Taylor and Francis Group is part of the Academic Publishing Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.


Unit 5. Implicature: (re)assessing the utility of the CP

In A5.1.3, we outlined the CP and its attendant Maxims (Quality, Quantity, Relation, Manner), and went on to detail the different ways of breaking these Maxims (in A5.1.4). In A5.2, we highlighted some of the issues that Grice himself raised in regard to the CP: they included his recognition that, potentially, there are ‘all sorts of other maxims ... such as “Be polite”’ (Grice 1975: 47). The remainder of A5 was then devoted to a discussion of how Neo-Griceans and Post-Griceans have sought to modify – and, in some cases, revolutionize – the Gricean approach to implicature, as a means of eradicating some of its ‘inherent problems’.

In this section, we want you to have the chance to test the utility of the CP for yourself. We begin by asking you whether Gricean implicature is indicative of linguistic or social cooperation. We then provide examples of how researchers have used the CP (and its Maxims) in conjunction with another procedural concept – specifically, activity types(Levinson 1979, 1992), metapragmatic framing strategies (Janney 2007) and reality paradigms (Archer 2002) – in order to be able to explain conflictive interaction: and you are given opportunities to test these methods.


  • Archer, D. (2002) ‘“Can innocent people be guilty?” A sociopragmatic analysis of examination transcripts from the Salem Witchcraft Trials’, Journal of Historical Pragmatics 3(1): 1–30
  • Grice, H.P. (1975) ‘Logic and conversation’, in P. Cole and J. Morgan (eds) Syntax and semantics 3: speech acts, pp. 41–58. New York: Academic Press. Reprinted in Grice 1989: 22–57
  • Janney, R.W. (2007) ‘“So your story now is that...”: Metapragmatic framing strategies in courtroom interrogation’. In: W. Bublitz and A. Hübler (eds.) Metapragmatics in Use, pp223–34. Amsterdam and Philadelphia: John Benjamins
  • Levinson, S.C. (1979) ‘Activity types and language’ Linguistics 17(5-6): 356–99
  • Levinson, S.C. (1992) ‘Activity types and language’ in P. Drew and  J. Heritage (eds.) Talk at Work: Interaction in Insitutional Settings, Cambridge: Cambridge University Press, pp66–100

5.1. Gricean implicature – indicative of linguistic or social cooperation?

Our focus on conflictive interaction necessitates that we also reintroduce here an issue that we briefly outlined in C6.3: whether the CP, as Grice defined it, was meant to capture linguistic cooperation or social cooperation.

Our position is similar to that of Bousfield (2008: 27), namely that Grice had in mind linguisticcooperation, not least because:

Immediately following his own definition of the CP; the maxims of Quantity, Quality, Relation and Manner, Grice says that, ‘There are, of course, all sorts of other maxims (aesthetic, social or moral in character)’ (Grice 1975: 47...). Now, it could be argued that had Grice intended his CP to be a model of social cooperation (and his maxims, therefore, as being socially directed maxims), then he would not have indicated ‘social’ maxims as being an ‘other’ type of maxim to the ones he himself had just stipulated for the CP; its categories and subordinate maxims.

What ‘really confirms Grice’s position’, for Bousfield (2008: 27), however, is Grice’s explicit indication that he considered – but then abandoned – the idea of the CP as a possible system of ‘social goal sharing’. Here, Bousfield has in mind Grice’s explanation of the way in which our expectations for social cooperation can grow as an interaction develops – to the point of becoming ‘quasi-contractual’. Yet, Grice nonetheless signalled that, even when a common aim might be evident, ‘like getting a car mended’, it is still possible that ‘the ultimate aims’ of the participants ‘may ... be independent and even in conflict – each may want to get the car mended in order to drive off, leaving the other stranded’, for example (Grice 1989: 29).

