Literary Stylistics:
Workshop Notes no. 23

Further Pragmatic Considerations:
Speech Acts and the Cooperative Principle

What's on this page

Speech Acts: Austin's Approach
        Further Terms
        Implicit & Explicit Performatives
        Appropriateness Conditions
Searle's Approach
        Searle's Speech Act Categories
        Table of Speech Acts
Grice's Cooperative Principle
        'Making Sense' in Conversational Interaction
        The Maxims: Table
        Maxim Overlap
        Conversational Implicatures
Breaking the Maxims
        Breaking the Maxim of Relation
        Who Judges the Break?
        The Indirect Addressee as Judge
Texts to Analyse

Speech Acts and Literary Works

Austin's Approach

As you may know, the first significant work on linguistic pragmatics was done by the philosopher J. L. Austin. It is to Austin that we must attribute the first systematic attempt to formally and clearly pin-point the shortcomings of formal semantics in the analysis of meaning in language.

What Austin initiated in the analysis of language, was the disjunction between the formal and functional (or preferably perhaps,[1] performative) approaches to the analysis of meaning, which we have already seen in workshop 21 in relation to clausal mood when viewed metaphorically. One may already notice that using the same terms imperative, declarative etc. for both grammar and pragmatics may be confusing. In relation to the classification of the intended or realised effects of utterances, John Searle has developed certain terms from the work of Austin, mentioned later in these notes, which we may find useful and less confusing.

Further Terms

Before we go to the classification of speech acts, a few more terms need to be explained, in order to help clear some confusion about their meaning, and to show how they relate to each other.

In later work in speech act analysis, the illocutionary and perlocutionary acts are virtually telescoped into a single concept: the illocutionary force. However, the illocutionary force does not seem to be a unitary concept, as there may be a disjunction between The fact that the illocutionary force conceived by the hearer is a reality, indicates that Austin's perlocutionary act is still applicable, although another term is now being used. Instead of the perlocutionary act, we now have In this connection, we may note that there may be more than one illocutionary force, whether intended or actual, for some utterances, and we will see this below, in relation to the imperative clause 'Don't do it'. In stylistic analysis, there is also the possibility that the actual illocutionary force registered by a character in response to another character's speech in a literary work, may be different from that registered by the reader, or, in the theatre, by a member of the audience.

Implicit and Explicit Performatives

One distinction Austin makes in relation to performatives, is that between

As performatives are seldom uttered using such a construction, it does seem to be the case that most of the performatives we encounter in the English language are implicit. Another and perhaps more frequently encountered term used for implicit performatives in speech act theory, is indirect speech acts. We may add here that the disjunction between intended and actual illocutionary force is more likely to occur with indirect speech acts.

Appropriateness Conditions

A performative, as noted earlier, does not depend on truth conditions for its meaning, but on certain appropriateness conditions, or what Austin calls felicity conditions. If, for example, a command cannot be issued by a particular person of lower status or power to another particular person of higher status or power, it will be infelicitous for the former to issue a command to the latter. A discrepancy between intended and actual illocutionary force may thus occur: for example, the person issuing the command may think that it is appropriate for him to issue the command, but the person to whom the command is directed, fails to see it as a command, or (even if he recognises it as an infelicitous command), refuses to obey it.

Searle's Approach

As I have said earlier, John Searle's classification of the various types of speech acts may help us see that the grammatical classification of clauses in terms of clausal mood may not be an adequate description of how these clauses actually function in discourse.

Searle's Speech Act Categories

As you know, Searle has classified speech acts into five categories:

  1. directives,
  2. commissives,
  3. expressives,
  4. representatives, and
  5. declarations (or declaratives, which we will not use, as the term may be confused with declaratives in grammar, which are defined constatively, and not performatively ).

Table of Speech Acts

The above terms are explained in the following table:

Table 23.1: John Searle's Classification of Speech Acts
If an utterance is:
a directive, the speaker[2] wants the listener to do something.
a commissive, she indicates that she herself will do something in future.
an expressive, she expresses her feelings or emotional response.
a representative, she expresses her belief about the truth of a proposition.
a declaration, her utterance results in a change in the external non-linguistic situation.

The discrepancy, and to a certain extent also, the consonance, between grammatical clause types and pragmatic utterance types, may be of significance in one's stylistic analysis.

[1] In order to distinguish it from the definition of the word functional in Halliday's book and elsewhere. Back to earlier position in the text
[2] For convenience of exposition, I treat the speaker as female here. Back to earlier position in the text

Back to index

Grice's Cooperative Principle

Introductory Notes

We dealt with speech acts above and in the previous workshop notes. We will be dealing today with the cooperative principle, which as some of you should know from your discourse analysis courses, is another pragmatic concept. I have explained that the constative approach to language would not give us an adequate picture of language in actual use, and that an analysis of speech acts could reveal facets of meaning which a more formal approach to semantics could not reveal. Another facet of meaning in language viewed in context will be revealed if we look at the cooperative principle, which as many of you know, was introduced by the philosopher H. P. Grice.

