Grice: Breaking the Code

Levinson's book (_Presumptive Meanings, The Theory of Generalised Conversational Implicature_, MIT, 2000) has it controversial. Did Grice break the code? The point is between THE CODED & the UNCODED. Levinson being an anthropologist likes that, and has this picture

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.author .semantic .deictic & .minimal .enriched .additional .
. .representation .reference .proposition.proposition.proposition.
. . . assignment. . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.Grice . . .
. . WHAT IS SAID . IMPLICATURE .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.Levinson . the coded . implicature .
. . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

which amounts to, inter alia, the great old polemic between Levinson and,
say, Sperber (of Sperber & Wilson fame). Sperber (and Wilson) think it
(i.e. communication) s all mainly "inference", rather than "code" (This
they do in their 1996, _Relevance: Communication & Cognition_, Blackwell --
which is their recast of Grice's programme in terms of just one _principle_
("be relevant") understood as a cognitive inbuilt guideline to process
information). Levinson, being a complementarist, and functionalist
anthropologist at heart, thinks there's (always= room for "The Code".

Incidentally, I've noticed, Levinson quotes "Horn" more than he quotes
"Grice" -- if we go by the name index. Grice is quoted (exactly) 73 times
-- I hate to add "exactly" but what with Murphy's Zero-Herd of Zero-Cows, I
guess I must --, and Horn a lot more! (I can't count them!).

Incidentally, this reminds me of what Horn said of Levinson's book. I had
asked him, "HAVE YOU READ LEVINSON's BOOK" (implicating, "say something
about it"), and he said (something about it): "Yes, in fact I was a reader
for MIT Press for it [...]. I'll be teaching a seminar this spring going
through it in detail (and I actually went through the first three chapters
when I taught a seminar at the Illinois summer institute in '99, when the
book was still in galley form)."

So, it's a small world! I mean that thing about the Galleys! Levinson does
acknowledge the fact that Horn was a MIT press reader for the book, when,
in the acknowledgements section, he notes, "A number of scholars hve
commented on the manuscript. Special mention must be made of the rich
annotations I was lucky enough to receive of the whole manuscript from
Larry Horn -- in thin disguise as referre for MIT press."

Wonder about all the implicatures of "thin disguise"! Levinson does not
deal -- much -- with "The Coded" vs "The Non-Coded" in his book, though,
alas. But I noted the other day, on my 4th reading of the book, how I liked
his use of that little Latin word, "alas". Levinson is discussing anaphora
in English (and Tamil -- Levinson finds Tamil such a great language -- for
some reason (<- don't you hate that blank Gricean silly implicaturish
phrase, "for some reason"? I do!) and writes:

Pronouns in English do not permit determiners,
relative clauses, modifiers, or complements,
especially in the accusative. One cannot (alas)
in modern English, say thinks like

2. Him of the lion heart.

so additionally commentary will require a prolix head
noun, as in


3. Lasnik thinks that I admire the Lasnik of Essays on Anaphora,
while in fact I admire the Lasnik of Lectures on binding &
empty categories
(Manner-Implicature to Disjointness overriden by the
semantics/syntax)."

Must say I was warmed by Levinson's use of "alas" there, since it was a
welcoming contrast for me with replies I would often (more often than not)
get from Horn to the effect that "you" (JL?) just cannot say THAT". Horn is
ever-lacking "alas"! Levinson, on the other hand (on _Levinson_'s hand)
sees it's a weakness of the English _code_, as it were. The debate ensues.
Since Tapper mentions "Natural History of Negation", I recall that I
reported here my polemic with Horn re the adj. "unboring". Horn finds that
as quite unexistent -- and I'm glad _our_ Larry (Tapper) has bought
Larry-2's book on negation. As Larry-1 (Tapper) has pointed out, Larry-2
(Horn) is probably _wrong_ in think "unboring" rather unexistent, since,
Larry-1 found some usages of "unboring" via Google... I get, nevertheless,
Larry Tapper's point re the uttering of