Mooney (2004: 902) also believes the CP can be ‘at work’ in socially uncooperative/conflictive contexts. But she believes the CP could be more powerful, interpretatively speaking, if ‘cooperation’ was linked much more explicitly to the notion of activity types (Levinson 1979, 1992). By way of illustration:

the activity type that is a cross-examination tells participants what is and what is not required and allowed. It is also embedded in the conventions of the courtroom ... Using activity types forces the analyst to reconsider the goals of [such] discourse activities ... The benefit of adopting the activity type in conjunction with the CP is [that i]t contextualizes the maxims and allows a clearer and more relevant construal of contributions. While it is not predictive in a detailed way, it certainly serves to make Gricean analysis more powerful and explicable ... (Mooney 2004: 905)


According to Mooney (2004: 905), ‘until one knows the “structural properties of an activity” one cannot understand the way in which these properties “constrain ... the verbal contributions that are made” (Levinson 1979: 370)’. As Mooney’s reference to Levinson makes clear, she is alluding to his definition of an activity type as:

a fuzzy category whose focal members are goal-defined, socially constituted, bounded, events with constraints on participants, setting, and so on, but above all on the kinds of allowable contributions. (1979: 370, 1992: 69)

Investigate a courtroom transcript1 as a means of familiarizing yourself with the structural properties of this activity type, noting in particular what types of verbal contributions are allowed, by whom they are used, and the goals of these different participants.

In A5.1.4 and A12.5, we suggested that the Quality Maxim is probably suspended in an interrogation/police interview. Do you think this is also true of the courtroom? Or is it more likely that participants opt to signal their inability to cooperate in the way the Quality Maxim ideally requires (cf. Grice’s definition for opt outs and/or clash(es))? Where possible, draw from your courtroom transcript to justify your answer.

How might the Quantity, Manner and Relation Maxims be affected by the courtroom procedure? Again, where possible, draw from your courtroom transcript to justify your answer.


  • Bousfield, D. (2008) Impoliteness in Interaction. Amsterdam and Philadelphia: John Benjamins (especially Section, ‘Grice: As social cooperation or linguistic cooperation?’ pp25–9)
  • Grice, H.P. (1989) Studies in the way of words. Cambridge, MA: Harvard University Press
  • Levinson, S.C. (1979) ‘Activity types and language’ Linguistics 17(5-6): 356–99
  • Levinson, S.C. (1992) ‘Activity types and language’ in Drew, P. and J. Heritage (eds.) Talk at Work: Interaction in Insitutional Settings, Cambridge: Cambridge University Press, pp66–100
  • Mooney, A. (2004) ‘Co-operation, violations and making sense’, Journal of Pragmatics 36: 899–920

1 Some American trials are available in full via the internet. In the UK it is more common for public inquiries to be made publicly available. The following link, for example, provides a range of data relating to the Shipman Inquiry: http://www.shipman-inquiry.org.uk

5.2. Gricean OPT OUTs and the social implication(s) they generate

As we highlighted in A5.1.4, an opt out ‘involves S explicitly indicating his or her unwillingness to cooperate in the way the maxims require’. As Grice (1989: 49) made clear, it is possible to ‘OPT OUT from the operation both of the maxim[s] and of the CP’. Opting out of the CP altogether equates to linguistic non-cooperation: an example would be not offering any response to a question, when a response is expected. In contrast, if someone states I cannot say any more: my lips are sealed, they are potentially opting out of the Quantity Maxim only (albeit by flouting the Quality Maxim; i.e. by stating that they cannot speak because of their sealed lips). According to Mooney (2004: 912), when ‘the CP is opted out of altogether’ a ‘social implication’ willbe generated nonetheless:

Social implication tells us something about the relationship between interlocutors. This relationship is important as different kinds of relationships support different ways of being cooperative. (ibid.: 909)

Mooney’s inclusion of social implication alongside linguistically generated implicatures suggests that we should perhaps pay as much attention to how ‘the CP is followed’ – and not followed – in a given activity type as we do to determining whether or not ‘it has been followed’ (ibid.: 909).


According to Grice (1989: 30), a speaker ‘may say, indicate or allow it to become plain that he is unwilling to co-operate in the way’ a maxim or maxims require. Below we’ve provided some typical opt outs of the Relation, Quantity and Manner Maxims:

  •  ‘I know I’m changing the topic, but...’
  •  ‘As you all know...’; ‘I cannot say any more...’
  • ‘This may be a bit obscure...’

Notice that, in each case, S makes explicit his unwillingness to adhere to the maxim in question.

