lin1280 pragmatics i - um.edu.mt · cooperation. (pete isn’t ... this gives rise to the...

26
LIN1280 PRAGMATICS I Conversational implicature

Upload: vokien

Post on 28-Apr-2018

216 views

Category:

Documents


1 download

TRANSCRIPT

LIN1280 PRAGMATICS IConversational implicature

COOPERATIVITY

We’ve outlined Grice’s maxims of conversation,

and emphasised that:

these are not “laws” imposed upon speakers, but

they are “conventions” that seem to capture our

assumptions when we communicate.

So, in conversation, the basic assumption is that

participants are trying to adhere to some kind of

cooperativity.

COOPERATIVITY AND IMPLICATURE

If we assume cooperativity, then violations of this

principle can give rise to inferences over and

above what is literally said.

(Why is my interlocutor violating Quality or

Quantity? Why is she saying something apparently

irrelevant? Maybe it is relevant after all...)

The kinds of inferences are called

conversational implicatures.

They arise in response to apparent violations of

the Maxims.

AN EXAMPLE

Jane: Did you get the car washed, the radio mended and the internet account renewed?

Pete: Well, I got the car washed.

Pete isn’t answering the whole question.

This gives rise to the inference that what is not being answered is in fact answerable in the negative.

Form:

Jane: A & B & C?

Pete: A (Implicature: therefore, not B and not C)

PART 1Generalised conversational implicatures

AN EXAMPLE

Jane: Did you get the car washed, the radio mended and the internet account renewed?

Pete: Well, I got the car washed.

Important things to note:

Jane’s conclusion that Pete didn’t do certain things is not directly based on what Pete literally says.

Pete in fact says nothing about the radio or the internet account.

Implicatures arise as a result of inferences by speakers. They are intended to preserve the assumption of cooperation.

(Pete isn’t refusing to cooperate. He is leaving things unstated because Jane will be able to infer what he means.)

GENERALISED CONVERSATIONAL

IMPLICATURE

The example we’ve seen is in fact much more general.

It seems that exchanges of the same form tend to give rise to the same sorts of implicatures.

Cat: Did you talk to Jake and Mike? (A & B?)

Jean: I talked to Jake. (A) Therefore: not B

Implicatures like these, which do not rely on specific contextual knowledge, are referred to as generalised.

ANOTHER EXAMPLE

A quite robust conversational implicature in English

arises with the use of an indefinite NP:

“a / an X” is often interpreted as “an X which doesn’t belong

to the speaker”

I was sitting in the living room when a child walked in.

Assumption: it’s not the speaker’s child, as otherwise he’d have

said my child.

If the speaker does mean my child, then saying a child is

uncooperative (he’s giving too little information).

The same seems to happen in Maltese with wieħed:

Kont fil-kċina meta daħal wieħed tifel.

Compare: Kont fil-kċina meta daħal it-tifel

PART 2Scalar implicatures

SCALES

Scales are pervasive in language.

Whenever we communicate quantities,

frequencies etc, we select from a scale:

all > most > many > some > few

always > often > sometimes

Note that languages make available a whole

range of expressions to communicate such

quantities and frequencies.

SCALES AND COOPERATIVITY

When a speaker selects a particular expression from a scale, the assumption of cooperativity means that we interpret their choice as the most “truthful” or “accurate”.

Example:

all > most > many > some > few

I’ve finished some of my work.

Observation:

If I’ve finished all my work, it’s technically true to say I’ve finished some of it.

But this may well appear misleading (why say some when we can say all?)

SCALAR IMPLICATURE

Scalar implicatures arise when:

A speaker uses an expression that involves a choice from a scale.

The choice is interpreted as being the most accurate, or “closest to the facts”

Under cooperativity, the choice gives rise to the inference that the other possible choices are not as accurate.

In general:

If there is a scale, and a speaker chooses some expression X on that scale, then: This gives rise to the implicature that higher values on the

scale do not apply.

SCALAR IMPLICATURE

Example:

Scale: all > most > many/much > some > few/little

I’ve finished some of my work.

Therefore: I haven’t finished most of my work

And: I haven’t finished all of my work.

Etc.

Note: use of some gives rise to the implicature of

not all.

SCALAR IMPLICATURE

Example:

Scale: always > most of the time > sometimes > never

Linguistics lectures are sometimes interesting.

Therefore: they’re not always interesting (or even

most of the time!)

Note: use of sometimes gives rise to the

implicature of not always etc.

OTHER SCALES

Scales seem to exist for things which aren’t really

quantities:

Scales of obligation:

Scale: must > should > can

This should be red

(implicature: it’s not the case that it must be red)

Scales of possibility:

Scale: possibly > probably > definitely

Possibly, they will be on time.

(implicature: it’s not probable, or definite)

PART 3Particularised conversational implicature

PARTICULARISED VS GENERALISED

So far, we’ve seen examples where implicatures

don’t seem to depend on context.

In some cases, they do rely on the topic of the

conversation and other features of the (linguistic

and non-linguistic) context.

EXAMPLE

Ron: Did you go to the pub last night?

Mary: My grandfather died.

Mary’s utterance seems to violate the Maxim of

Manner (Be Relevant). But it doesn’t.

In this particular context, it is answering Ron’s

question in the negative (No, I didn’t go to the

pub) and giving the reason for the answer.

EXAMPLE

Ron: Did you go to the pub last night?

Mary: My grandfather died.

Notice that in order for the implicature to go

through, one needs to make certain assumptions,

for example:

Death in the family usually gives rise to an

emergency that needs to be dealt with, and usually

overrides any desire for enjoyment.

These assumptions are particular to the context

(cultural and conversational).

A WORD ON IMPLICATURE GENERALLY

Both generalised and particularised implicatures rely on what is communicated by speakers, but not said. (They are not entailments)

They always rely on an assumption of cooperativity.

They are usually seen in conversation (not monologue, for example).

They can be denied: I never said that, did I?

But they can also be reinforced: You’ve got ten minutes to live. Well, in fact you’ve got

several years.

PART 4Conventional implicatures

CONVENTIONAL IMPLICATURE

Unlike the forms of implicature we’ve seen so far,

conventional ones can’t be denied or reinforced.

They don’t have to occur in a conversation and don’t

need an assumption of cooperativity to be understood.

They don’t depend on context, conversational, cultural

or otherwise.

Rather, they always rely on the use of a particular

expression, which is conventionally associated with a

given meaning, irrespective of speaker or context.

SOME INTERESTING EXPRESSIONS

The use of but usually carries an implicature of

contrast:

I suggested pink but she chose black.

But is a conjunction, so we have something like:

p but q

where both p and q are true

But tthere’s also an additional implicature that q

contrasts in some way with p

SOME INTERESTING EXPRESSIONS

The use of even usually carries an implicature of

violation of expectation:

You should’ve come last night. Even Ted came!

With even, there’s a suggestion of something that

is contrary to what one might have expected in

normal circumstances.

Note: this doesn’t depend on a speaker or a

context, it’s a convention that has developed

surrounding the use of this expression.

SOME INTERESTING EXPRESSIONS

The use of yet usually suggests that the present

situation is expected to change to something

different later on.

I haven’t murdered anyone yet.

(Implicature: I’m likely to murder someone shortly)

SUMMARY

Implicature is one of the most central concepts in

pragmatics.

It primarily has to do with how meaning is

conveyed over and above what is said.

With the exception of conventionalised

implicatures, such inferences usually depend on:

The dynamics of conversation

The assumption of cooperativity among speakers.