Making 'Sense' in Conversational Interaction

We may begin our observations on the cooperative principle by noting that in the language of conversational interaction, what has been said by each of the two (or more) parties in a conversational interaction does not always seem to 'follow' (in the logical or formal-linguistic sense) from what has gone before. Yet what is said by each speaker usually appears to 'make sense' to the listener(s). Part of the reason of course, may be due to ellipsis, which we have dealt with in the workshop notes on cohesion. The retrieval of the elliptical elements from elsewhere in the text does restore some logical sense to what has been said, and makes normal conversation more amenable to formal-linguistic analysis. But the analysis of ellipsis does not completely explain what goes on in conversational discourse, and there are other features of discourse which are not explicitly stated (including non-cohesive elliptical elements), and which cannot be retrieved by looking elsewhere in the text. By taking a different perspective, the Gricean approach, in addition to giving another plausible explanation for the speaker's resort to ellipsis and substitution, can account for at least some of the implicit features of conversational discourse which are not accounted for in the Hallidayan approach.

To Grice and his followers, one reason why what is said in conversational discourse 'makes sense', in spite of the missing implicit elements, is due to the cooperative principle, which usually operates between the speakers and listeners in conversational interaction.

The Conversational Maxims

According to Grice, cooperative behaviour in conversation can be described in terms of four conversational maxims:

Table 23.2: Grice's Maxims
1. the maxim of quantity, where one tries to be as informative as one possibly can, and gives as much information as is needed, and no more
2. the maxim of quality, where one tries to be truthful, and does not give information that is false or that is not supported by evidence
3. the maxim of relation, where one tries to be relevant, and says things that are pertinent to the discussion
4. the maxim of manner, when one tries to be as clear, as brief, and as orderly as one can in what one says, and where one avoids obscurity and ambiguity

Maxim Overlap

As the maxims stand, there may be an overlap, as regards the length of what one says, between the maxims of quantity and manner; this overlap can be explained (partially if not entirely) by thinking of the maxim of quantity (artificial though this approach may be) in terms of units of information. In other words, if the listener needs, let us say, five units of information from the speaker, but gets less, or more than the expected number, then the speaker is breaking the maxim of quantity. However, if the speaker gives the five required units of information, but is either too curt or long-winded in conveying them to the listener, then the maxim of manner is broken. The dividing line however, may be rather thin or unclear, and there are times when we may say that both the maxims of quantity and quality are broken by the same factors.


The reason why what one says makes sense in spite of some missing elements, is that these elements have been implicated, and these so-called implicatures are made possible by the cooperative behaviour of the speaker and listener. At one level, cooperative behaviour between the interactants means that the conversational maxims are being followed; at another and more important level however, cooperative behaviour still operates even if the conversational maxims are apparently broken. It is the latter level which usually gives rise to implicatures, and will be dealt with later in these notes. These implicatures may not be, strictly speaking, cohesive in the Hallidayan sense, because the text's phoric or ellipsed elements may not be physically present elsewhere in the text itself, or within its immediate or available environment, and cannot therefore be completely retrieved by looking at the text itself or its immediate or available contexts.

Conversational Implicatures

As these implicatures arise out of the observance or contravention of what Grice has called the conversational maxims, he has termed them conversational implicatures, which he distinguishes from conventional implicatures. As we are not doing a full-fledged course in semantics or pragmatics, we will not concern ourselves with how to distinguish conventional implicatures from entailment and presupposition, and with the exact dividing line between conventional implicatures and conversational implicatures. Grice himself does not seem to have clearly distinguished conventional implicatures from entailment and presupposition, and as an extension to this, it may be sufficient for us to note that conversational implicatures arise out of the breaking of the conversational maxims, whereas conventional implicatures are those which do not use the conversational maxims as a yardstick.

Back to index

Breaking the Maxims of the Cooperative Principle

It should be made very clear here that the breaking of any of the maxims of the cooperative principle does not mean that there is invariably a breakdown of communication. There may appear to be the apparent breaking of a maxim to the analyst, whereas the interactants may feel that the cooperative principle in general, or even the maxim itself, has not been broken. The breaking of a maxim may also involve some kind of trade-off, where one breaks a maxim in order to uphold another maxim, the latter of which may be more important in the communicative situation than the former. A breakdown of communication may not occur even if the listener perceives that the speaker has not been very cooperative with regard to a particular utterance, as the listener may have the opportunity to ask for confirmation or elaboration, thus repairing a communicative damage done by the earlier contravention. A useful differentiation of how the maxims can be broken has been made in the analysis of the cooperative principle, but before we go into that, let us ask two antecedent questions:

