Analysing Talk In Interaction
Lectures 1 2 3 4 5 6 7 8 9 10 11 12 13
Charles Antaki
Loughborough University

Lecture / Seminar 2: Pragmatics of language II: Grice's Principle of Co-operation


You will find all you need under the entry for "Grice' or 'the co-operative principle' in any pragmatics textbook (see Readings, above).

And as before, if you'd like to read a specialist account, the best is in Levinson, S. (1983) Pragmatics. CUP.


Last week we wondered about how words do things, which led us to understand that the idea was helpful, though it seemed idealised and a long way from actual talk-in-interaction.

Today we'll look at a fascinating insight from a philosopher, Paul Grice, meant to explain some absolutely fundamental things about what happens when people speak to each other in real encounters.

We shall see how people use certain regularities about spoken language in interaction to help them make sense of what's being said (and what it means, and how one should reply).

Intro: Exploiting a basic principle

Perhaps the most basic principle of actual exchanges of talk is that what is said at any moment is inevitably coloured by what has gone just before (even if what is before is a silence, a scream, or anything at all). Take this example:

Speaker B: The dog looks happy.

Now on its own, that's a statement about the dog. But what did Speaker A, say, immediately beforehand? What if we discovered that she said:

Speaker A: Where's the roast beef?

Then we have a different sense of the force of the observation about the dog. That's the kind of thing I want to describe in this lecture: the relationship between the things that two people say. The basic human contract we have with each other as competent members of society is that what I say is somehow related to what you've just said, and we'll see what that relation is like.

The first person to have a go at putting the relationship on a systematic footing wasn't a psychologist, interestingly enough (in fact nobody I'm going to mention in this lecture is a psychologist, strictly speaking), but a philosopher, Paul Grice. The aspect of his theory that most interests us is hs notion that people were bound to what he called a general Co-operative Principle.


The Co-operative Principle: Be relevant, clear, truthful and say as much as is necessary
The co-operative principle is to speak in such a way as it helps move things along. It comes out in four main ways. People try to to be relevant; to say (only) as much as is necessary; to say what is true; and to be clear.

Who would argue with any of that? No-one. But they are not meant to be prescriptions like the ones you see in etiquette books ('speak clearly and be courteous at all times.."). They are mean to reveal what the listener can assume about the speaker's intentions. Only by making those assumptions can talk be understood that would otherwise be unintelligible. Let me show you what I mean.

Maxim of Relevance: Make what you say bear on the issue at hand
Let's work on that snatch of (invented) talk I mentioned above:

Jim: Where's the roast beef?
Mary: The dog looks happy.
Any competent speaker knows that Mary means something like "In answer to your question, the dinner has been eaten by the dog". Of course, she doesn't say that - we work it out on the basis, first, that what she says is relevant to what she's been asked. If she's mentioning the dog, then the dog must be some kind of answer. This is perhaps the most utterly indispensable and foundational assumption we make about the talk we hear - that it's relevant to what has immediately gone before.

It's hard to overstate the case, or to say how much for granted we take it. As listeners, we will work very, very hard (but unconsciously) to find a relevance in even the most apparently unconnected utterances, so long as one follows the other.

Use the table below to generate a few random pairs of utterances. I think that even when the result is odd or daft you'll be able to find a meaningful relation between the two utterances (though sometimes you'll have to work at it):

Choose any one from here Then any one from here

Speaker A:

Speaker B:

What a beautiful coat
How about a cup of tea?
It's raining
Great news about Pat
Did I ever tell you about my crazy uncle?
I'm in the mood for love.

My head hurts
I'll be back around ten thirty
I'd rather be in Torremolinos
What do you know about it?
Don't you ever think about anything else?
Not round here.

Anyone who overheard an exchange like A: "Great news about Pat" / B: "Don't you ever think about anything else?" would automatically make sense of it on the assumption that B was being co-operative - that they said what they said relevantly. In this case, it might be that B is accusing A of always going on about Pat, is obsessed by Pat and so on. Or if the random pair was A: "How about a cup of tea" and B: "Not round here", we'd assume that we were overhearing a couple discussing a break from shopping and rejecting the local cafes, and so on. We can always make some sort of sense out of two utterances on the basis that they are relevant to each other.

Imagine life without that basic assumption. Talk would be extremely tedious, because Speaker B would always have to say "In response to what you just said...."

People work hard to ignore the apparent un-cooperativeness of the next utterance and to find something relevant, clear, truthful etcetera about it. That means that we can work miracles of implication, and get meanings across very laconically and economically, like the 'roast beef' example above. Let's see the same thing in a couple of other maxims, keeping an eye out to see how people can exploit the maxims to lead others to draw conclusions.

Maxim of Quantity: Say as much as is helpful

The maxim is "say as much as is helpful but no more and no less". Let's use it to see how you can misdirect someone by playing on the fact that they will assume you are saying as much as is needed. If someone asked you

Where did you go yesterday?
and you said
Odense train station
Then they would automatically assume that you went to no less and no more a place than the train station (to buy a railcard, or to meet someone or whatever). If they later discovered that in fact you'd then gotten on the train and gone to Copenhagen and spent the day shopping, they'd be entitled to feel surprised and even perhaps a bit cheated. You had been 'economical with the truth', in the sense that you had disobeyed the maxim of quantity. Every listener assumes that the next speaker will say enough - not too much, but certainly not too little (even though what little they said, like "Odense train station", would be 'true').

Maxim of Manner: Be concise, to the point, etc

Let's use this maxim to see how a speaker can make listeners draw quite extensive implications by the way they flagrantly go against, or flout, as Grice called it, the maxims. Suppose you overheard two parents say to each other (the example is from Levinson, p 104; see ref above)

A: Let's get the kids something

B: OK but not I-C-E C-R-E-A-M [spelling it out]

B is going out of their way to be a bit obscure, spelling out the words rather than simply saying them (or B might have said 'pas des glaces pour les enfants' or some such). B is utterly failing to co-operatively follow the maxim of clarity and conciseness. B is being so flagrant that A can infer that there must be a special reason for being so uncooperative: the likely inference, of course, is that B doesn't want the kids to complain that they're being denied a treat.

Maxim of Quality: Be as truthful as is appropriate

The hearer assumes that the speaker is not knowingly telling a lie or fantasising. Fair enough, and maybe routinely that is indeed the case. What is more interesting is what happens when someone flouts that maxim (as we saw in the case above). Suppose the conversation went like this:

A: I might win the lottery

B: Yes, and pigs might fly.

The obviousness of the 'untruth' of B's reply gives our cognitive system a huge nudge. B is flouting the maxim of quality, so there must be something else going on, and so we start a hunt for likely inferences we can make. Here. of course, we quickly settle on the implication that A's chances of winning the lottery are about the same as pigs flying. Flouting the maxim of quality is the driving force in irony - try thinking of ironic comments you've heard recently, or generate some witty repartee of your own (hard to do just like that, I know) and see how they achieve their ends by what they do to expectations of 'truth'.

This week's seminar: Working with the co-operative principle.

We will discuss the idea of 'flouting' the principles, based on your readings of relevant material in the introductory texts.

For next week's seminar:

Start to look into Conversation Analysis. It would be helpful if you read the first chapter of Hutchby and Wooffitt's book 'Conversation Analysis' (see Readings).

back to top

A course for the University of Southern Denmark, Odense 2003