At A Bar In Texas With Grice

Rogers A & R Wall and JP Murphy, _Conversational Implicature_, Proceedings of Conference. Arlington: Virginia: Centre for Applied Linguistics.

Contents:

0. Editors's Intro

1. G Harman, Against universal semantic representation.

2. JD Mccawley, Remarks on the lexicography of performative verbs

3. JR Searle, A classification of illocutionary acts

4. JR Ross, Where to do things with words.

5. JM Sadock, Aspects of linguistic pragmatics.

6. RT Lakoff, What you can do with words: politeness, pragmatics, performatives.

7. GN Lakoff, Pragmatics in natural logic.

8. RC Stalnaker, Pragmatic presupposition.

9. L Karttunen, Presupposition and linguistic context.

10. RH Thomason, Where pragmatics fits in.

--- "Unfortunately, Grice's paper, "Desirability, Probability and Mode-Operators" was too crytptic for us to accomodate."

From the intro:

"Ross -- in his essay supported by the National Institute of Mental Health -- basically accepts the Gordon/Lakoff _transderivational_ approach to conversational implicature, a topic first extensively explored by Grice in _Logic & Conversation_.
"Ross's claim is that _SLIFTING_, which appears to be an ordinary syntactic rule, is governed in part by conversational implicature. Since conversational implicature is thought to be a pragmatic phenomenon, Ross is claiming that there are pragmatic conditions on syntactic rules".
"Ross argues that the kinds of _trans-derivational_ constraints postulated by Gordon/Lakoff to account for conversational implicature must be allowed to apply to embedded questions such as
(1) When woud dinner be, Archie-1 told Edith-2 that he-i wanted her-2 to
tell him-1.

"The analysis R. T. Lakoff offers is explicitly Gricean. She takes Grice's maxims of conversation to be rules of clarity, and offers her own rules as rules of _politeness_, rather. Thus, paralleling Grice's maxims of quantity, quality, relevance, and manner, she offers maxims of Formality ("don't impose. Remain aloof"), Hesitancy ("allow the addressee his options"), and Equality ("act as though you & the addressee were equal, make him feel good"). Her maxims, however, unlike Grice's, tend to be _mutually exclusive_, to apply in different types of social interaction."
"RT Lakoff notes that typical conversations do not adhere to Grice's maxims in a straightfoward way, but that, given that a conversational contribution does not _clearly_ conform to the maxims, the addressee can generally arrive at an interpretation of the contribution consistent with the maxims by employing Grice's cooperative principle."
"There is an objection which might be raised against both her work and Grice's, that it is non-empirical, both in that it is not based on systematic observation and that it is not predictive, hence, is _unfalsifiable_, though perhaps of heuristic value.".
Re: GN Lakoff's essay: "Finally, conversational implicatures of the Gricean sort are foreseen to be ultimately formalisable as _context-dependent entailments_ which are no different in principle from garden-variety _semantic_ entailments. This last suggestion is quite programmatic and leaves perhaps more room for doubt than do the other proposals as to the ultimate success of the undertaking. One serious difficulty that stands in the way is giving a precise account of terms such as 'cooperation' and 'relevance', which figure significantly in Grice's maxims of conversation. Lakoff suggests that the means for accomplishing this is by the techniques of 'natural logic' (G. Lakoff, 1972), a cluster of proposals which, it must be acknowledged, are themselves still highly programmatic and of unproved effectiveness in the face of so difficult an analytic task."

==

HARMAN:

"the theory of conversation (as part of a theory of performance) includes an account of implicature in Grice's sense. This theory of conversation is concerned with such things as implicatures and distinguishing these from what is, in a strict sense, _said_. Part of the theory will probably be universal, giving general characterisations, of implicature and the conversational strategies on which they depend."

==

ROSS's essay is great. He's a MIT man.

==

Three examples discussed by RT Lakoff

(1) a. We're not getting any younger
b. +> you look so old

(2) a. Margaret is so capable
b. +> i.e. ugly and undesirable,
but washes dishes nicely

(3) a. Her cakes are so _moist_.
b. +> they taste like cotton batting, but
_wet_ cotton batting.

Anyone familiar with American conversational ritual can translate, but the problem is that of course we may be dealing with a real ambiguity. I.e. not only has the utterer _SAID_ (a), he _MEANS_ (a). Grice has discussed such cases under the rubric of conversational implicature."
As Grice points out, the longer version is not equivalent to the shorter version, exactly. Her example is:

(4) a. Do you think you could perhaps lower your voice
a few hundred decibles?
b. shut up!