Breaking the Maxim of Relation

If a speaker wants to be cooperative with her listener,[1] the maxim of relation is probably the most difficult to break. Although it can be argued that communication of some kind can still go on if, for good reasons, any of the other maxims are broken, we cannot say that the cooperative principle is upheld at the points where the maxim of relation is broken. It has also been argued that the other maxims can, in varying degrees, be viewed in terms of the maxim of relation. For example, an undetected lie breaks the maxim of quality, as one is telling a deliberate falsehood, but in a sense, it may also break the maxim of relation, as the relevant utterance to the listener is an utterance which is propositionally true, and not one that is false. (This of course depends on the communicative situation; there may be cases, as we will see below, where, in spite of the breaking of the maxim of quality, it is more relevant to give a statement which is propositionally untrue than one which is truthful). Further, if five units of information are needed by the listener, but four or six units of information are given, then, either one relevant unit of information is not given, or one irrelevant unit of information is given, which indicates that both the maxims of quantity and relation may be broken. For these and other reasons, it has been argued, as you know, that much of (intentionally cooperative) interactive discourse can be understood in terms of only one underlying principle: the principle of relevance,[2] and the other maxims can be subsumed under this principle. The Gricean approach recognises the importance of the maxim of relation, but it does not subsume all the other maxims under it.

Who Judges When a Maxim is Broken?

Although we can say that it is always the speaker who breaks the maxims of the cooperative principle, the judge or arbiter of when or whether a particular maxim has been broken is not a predetermined factor. We can say that in some cases, it is the speaker who knows that a particular maxim has been broken; the speaker of an undetected lie, for instance, knows that she is not being cooperative with her listener(s). In other cases however, it is the listener or addressee who decides whether any maxim has been broken; indeed, we can say that it is the listener, or rather, the direct addressee (as he is normally the intended decoder of the message from the addresser), who is usually an important judge of whether a maxim has been broken. There are also instances when both the speaker and listener can be said to know that a maxim has been broken, and this does seem to be the ideal situation, where cooperative behaviour is fully maintained, in spite of the breaking of a maxim. Metaphorical uses of language, for example, can be described as breaking the maxim of quality, but both speaker and listener (unless we have a literal-minded listener) probably know that the literal truth is not a relevant factor here; in certain cases indeed, the metaphorical expression is more relevant than the literal truth (as when one gives bad or unpleasant news to sensitive listeners), in which case, one may be breaking the maxim of relation if one does not use the appropriate metaphorical expression. The maxim of relation may be crucial for the maintenance of cooperative behaviour in conversational discourse, and there may indeed be a breakdown or immediate termination of communication if, in order to uphold the maxim of quality, one delivers 'the unvarnished truth'. Given the importance of the maxim of relation, there may be instances when the speaker may perceive that she is being relevant to the discussion, but what is being said is not perceived as relevant by the listener. There may thus be a discrepancy between the speaker's and the listener's perceptions of relevance, which sometimes happens with egocentric speakers, or when the speaker does not know or is unable to gauge her listener's background knowledge, or when she talks 'at a different wavelength' from that of her listener.

The Indirect Addressee as Judge

To complicate matters, there may be a third party who judges whether a maxim has been broken, and this person, who is not directly addressed by the speaker, witnesses or overhears the conversation, often without the knowledge of the speaker or listener, or both of them. In the person of the reader (or a member of the audience in a theatrical performance) the indirect addressee here may be an important consideration in the analysis of conversation in literary works. We can see how a complication may arise when there are different perceptions of relevance between the reader on the one hand, and the speaker or listener (or both) on the other. But as I have said above, it is the direct addressee who, as the intended recipient of the message, is an important judge on whether a maxim has been broken. When analysing conversation in literary works, the reader often tries to determine, by looking at contextual cues available from the text, whether the direct addressee perceives that the conversational maxims have been upheld or broken. On occasions, the reader is also in an advantageous position to know that the speaker has consciously broken a maxim without the direct addressee knowing it. With regard to the speaker's perception of the breaking of the conversational maxims, if the speaker consciously breaks a maxim, and, in the spirit of cooperation, perceives or at least believes that the listener understands or will understand that the same maxim has been broken (and the listener in fact does perceive it accordingly), then the speaker can be described as flouting the maxims. But if the listener does not perceive that the speaker has broken a maxim when she has deliberately done so, or, when the listener perceives that the speaker has broken a maxim when the latter believes that she has followed it, then the speaker is described as having violated the maxim.

[1] For convenience of exposition, I treat the speaker here as female, and the listener as male. Back to earlier position in the text
[2] See for example, Dan Sperber and Deirdre Wilson, Relevance (Oxford: Blackwell, 1986). Back to earlier position in the text

For an early study of the use of Grice's Cooperative Principle for the analysis of literary works, see Michael Hancher's 'Grice's "Implicature" and Literary Interpretation: Background and Preface'.

Back to index

Texts to Analyse

Extract from D.H. Lawrence's Daughters of the Vicar.
Another text, in hard copy, will be analysed

Click here for the notes for the next workshop.

Click here for the notes for the previous workshop.

Back to Homepage: EL3221.

Last revised: 16 May 2017
Ismail S. Talib 1995-2017.