4. Ba ba blah ba ba blah party

(Tamil for 1 = "Some of the guests have already left") before some American
monoglots. But surely, you cannot make the logic of language depend on the
fact that some Americans can't tell Tamil from Austronesian (the other
favourite with Levinson)! My point was _conceptual_! I meant to say: "the
code" is what it is because utterer and addressee _work_ on the
presupposition of a shared system of production and recovery (interpretive)
procedures (I am reminded of a PhD cited by Levinson on this: K. Welker,
Plans in the common ground: towards a generative account of conversational
implicature. PhD dissertation, Ohio State University, Columbus). There are
_two_ ways of looking at those "common ground" (as it were) procedures:

5. By uttering "ba ba blah ba ba blah party", the Tamil speaker meant
that some of the guests had already left.

5b.By uttering "ba ba blah ba ba blah party", The Tamil speaker
was relying on the fact that he & his intended recipient had the
procedure in their repertoires: to Use "ba ba blah ba ba blah party"
to mean that some of the guests had already left

It is this _having one's procedure in one's repertoire_ that amounts for
what Tapper and SR Brown have been calling here "core meaning", which I'd
prefer to call (some version of) (timeless, non-occasional) "expression
meaning". The issue is very tricky. For an analysis of "core meaning" in
terms of procedures I am reminded of R. Grandy/R. Warner (the "Richards" in
Grice's "Reply to Richards") in their intro ('Paul Grice: a view of his
work') to PGRICE (Philosophical Grounds of Rationality: Intentions,
Categories, Ends, OUP):

====INTERLUDE.
Grandy/Warner on Grice on _core_ meaning:

"Utterance type meaning. We are now ready to return to the concept of
utterer's type meaning. Grice offers an explication of utterance meaning in
terms of the notion of _having a procedure in one's repertoire_. Grice
takes this notion more or less for granted. he says:

"The idea seems to me to be intuitively fairly intelligible and to have
application outside the realm of linguistic, or otherwise communicative,
performance, though it could hardly be denied that it requires further
explication. A faintly eccentric lecturer might have in his repertoire the
following procedure: if he sees an attractive girl in his audience, to
pause for half a minute and then take a sedative. His having in his
repoertoire this procedure would _not_ be incompatible with his also having
this two further procedures: (a) if he sees an attractive girl, to put on a
pair of dark spectactles (instead of pausing to take a sedative); (b) to
pause to take a sedative when he sees in his audience _not_ an attractive
girl, but a particularly distinguished colleague" (_Logic & Conversation
VI:', _Studies in the Way of Words_, p.126).

Grice first uses thi notion to explicate meaning for _unstructured_
utterance-types. There are _nonsentential_ items, like flag signals. For
example, in yacht racing, a blue flag means that there are 10 minutes to
the start. The flag has no syntactic structure and no components that
themselves have a meaning that contribute to the meaning of the whole. What
is for such an _unstructured_ utterance type to mean something?

Grice answers this question by considering a _group_ (*actually Grice first
considers a single speaker and then turns to groups. We are simplifying for
ease of exposition) of utterers each of whom HAS IN HIS REPERTOIRE the
procedure of making a certain gesture (call it HW for hand-wave) if he
WANTS HIS AUDIENCE TO THINK he knows the route. [...] So if utterers in the
group have the H-W procedure in their repertoires, given that they all know
this -- H-W is a particularly effeciatious way of meaning that the utterer
knows the route. In such a case, Grice suggests that H-W means (_in the
sense of _utterance type meaning_ _) that the utterer knows the route.

