Rules of Conversation

Grice says that in carrying on a conversation, there are rules. There is one over-arching rule, the _cooperative principle,_

Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk-exchange in which you are engaged. 191

That rule has many corollaries like, be as informative as is required, but no more, don't lie or mislead, and so forth. The reason that such rules are important is that they go a long way toward explaining how speaker meaning can differ from literal meaning.

Examples:

the guy over there drinking the martini

There's the door.

Well, his handwriting is excellent.

Ms. X produced a series of sounds that corresponded quite closely to the score of Handel's "I Know That My Redeemer Liveth."

Grice's idea is that we often use sentences to express other than their literal meaning, but we almost never flout the rules of conversation, and so we use that fact that an utterance appears to flout the rules of conversation to determine what the speaker meaning actually is, since the speaker meaning must be consonant with the rules of conversation. There is standard movie comedy routine type where person A is drunk or stoned or whatever and person B doesn't know it. Person A is flouting the rules of conversation (an extraordinary occurrence explained by inebriation), but, since Person B doesn't know Person A is inebriated, Person B thinks Person A is speaker meaning something humorously inappropriate.

Grice introduced the term "implicate" for what is said without literally being "implied" or (logically) "entailed." Hence, the term "conversational implicature."

"Relevance" theorists (Carston, Recanati) argue that Grice is wrong, that instead of "rules of conversation" there are general principles about how we transmit and receive information that lead to the implicatures.

Both theories have the problem that they lack positive explanations. Grice can do amazingly well at accounting, using not too many rules, for when we should expect an utterance not to have its literal meaning. But the second part, determining what meaning an utterance does have is not adequately explained.

Conventional Implicature

Some words cue expectations that are not literally part of what is asserted: If you say "A and B," the conjunction is false if either conjunct is, and the same is true of "A but B." But 'but' carries the implication that A somehow inhibits or contrasts with B. Unlike conversational implicature, conventional implicature is "noncancellable." You can block a conversational implicature by explicitly denying it:

Is Sally a good philosopher?
She's a great ping-pong player.
But don't get me wrong, she's also a good philosopher.

On the other hand,

She's poor but honest.
* But don't get me wrong, poor people are no less likely to be honest than anyone else.

Presuppositions

Strawson takes "The present King of France is bald" to conventionally implicate that there is a present King of France. He takes that there is a present King of France to be a "presupposition" of saying "The present King of France is bald." His theory of presuppositions can be understood as a theory of conventional implicature. His evidence for the presupposition is that one wouldn't say "No, he isn't" to someone who said "The present King of France is bald" and that "The present King of France is bald" is therefore not false. What one would say is "France doesn't have a King." But a Gricean can explain that even if "The present King of France is bald" is false: to say "No, he isn't," while true, violates a rule of conversation since one can say something more informative, namely, "France doesn't have a King." Such considerations have led many to conclude that there are no presuppositions in ordinary language.

Similarly, philosophers often want to say that "if A, then B" just means that either A is false or B is true. That seems wrong because it seems that A and B must be suitably related for "if A, then B" to be true, but that can be explained away as a matter of implicature.

-- ShaughanLavine - 09 Mar 2006