Conversational Inferences -- and how to fail them

By JLS
--- for the GC.

Again, this was meant in thread "Bandaged-Leg Squash Player" but I was unable to submit it in that thread, so copying and pasting it in this blog post.

We are discussing the 'if'-schema of most implicature stuff, according to Kramer.

Alas, it does not seem to be obvious (I hate that cumbersome phrase but hey) how 'modus ponens' (which Kramer does not use) relates to his 'fine-tuning' his original example. But let's have a look:

Kramer writes:

"One could fine-tune this line of thought [i.e. modus ponens as per above] to include inferences from bandaged leg to bad leg, or from "cannot play" to "will not be playing," but the point is that only the minor premise, the bandaged leg, is made explicit by word or deed, i.e., by CB. The rest of the syllogism is already available to A, so the CB need only supply the bandaged leg and, thereby, invoke those other elements."

I wonder if the idea is that ALL implicature works like this. Consider Grice's first example of an implicature:

How is Smith getting on in his new job at the bank?

B: Oh, quite well. He likes his colleagues, and he hasn't been to prison yet.

---- What would the minor premiss be here?

I submit:

"If a person hasn't been to prison yet, he is an honest chap".

But then is that SO?

I don't think so. I think the implicatum is triggered by the use of 'not': "He has NOT been to prison yet".

Rather, then, the minor (or major, because compared to the frivolities one endures most of the days -- on television --e.g. whether Nicole Kidman is pregnant, or if Ricky Martin is gay -- I cannot see how one could call a premiss (an important part of one's reasoning) 'minor') premise seems to be, rather,

"If you negate something, it's because someone (or other) has affirmed it, or will affirm it".

---

I do agree with Kramer that SOME implicatures are enthymematic (to 'implicate' is BY DEFINITION, to leave something 'unsaid' -- it's the sousentendue of J. S. Mill) and have the conditonal form he suggest they have. I'm less sure that ALL do. But then blame ME for thinking he should have meant that ALL do in the first place.

Consider:

"What month has 28 days?"
"All of them do".

Why suspect that the questioner is trying to trick his co-conversationalist. Why would a flouting of 'be as informative as is required' be judged to rely on a modus ponens involving an enthymeme?

What colour is the Japanese flag?
Red.

Kramer disagrees with me, and think that that would be a FALSE answer. So, he seems to be wanting to say that if a conditional enthymeme is involved here, it constraints the 'entailments' of what one reads between the lines. I don't read between the lines of "What colour is the Japanese flag?" as a request to provide ALL the colours that (rather unimaginative) flag displays. In any case, I would not regard "The Japanese flag is red" as a false thing to say. "The only colour that the Japanese flag is red" IS on the other hand, a falsity (insofar as 'white' IS a colour -- and the flag is clean of course).

---

What we may need is a symbolisation alla Grice in terms of letter schema: p, q. Using "that p" and "that q" as contents of one's beliefs or desires. And see why a conversational implicatum relies on 'if' structures like that and what IS that is left unsaid.

Of course I disagree with all the emphasis on the unspoken and the unsaid, and the latent and the implicit. It would seem that while Grice CAN be a literalist (he speaks of 'say') some of US want to generalise what he has to say about communication into channels other than the verbal.

If a man makes the gesture, "Are we jogging tonight?" (I can do that), and the other answers, by making a gesture, and pointing to his bandaged leg, a 'conversation' did take place if albeit a silent one. It would be otiose to ask what was implicated here, since, hey, nothing was said.

Or if I make a gesture (as I can), 'are we drinking tonight?' and my partner makes the gesture, that he has a stomach ache (he can do that), I can infer that, no, he is not drinking (alcohol) tonight -- I CAN distinguish bewteen a gesture, 'let's get boozed' and 'let's drink water', or tea.

Again, I would think that the point of communication (the exchange of info, via the separate impartings of info by A and B) was fulfilled: communication TOOK place. But again, it would be otiose to say how much was implicated. All of it, or none of it. No words were uttered.

Similarly, if I make a gesture, as if to mean, "Do you have a light?" for a cigarette, and my partner answers with a gesture, with the middle finger. I can infer that what he implicates is NOT that he hasn't got a lighter but that perhaps, I shouldn't be, in the first place, smoking (or willing to smoke).

The middle finger carries indeterminate implicatures, one may argue. (While, rumour has it, what it explicates is "Pluck you").

----

By making the gesture, "No way I will light your cigarette" the utterer may be held to have meant, "And you shouldn't be smoking in the first place" but to decide to include (or not) THAT as part of what-is-meant (never mind 'implicated') depends on our willingness to ascribe to the utterer the intention to have us think that he does care for us via the recognitio of the same-self intention.

Sex works a lot like that, Ovid reminds us in "Ars Amatoria". I read the book with eagerness to check how Gricean Ovid can be. I was disappointed. He writes: "A man should keep the nails of his feet clean -- especially if he's wearing sandals in the theatre. A girl dislikes dirty feet". Give me a break!

(How can you fail an inference? By proving it fallacious. Or not).
Etc.