speakers run complicated psychological occurrences. Grice disclaims “peopling all our talking life” with them. Intentions are read by public criteria: “an utterer is held to intend to convey what is normally conveyed”, and context decides between candidates (pp. 386–388).
• That the analysis is exact. It is stated “roughly” throughout, and the condition on reasons (C4) rests on one case.

Connections

• Extended by THEORY-511. Strawson adds a fourth intention and carries the analysis from meaning to illocutionary force.
• THEORY-172. The rational-speech-act listener inverts a model of a speaker who chooses by informativeness, and its papers credit Grice. In the formulation of LIT-821 nothing is an intention that the choice be recognised, so that account is Gricean in its inference, not in this analysis of meaning. That observation is NOTE-735’s, not Grice’s.
• QUESTION-073, CLAIM-221. The cut in the street counts as meaningNN because the offence depends on recognising the intention to offend, a speaker-internal individuation in CLAIM-221’s sense. But the paper individuates no act types, so it does not answer QUESTION-073, and no answers edge is declared.
• Brandom (THEORY-665) explains content by normatively described social practice, “not intentional” vocabulary, which reverses Grice’s hoped-for order of explanation. As the two accounts are stated here they do not exclude each other, so no rivals edge is declared.