She notes that in high school one is taught not to begin a sentence with "I don't think" "because that showed that you weren't thinking",
"Your average English teacher showed that she had an excellent _inborn_ intuition about the rules of conversation and how they were to be applied".
"We have, fortunately, a preliminary undersatndin gof what we might call the Rules of Clarity, viz. Grice's rules of conversation which exist precisely for the reaason of communicative clarity. They dictate saying as much and no more than is necessary: saying what is true: what is relevant, and saying it in a non-confusing way. Grice's cooperative principle dictates that we strive to be unambiguous in verbal interaction; that, further, when there appears to be ambiguity, we assume there is something deeper going on. We might, in fact, want to refer to something like a _principle of mutual sanity_. i.e. each participant in a transaction assumes, unless given _strong_ reason to abandon this belief, that the other person is acting rationally. Maybe too we want to suggest an application here of the PLATONIC ART OF MEASUREMENT -- if you have to work really hard to understand something, you should get correspondingly more out of it than if the communcation were crystalline in the first place."
She notes parallel with "walking": "superfluous movements like arm-flailing and kicking are viewed strangely, and concealing your intention to turn a concern until the last moment will also throw your partner off".
"There always remains one grate problem with Grice: my relevance is your irrelevance, my necessity is your necessity, and so on."
===but some generalisation may be possible.
She notes that Americans may find Gricean maxims annoying.
"why can't he say what he means?" (p.96).
"why don't you get to the point, dammit"
She refers to a "shaggy dog story". Nice that Grice should consider shaggy dogs in his analysis of word-meaning (Studies in the Way of Words, Essay VI).
"I propose to view Grice's maxims as a subclass of _my_ maxims: Gricean
maxims are there to ensure dlarity, but clarity in a conversational contribution just means that the utterer will not be wasting the addressee's time. That is, he is not imposing on the addressee." untangling the entanglement: "the addressee can figure out WHY the utterer has made him go to the trouble on untangling the implicature, for certainly it would be unreasonable to make participants in a conversation go to special trouble to understand one another unless something were to be gained. And we can suggest that this _something_ may be, as Grice does not!"
"don't take any wooden nickles" is an American way of saying "bye-bye".
Why?

===

GN Lakoff:

"I would like to discuss one aspect of pragmatics that in recent years has been treated very differently: conversational implicature. Implicatures were taken by Grice to be by nature informal inferences of a fundamentally different kind than logical inferences, and hence _not_ to be dealt with by the apparatus of formal logic. In other papers I have dropped hints to the effect taht implciatures should be treated somewhat differently that they are in the Grice proposal."
"I would like to suggest (modestly) that implicatures are not 'loose' or informal inferences. Given my analysis, an implicature should turn out to be a species of semantic entailment, providing one had an adequate natural logic and an adequate analysis of the relevant principles of social interaction. Grice's theory of conversational implicature is based on the cooperative principle, the idea that certain 'maxims' are to be followed in conversational situations in which the participants are cooperating. Grice's maxims can be restated as principles like

(1) If x is cooperating with y, x wil do only what is relevant to the enterprise at hand, unless his actions make no difference to the enterprse. And if x is cooperation with y, x will _not to less than is necessary to make the enterprise successful. Anf if x is cooperation with y, x will not greatly exceed his needed contribution.

It seems to me that (1) follows from the meaning of "co-operate", and thus within natural logic."

Lakoff's example:
(2) If Sam asks Prof Snurd to write him a recommendation to graduate school, and Professor Snurd writes the recommendation saying only that Sam has nice handwriting, Sam will regret that Prof Snurd wrote him a bad recommendation.
"Now, the inference from saying only that a student has nice handwriting to giving a bad recommendation is a classic case of Gricean implicature.
And so (2) indicates that implicatures work like entailment-in-context with respect to cancellation.
"Probably the main objection to trying implicature via formal semantics is that implicatures are cancellable, while entailments are not".
His example is

(2) a. John has three children
b. John has three children- and he may even have 6.

"the function of the cancellation phrase is to limit the contexts appropriate for thus use of the entence to those in which the entailment does not hold". Lakoff considers

Conversation One:

A: we've got a job for a welfare recipient who has at
least 3 children -- and the more the better.
Do you definitely know someone who fills the bill?
B: John.