But what of _sentences_? Sentences are _structured_ utterance-types. The
meaning of the whole sentence is consequent (in ways determined by
syntactic structure) on the meaning of its parts. Moreover, there are
_infintely_ many sentences. If an utterer is to associate a PROCEDURE with
EACH sentence of his language, he must have INFINTELY many procedures. If
he has to acquire them one by one, it will take him an infinite amount of
time. These considerations lead Grice to

"the notion of a 'resultant procedure': as a first approximation, one might
say that a procedure for an utterance-type X will be a _resultant_
procedure if it is determined by (its existence is is inferrable from) a
knowledge of procedures (a) for PARTICULAR utterance TYPES which are
elements in X, and (b) for any sequence of utterance-types which
exemplifies a PARTICULAR ordering of syntactic categories (a particular
syntactic form)" (Grice, op. cit, p.129).

The idea is that an utterer has a FINITE stock of _BASIC_ procedures. These
GENERATE an inifite set of RESULTANT procedures, including at least one
procedure for EACH SENTENCE of the utterer's language. So, for example, if
U is an English speaker, U will have the _resultant_ procedure of uttering

6. Snow is white.

if U wants his audience to think that snow is white.

"Let us assume a given relation S between sentences and propositions. Let
S(* + R) be the set of all propositions associated with any sentence with
the structure (* + R) [askerisk indicating mood, plus a radical or Harean
phrastic]*. (*Grice does _not_ proceed in this way. He does _not_ use the
notion of a _proposition_ in formulating resultant procedures). Now, where
p belongs to S(* + R), a resultant procedure of "* + R" takens the form:

7. Utterer has the resultant procedure of
uttering "* + R" if Utterer wants the
Addressee to think Utterer to think that
p.

"A complete theory of _sentence_ meaning would specify such procedures.
Such a theory would consist of a finite stock of basic procedures from
which one could derive infinitely many resultant procedures. As a
definition of STRUCTURED UTTERANCE-TYPE MEANING"

(core meaning?)

"we can say (provisionally, at least) that, where "p" belongs to "S(* + R),
"* +R" MEANS "p" in a group G iff members of G have the resultant procedure
of uttering "* + R" if the utterer wants his addressee to think that the
utterer thinks that p." Grandy & Warner do note that problems still remain!

--- end of excursus.

After this rather perhaps irrelevant excursus on "utterance-type" meaning
and its connection with utterer's meaning, must say I will have to analyse
once I figure out some resultant procedure RB Jones's remarks on modality.
Indeed, I guess I was using "doxastic" and "epistemic" quite
interchangeably to indicate some "essential" reference to the speaker's
state. Indeed, Jones is right that we must distinguish between the
implicatures of "possibly" from the implicatures of "<>". I guess the
Gricean point would be that "<>" models either the _core_ meaning of
"possibly", or the ceteris paribus standard utterer's meaning (the
explicature? -- I'm currently analysing this, since, with Levinson, I'm
quite sceptical that we need a notion of "explicature" as different from
"truth-conditions"). Levinson discusses quite a bit the implicature that
interests R. B. Jones, viz.

8. p & <> - p

Notably, I was refreshened to note that one problem with Noel
Burton-Roberts (in his "Modality & Implicature" (Ling & Philosophy, vol.
7)) was that _he_ (Burton-Roberts) thought that there was a Horn entailment
scale (after Horn's 1972 PhD):

6. <[]p, p>

but this assumption can be shown to be wrong. There is _no_ such scale.
I.e. "Nec" or "poss". just can't occur in an entailment scale along with
bare propositions. I will have to examine all this, though.

On the other hand (_Murphy_'s hand) -- Gosh we have a lot of hands here --
Murphy poses yet another interesting problem to the Gricean re: "some" cows
and, say, "three" cows. It's good that G. Koenig defends Grice in terms of
the Empty-Set Theory of General Set Theory... I will have to examine
Murphy's examples from an ordinary-language perspective, too.

I am fascinated to learn that, although Grice _does_ quote "(Ex)" and its
natural-language counterpart (two versions cited by Grice: "some", "at
least one") he does not further the topic. He _had_ done that as cited by
Strawson in _Intro to Logical Theory_ though. And yet, and much to Horn's
credit, Horn manages to write a whole PhD for UCLA about what Grice would
have said about this, and even before the 'Logic & Conversation' Lectures
were officially published!