Archer (2002: 19) found the following example of an opt out in a transcript taken from the Salem Witchcraft Trials (1692), which is less overt than the above. A witness, Martha Corey, was advised to ‘tell ... why [she] hurt these persons’ at the beginning of her examination. Rather than responding, however, she insisted that she be allowed ‘to goe to prayer’ – and she continued to make this request ‘sundry times’, to the extent that her questioner (Hathorne) was forced ‘to reinstate his own agenda by asserting “We do not send for you to go to prayer But tell me why you hurt these!”’

  • Archer argues that the above example constitutes both social and linguistic (i.e. Gricean) non-cooperation. Do you agree?
  • Examine some modern courtroom transcripts to determine whether opt outs are used in today’s courtroom, and, where they do occur, whether the opt out is signalled explicitly or ‘allow[ed] to become plain’ by some other means (see Grice ibid.: 30). Also pay attention to the type(s) of response they prompted, as a means of assessing their social implication (see Mooney 2004).


  • Grice, H.P. (1989) Studies in the way of words. Cambridge, MA: Harvard University Press
  • Mooney, A. (2004) ‘Co-operation, violations and making sense’, Journal of Pragmatics 36: 899–920
  • Archer, D. (2002) ‘“Can innocent people be guilty?” A sociopragmatic analysis of examination transcripts from the Salem Witchcraft Trials’, Journal of Historical Pragmatics 3(1): 1–30

5.3. Metapragmatic framing strategies and the Gricean Maxims

Janney (2007: 223) uses the term ‘Metapragmatic Framing Strategies’ (henceforth MFSs) to describe the ‘strategies used by the prosecutor’ in a cross-examination context so that they can ‘frame questions in ways that steer the defendant’s answers while at the same time commenting on their adequacy, clarity, relevance, or truthfulness’. Janney goes on to explain that these strategies are metapragmatic in two senses:

on the surface, they often appear to be ostensibly directed to regulating communication between prosecutor and defendant. We tend to perceive them as procedural requests, directives, reminders, reprimands, objections, etc. ... at a deeper level, their goal is to steer or ‘manage’ the jury’s perceptions of the defendant’s testimony ... (ibid.: 224)

Janney further claims that the US jury trial in particular presupposes that, when giving their evidence, defendants are expected to follow a kind of legal CP (i.e. they are expected to be clear, relevant and truthful). Not surprisingly:

the four maxims of the CP (quantity: don’t be more or less informative than necessary; quality: don’t lie or say things for which you have no evidence; relevance: don’t say things that are not relevant; manner: don’t say things that are ambiguous or obscure) all prove to come into play as criteria for judging the extent to which the defendant is, or is not, being [legally] cooperative.

Drawing from transcripts relating to the 1997 civil trial of O.J. Simpson, Janney demonstrates how the prosecution lawyer in this case used MFSs to make it appear to the jury that Simpson’s testimony was:

incomplete or overly generalized (violating the maxim of quantity), incredible or  unfounded (violating the maxim of quality), off-question or beside the point (violating the maxim of relevance), or obscure (violating the maxim of manner) ... (ibid.: 228)

He achieved this, in the main, using a combination of the following procedural MFSs (ibid.: 228–9):

Requests for clarification

= Quality and/or Manner flout (as they ‘create the impression that the defendant’s testimony is ambiguous or unbelievable’)

Reminders of questions

= Quantity or Relevance flout (as they ‘create the impression that the defendant’s testimony is insufficient, beside the point or irrelevant’)

Procedural directives

= Quality flout (as they imply ‘defendant is not following the rules of courtroom testimony)


Find evidence of the above MFSs in the following extract, taken from Simpson’s 1997 civil trial (you might find it useful to start with the emboldened sections of the extract):

Q:        Is that scar the cut that you incurred in Los Angeles between 10 and 11p.m. that you recut in Chicago, yes or no?
A:        I would have to say no.
Q:        Is that the cut you reopened in Chicago, yes or no?
A:        I would say no.
Q:        What do you mean ‘you would say’? What happened? Can’t you just answer the question?
A:        You want me to explain?
Q:        I’m not asking you what you would say. You told the police that you cut your finger in Los Angeles, and that you cut it again in Chicago. I would like you to tell us what cut you were referring to.
(adapted from Janney 2007: 230)

Now, focus on the defendant’s response. Is he being evasive in your view? Justify your response via a close textual analysis of the defendant’s turns.