Conversation Two:

A: we've got a job for a junior executive with children,
but no more than three. Do you definitely know someone
who fills the bill?
B: John

Conversation Three:

A: Exactly how many children does John have?
B: John has three children

"If "John has three children" is a prely to "welfare-recipient" question, the implicature is cancellable, as shob by the fact that "John has three children, and he may even have 6" is a relevant and appropriate response. However, if "John has three children" is taken as a reply to "junior executive" or "exactly"question , the implicature is _not_ cancellable, as shown by the fact that adding "and he may even have six" is _not_ an acceptable response in these cases. The reason is fairly clear: the implicature is based on principle (1) about "cooperate". That principle
will be part of X in 'X U (P) ENTAILS Q. In this case, the relevant matter is whether it _matters_ that John has more than 3 children. If one case, it does not (the welfare recipient), in two cases, it does. My claim is that if this could be suitably formalised, the presence or absence of a conversational implicature could be handled using context-dependent entailment. Alas, no significant work has yet been done on the problem of formalising Gricean implicature".

====

STALNAKER's essay
example

(1) My cousin isn't a boy anymore
1. he's grown up
and +> he's male
2. the cousin changed sex.
and +> she hasn't necessarily _grown_ up.

He quotes Grice on "exploitation" of a maxim and the conversational/conventional implicature contrast. Stalnaker contributes to "Symposium on Grice", published in the Journal of Philosophy, chaired by J. F. Bennett.

===

THOMASON's essay

"Implicature has to do with the interpretation that language users give to signs when they are used in communcation. Grice's work provides the starting point here. It provides us with an organised view of a body of data, and with tools for classifying and interpreting it" Four examles for application of Grice:

1. clefting

1. John ate beans

2. What John ate was beans

2. nonrestrictive relative clauses and parentheticals.

1. John, who was hungry, ate beans,

2. John, I believe, ate beans.

3. demonstrative determiner

1. This elephant is not clean

*+> Either this is not an elephant or this is not clean

("though reference to an elephant is _implicated_ by an
utterance using "this elephant" it will not be part
of the semantic interpretation of "this elephant".

4. constraints of heterophora on the deletion of noun phrases

1. John said for Bill's father to go.

It would be a happy solution if we could have a theory
stating that utterances of the form "a said for b to go"
implicates that a is not the same as b.

"My problem then, is to develop a mathematical model of language use that will permit the explanation of a reasonable portion of Grice's phenomena". (Interestingly, Thomason develops this in "Accomodation" paper that Levinson does quote.
"Conversational implicatures are difficult to formalise. Yet, an account of implicature would make possible a much simpler semantics.

===
The refs section -- by organisers -- include

KASHER, conversational maxims and rationality,
Work by London School: R Kempson's and D Wilson's book on Presupposition
with respectively CUP and Academic Press.
WALKER in Blackburn on 'implicature'. Cited by Levinson, Pragmatics.
P. COLE, the synchronic and diachronic status of c. i.
S. Schmerling, assymetric conjunction and rules of conversation.
R. Wright, meaning and conversational implicature.
(the last three in Cole/Morgan)
OCHS, universality of conversational postulates. Now repr in Kasher,
_Implicature_ RKP.
KAHSER, A Mood implicatures. In Theoretical Linguistics I.
KARTUNNEN L Conventional implicature in Montague Grammar. Berkeley
Linguistics Society 1.
HORN LR Negative raising predicates. In CLS 11
PETERS S & KARTUNNEN L, what indirect questions conventionally impicate.
CLS 12.
PETERS S, Presupposition and conversation. Texas Linguistic Forum 2
GAZDAR J. Implicature. PhD Reading, Berkshire. University of Reading, England.
Later published as book with Academic.
Discussed by Levinson at length in section "Gazdar's bucket".
SEARLE, 'Meaning communication and representation'.
Later to be published in PGRICE, ed. R. Grandy, OUP.

==

Of course, Levinson's question still remains: to what extent is "metaphor" and other tropes not so much a question of _generalised implicature_ but the rather different topic of _particularised_ implicature, I can't tell! (and Rogers also quotes Fauconnier's early essay on Scalar predicates -- and doesn't that show some kind of _interface_ here?
JL
Grice Circle