The Conversationalist Hypothesis

Grice on 'an X':

-- By JLS
----- for the GC

Again: this is meant as reply to comment by Kramer in "And then" -- bloggers.com disallowed me to post it in that thread:

Yes. Kramer is right. I suppose Kramer's position reminds me of L. J. Cohen's. His claim of fame was to criticise Grice (in Cohen, "Grice on the logical particles of natural language"). This was taken very nicely. R. Walker, a Scots, who, like Cohen, had Oxford associations, replied to Cohen's attack in full. (Cohen is the 'attacker'. He had previously challenged Austinians with "Is there such a thing as illocutionary force?"). Walker's essay appeared in a pretty influential edition -- ed. Platts et al. Walker's essay was called, unimaginatively, "Conversational implicature". It is a long, and somewhat boring, essay.

Cohen then re-attacked with "Can the conversationalist hypothesis be defended" (and then he died). The "Can the conversationalist..." appeared in "Philosophical Studies" and it's pretty short.

The negative parts of Cohen are well taken, and indeed, Grice seems to have been aware of the fault line -- Cohen criticised Grice for failing to account for the implicatum in an 'embedded' context. If x implicates p, should we accept that -x implicates p, or not p, or what? etc. ---

---

NOW, what reminds me of Cohen about you -- is his POSITIVE, constructive line of thought. He called this the "Semantic Hypothesis", versus the "Conversationalist Hypothesis". He wants to say, like you, that 'and' has a couple of 'senses': one is the truth-functional connective, and it has others, as in 'safe and sound' and 'and then', etc. --.

You are right that polysemy is a bad word for me. I suppose a good passage by Grice that articulates my antipathy on this front is as per what follows. He is concerned with the alleged sense of the word 'a':

as in 'a bitch'.

-----

Suppose we gather a few samples of the uses, in natural English, as you call it, 'a':

--- "I was walking towards the house and I found a tortoise inside the front door".

---- Another sample:

"Smith is meeting a woman this evening".

---- Another sample:

"I broke a finger yesterday".

---- Yet another sample:

"I've been sitting in a car all morning".

So now we are asked to define 'a':

Grice imagines someone pretty much like Cohen. He is NOT concerned with lexicologists which are the most boring people you'll ever meet (unless you love them). Grice is concerned with philosophers criticising philosophers doing 'lexicology'.

Grice writes:

"I am inclined to think that one would NOT lend a sympathetic ear to a philosopher who suggested that there are three senses of the form of an expression 'an X'. One in which it means, roughly, (1)"

1. [something that satisifies the conditions defining the word X]

This would be Kramer receding to the meta-level. I find that particular first shot pretty unhappy in that surely you are NOT allow to recede to the meta-level when writing a dictionary!

Second:

"Another in which 'an X' means, approximately,"

2. [an X (in the FIRST sense) that is ONLY REMOTELY connected in a certain way to some person indicated by the CONTEXT].

This sins (Grice holds) and it uses Kramer's favourite word, 'context', or co-text, as I prefer. Cf. pre-text.

Then there's

"yet another in which it means"

3. [an X (in the FIRST sense) that is CLOSELY related in a certain way to some person indicated by the context].

This one Grice alleges this 'philosopher' would appeal to, to account for the uses of 'a' in cases such as "I broke a finger yesterday". (Horn provides the cancellability here in having a NURSE uttering that in a hospital: I broke five fingers yesterday).

Grice goes on and asks, rhetorically:

"Would we NOT much prefer an account on the following lines?"

"When someone, by using the form of expression "an X", implicates that the X does NOT belong to or is NOT otherwise clsely connected with some identifiable person, the implicatum is present because U has failed to be specific in a way in which he might have been EXPECTED to be specific, with the consequence that it is likely to be assumed that he is NOT in a position to be specific"

--- I found a tortoise in a house.

---

"My hearer would be NORMALLY surprised if some time later I revealed that the house was MY house."

For the other cases, the implicatum is different. But the above is, Grice notes, "the familiar implicature situation and is classifiable as a failure, for one reason or another, to fulfil the first maxim of Quantity."

----

He does not provide a gloss for the other two 'alleged' "senses" but I can. The Quantity flout accounts for the alleged 'remote' sense. For the 'closely related' sense: "I broke a finger yesterday", I would assume that the situation is the reverse.

It usually cannot be for lack of evidence that U fails to specify which finger he broke. So in this case, a specification would flout the maxim 'be brief' -- along with 'do not be more informative than is required' -- "I broke my middle finger".

As for the 'first sense':

"A mind is a terribly thing to waste". "A waist is a terribly thing to mind".

This is possibly idiomatic. And it is NOT rendered by the use of (Ex), unlike the other two alleged 'senses'. So, it is a matter of the idiom -- in the given language -- that one uses the indefinite:

"Homo hominis lupus" -- a man is a wolf for another man. Note that 'a' is dropped in Latin:

"Unus homo"

WAS available to Cicero, but the implicatum, in Latin, would have been that it's only JUST one man who is a wolf to just another one man. And we don't want that.

"A rose by any other name would smell so sweet".

Here 'the rose' will not do. So, again, the 'a' is merely idiomatic. And it means, for any x, if x is a rose, then translating that to a higher order predicate such that we want to refer to a different label for x, any label other than 'rose', the odour of x would remain constantly sweet. Or not.