According to Janney (ibid.: 231):

While the attorney’s questions and comments appear ostensibly to be aimed at getting the defendant to be ‘more specific’, ‘more accurate’, ‘more relevant’, ‘tell the truth’, etc., they are actually part of a larger game designed to achieve just the opposite effect. In this metagame, the prosecutor’s strategically feigned inability to elicit ‘adequate’ answers becomes evidence of the defendant’s evasiveness.

To what extent do Janney’s comments suggest the CP works differently for different participants in a courtroom context? How might you go about proving Janney’s hypothesis? How might you disprove it?


  •  Janney, R.W. (2007) ‘“So your story now is that...”: Metapragmatic framing strategies in courtroom interrogation’. In: W. Bublitz and A. Hübler (eds.) Metapragmatics in Use, pp223–34. Amsterdam and Philadelphia: John Benjamins

5.4. ‘Reality paradigms’ and Gricean implicature

According to Archer (2002: 25), if Grice’s (1975) theory of implicature is to adequately account for conflictual situations like the courtroom (modern and historical), we need to recognize how:

the different assumptions two parties (may) bring to the discourse can lead to the adoption of different reality paradigms that, in turn, impact the operation of the maxims by [potentially] impeding the inferencing capacity of the addressee.

Reality paradigms are the truth filters interlocutors use to interpret/make sense of their worlds. They are thus more specific to the individual than expectation frames (see Tannen 1993), which effectively equate to schemas that provide us with a socially prescribed feel for the game (Bourdieu 1990). However, they tend to overlap with or feed into representational frames – that is, the way(s) in which interlocutors opt to ‘represent the character traits, ideas and opinions of and even statements made by others’ (Locher and Watts 2008: 99, n9).

A particularly noteworthy feature of reality paradigms, expectation frames and representational frames is their tendency to clash or be in opposition in a courtroom context. This is not surprising given that the point of the Anglo-American adversarial system is to discover truth via the contention of counsel (i.e. by choosing between opposing crime narratives). However, as Archer (2002, 2011b) has shown, such clashes – when extreme enough – can affect interlocutors’ ability or willingness to correctly perceive the implicatures of others.

Consider George Jacobs, Snr, a defendant in the infamous Salem Witchcraft Trials. Jacobs was so confident that he would be exonerated of charges of witchcraft in 1692 that he told the court ‘If you can prove that I am guilty, I will lye under it’. He was to regret his words, however, for the magistrate (William Stoughton) was operating within a reality paradigm that assumed guilt:

Here are them that accuse you of acts of witchcraft. Well, let us hear who are they, and what are they.
Abigail Williams – Jacobs laught.
Because I am falsely accused. – Your worships all of you do you think this is true?
Nay, what do you think? 
(adapted from Salem Witchcraft Papers 1977: 474)

Given that assigned roles help to determine ‘what kind of inferences will be made from what is said’ (Levinson 1992: 97), Jacob’s laughter is salient in this context. Archer (2002: 21) suggests the magistrate probably found it contemptuous – hence the need for Jacobs to explain himself. Notice that Jacobs went on to ask those present whether the allegations were true:

Stoughton’s response, ‘Nay, what do you think?’, implied that he did. Recognising the implicature, Jacobs immediately affirmed his innocence. That affirmation, ‘I am as innocent as the Child born to night, I have lived 33 years here in Salem’, [flouted the Quantity Maxim as it gave] more information than was necessary ... the seemingly irrelevant and ambiguous remark regarding his years in Salem also flouted the Manner and Relation Maxims, to generate the implicature that if he was involved in witchcraft, it would have been discovered long before. Jacobs was then asked to identify who was responsible for afflicting the ‘children’, to which he replied – ‘Dont ask me’. Jacobs presumably meant to imply (via the Maxims of Relation and Manner) that he either did not know and/or was not the right person to ask because of his innocence, but Stoughton missed or ignored the implicature and responded to the imperative’s literal meaning via a rhetorical question that affirmed his socially legitimated ‘right’ to question Jacob’s, ‘Why should we not ask you? Sarah Churchwell accuseth you, there she is’. (Archer 2002: 21) 

Archer (2011b) has shown how representational frames and reality paradigms can also be manipulated for strategic purposes in a courtroom context. By way of illustration, Carson, the defence counsel for Queensberry in the Trial of Regina vs. John Sholto Douglas, sought to ‘make evident [Oscar] Wilde’s “peculiar” view of literature, “truth” and life in general when compared with those of the average Victorian man’ (ibid.: 86):

