Grice on "Not"

--- by JLS
------ for the GC

--- WE HAVE DISCUSSED PROFUSELY at this club the idea of 'not', mainly with regard to what J. Kennedy identifies as an Irish 'tic': "Don't you think he is always wrong?", rather than the more natural, "Do you think he is sometimes right?". Also in connection with what Kramer calls "Queen Anne is dead" "or is not dead".

Kramer was asking me to revise Gazdar's arguments. Believe it or not, I have a copy of Gazdar's book in my Swimming-Pool Library. I treasure it, because when I had to be evaluated for a course as a student, I cited from Gazdar, and the person who was grading me -- asked me for a copy of the book. (It happened to me often, that I had to instruct my instructors. In the case of providing copies of Grice, "Method in philosophical psychology" to instructors who had to grade me on philosophical psychology the thing was kind of amusing).

In any case, this is what Gazdar says about the four possible truth-functors of monadic appearance, like "not".

The others are T, P, and Q.

--- where he found the rubrics from escape me (i.e. go beyond my whatever).

We are concerned with Kramer especially the rationale for this. Kramer was saying that indeed, there is a rationale for the T, the P, and the Q, and not just 'not'. And I agree wholeheartedly with Kramer.

Gazdar is a linguist, not a philosopher. As such, they really don't care for philosophical arguments. They care for realizations in natural language, provided they are not just realisations in philosophers, even if native speakers. They are not ideal enough.

So this is what Gazdar rather dogmatically, authoritatively and unphilosophically says:

"There are four one-place truth-functors definable for a language having a bivalent semantics."

Their definitions are:

T N P Q
1 0 1 0
0 1 1 0

----

"Why is it that the only one which occurs in natural language is N?"

Why not? ie. Not why? The question is loaded already.

The sad is that the brings Grice onto the picture he wants to draw (This was a mere PhD dissertation under Palmer at a provincial university -- not far from Oxon, indeed, but provincial enough -- I love the provinces, don't get me wrong (* It was Berkshire).

Gazdar writes:

"Grice's maxims provide us with straightforward answers".

But Grice's maxims were UNPUBLISHED. Gazdar -- does he quote from --- where? Because he does have the decency NOT to quote from unpublished things by Grice. (Recall that Grice's 1967 William James lectures were only published posthumously in 1989 -- so that all of the Griceian influence was in areas where he'd rather not be seen cited at all).

"The maxim of manner," Gazdar goes on, ignoring or minimising the fact that this is AMUSING Grice. I.e. Grice amusing himself (or his-self) and his Harvardite audience back in 1967. No such thing as SERIOUS manner. This is the MODUS of Kant, as mistranslated as "Manner" by, of all people, Kemp Smith.

"motivates exclusion of T".

But a bit of undogmatic thinking along TRULY Griceian lines motivates again its inclusion. And never mind obscure lingos that no-one speaks. We's talkin' English! And I have some native speakers on my side, too!

"Because," Gazdar continues, Tphi iff phi, but Tphi is a LONGER expression than phi, use of Tphi in preference to phi would involve unnecessary prolixity, which violates the maxim."

He means the SUPER-maxim, 'be perspicuous' (sic -- see the amusement by Grice in calling a clarity maxim by reference to a very obscure way of naming clarity: perspicuous-being). Rather, it may be the maxim proper, "avoid unnecessary prolixity (sic)". But this is of course NOT violated since Grice dedicates the full last bit of lecture III to Strawson on 'it is true that...'.

"Furthermore," Gazdar goes on, P and Q are eliminated by the maxim of [relatio]." He uses 'relevance', but Grice would rather be seen talking French as a native speaker than using 'relevance' as the name of the maxim!!

Gazdar goes on: "Because Pphi iff Ppsi and Qphi iff Qpsi for ANY sentences phi and psi, it follows that the arguments of such functions are quite irrelevant to the truth valuation of sentences in which they appear."

Typical PhD dissertation talk. Impressive, i.e. meant to impress F. R. Palmer, a linguist whose claim to fame is to have written a short commercial intro to semantics where he cursorily eliminates Grice in one passage or two. Recall that Gazdar's thing was DISTRIBUTED before book form, and then, hocus pocus, got published by the very 'academic' "Academic Press", which is the one I have in front of my eyes -- and not out of the pleasure of reading it, but because my love for Grice!