Carson was able to demonstrate to the jury that Wilde didn’t believe in ‘truth’ as a fixed, absolute entity, for example: when asked about his Phrases and Philosophies for the Use of the Young, Wilde admitted that much of what he wrote was ‘not true ... in the actual sense of correspondence to actual facts of life’. This was not problematic for Wilde, as the purpose of the Phrases was to ‘stimulate thought in people of any age’. But Carson went on to link the notion of truth with (im)morality, by asking Wilde whether ‘anything that stimulate[d] thought’ was acceptable, ‘whether moral or immoral?’ Wilde’s response was to counter Carson’s presupposition that thought could be moral or immoral. Carson read aloud several more of Wilde’s Phrases, for the benefit of the jury, including one which claimed that ‘a truth ceases to be true when more than one person believes in it’. When asked whether he thought that this was ‘right’, Wilde responded that he did, adding ‘the same truth can never be apprehended by two minds’ as ‘to each mind there is its own truth’.


At this point in the trial, Carson asked Wilde whether one person could influence another adversely. Oscar Wilde did not think so:

WILDE:             I don’t think there is any influence, good or bad, from one person over another. I don’t think so.
CARSON:         A man never corrupts a youth?
WILDE:             I think not.
CARSON:         Nothing he could do would corrupt him.
WILDE:             Oh, if you are talking of separate ages it is nonsense.
CARSON:         No sir, I am talking common sense.
WILDE:             Do not talk like that ... personally, as a mere philosophical point, I don’t think – I am talking of grown human beings – that one person influences another. I don’t think so. I don’t believe it.
CARSON:         You don’t think that one man could exercise any influence over another? I may take that as a general statement?
WILDE:             As a general statement, yes. I think influence is not a power that can be exercised at will by one person over another: I think it is quite impossible psychologically.
CARSON:         You don’t think that flattering a young man, telling him of his beauty, making love to him in fact, would be likely to corrupt him?
WILDE:             No.
(Cross-examination of Wilde, Wednesday p.m. 3 April 1895 (Holland 2003: 102))

Which maxims are being flouted here, and what inferences are we meant to draw?

To what extent does the lawyer (Carson) flout maxims as a means of intimating – to the jury – that Wilde was ‘corrupting’ young men? And why does he make references to ‘common sense’?

To what extent do Wilde’s answers confirm that his view is different to the accepted societal norm, at this time?


  •  Archer, D. (2002) ‘“Can innocent people be guilty?” A sociopragmatic analysis of examination transcripts from the Salem Witchcraft Trials’, Journal of Historical Pragmatics 3(1): 1–30
  • Archer, D. (2011b) ‘Libelling Oscar Wilde: the case of Regina vs. John Sholto Douglas’, Journal of Politeness Research 7(1): 73–99
  • Bourdieu, P. (1990) Language and symbolic power, Cambridge: Polity Press
  • Boyer, P. and S. Nissenbaum (eds.) (1977) Salem Witchcraft Papers: Verbatim Transcripts of the Legal Documents of the Salem Witchcraft Outbreak of 1692, vols 1–3, New York: Da Capo Press
  • Briggs, C.L. (1997) ‘From the ideal, the ordinary, and the orderly to conflict and violence in pragmatic research’, Pragmatics 7(4): 451–9
  • Holland, M. (2003) ‘The real trial of Oscar Wilde’, The first uncensored transcript of the trial of Oscar Wilde vs. John Douglas (Marquess of Queensberry), 1895, New York: Perennial, HarperCollins
  • Levinson, S.C. (1992) ‘Activity types and language’ in P. Drew and  J. Heritage (eds.) Talk at Work: Interaction in Insitutional Settings, Cambridge: Cambridge University Press, pp66–100
  • Locher, M.A. and R. Watts. (2008) ‘Relational work and impoliteness: negotiating norms of linguistic behaviour’ in D. Bousfield and M.A. Locher (eds.) Impoliteness in language: studies on its interplay with power in theory and practice, Berlin and New York: Mouton de Gruyter, pp77–100
  • Tannen, D. (ed.) (1993) Framing in discourse. Oxford: Oxford University Press