I refer to Kramer's commentary for a truly Griceian (and Kramerian and Speranzian) account of P and Q (Minding your P's and Q's we could call it, echoing Robin Talmach -- married to Lakoff).

Gazdar, not happy with having impressed Palmer and been awarded the title of a doctor (PhD but this is a misnomer in that it's NOT in "philosophy") (I love Gazdar) goes on:

"If Grice's maxims capture universal principles of language usage".

-- or "universal principles of language usage at Oxford", as I rewrote the thing in the ch. viii of my philosophy thesis, "The cunning of conversational reason", the title of that chapter --

Gazdar continues:

"then it is is HARDLY surprising that NO languages [that Gazdar speaks -- they exclude English, as thought out by people like Grice, Kramer, and me] have truth-functional operators which would violate those maxims whenever they are used."

--- And the circus? Can't language be used at the circus too? I mean, Gazdar is supposed to have brought a 'fresh air' to linguistics: hey, the man was FORMALISING pragmatics! But in the proceedings, he is violating Grice! For surely the linguistic idiosyncrasies of one, say, H. Corey HAVE to be taken into account. If you CAN Think of things like "NAND" and "NALL" then these operators DO Exist. Ditto for T, P, and Q. They are not inconceivable. Perhaps THAT's the big mistake that linguists commit, from the more perspicuous vision of the philosopher. For the philosopher (and Grice was a philosopher, not a linguist) argues (rationally) about the CONCEIVABLE. If Grice or I had wanted to study linguistics, imagine the bore! Grice quoting from Firth!

--- Anyway, Gazdar goes on:

"Use of N will NOT in general violate any of Grice's maxims"

My! What about "That man over there is NOT lighting a cigarette with a 20-dollar bill". THAT does violate the maxims alright. And yet it is a perfectly well-formed logical operator and attending formula. So what is Gazdar talking about? (Of course we know, but we are playing the philosophical devil's advocate -- which Grice needs most when discussed by linguists).

--- Gazdar continues:

"Besides this negative advantage, N can be used to define other operators and connectives in a way which T does not permit."

This IS a good point.

"Thus N can be used to define T (Tphi iff NNphi), but the converse is not possible."

Good point. No iteration of Queen Anne can yield "NOT" dead, or something. So what? Who says that the virtue of x is that it allows to define y? Imagine if we say "Jupiter is less important than Mars, because from Mars you can travel to Jupiter, but not vice versa". Fallacious. Etc.

Gazdar goes on:

"Furhermore, if we add T to a language containing just conjunction and inclusive disjunction, then we are no better off. But if we add N to such a language, then we can define all the other connectives and thus increase the expressive power of the language."

Again, this begs on the question that definability of one operator in terms of another is a rational requsite. Indeed, quite the contrary. If definability were SUCH a rational requisite, people would NOT be introducing otiose operators which can be defined elsewhere in the FIRST place. It's not like Grice's analysis of 'mean' in terms of 'intend', say. I can teach someone the meaning of 'mean' by explaining how 'mean' is constructed out of 'intend'. But it would be stupid of me to teach someone the meaning of 'if' by relying on 'not'.

---

Gazdar finishes this rather inconclusive section thus: "This fact makes N invaluable in such languages."

Not for Jenny:

"Jenny made her mind up when she was twelve,
into furrin languages she would delve,
but at seventeen at Vassar it was quite a blow,
For in twenty four languages she couldn't say 'no'"

---- Gershwin.

--- Gazdar finishes it off this section of his Reading MS: "If the arguments to be given later can be maintained," -- typical PhD phrasing: it's all about the argument and how things can be maintained till you are awarded the degree -- never mind afterwards! Note that Gazdar never again publish ONE thing in formal pragmatics -- "then it may well be that natural languages [sic in plural -- i.e. not logic, not Griceish, not anything resembling a rational calculus alla Leibniz and Whithead and Russell that Grice is defending --] are of just this type."

The type that typifies "not"!