Socialising Grice:

on interlocutors reasons for co-operating in conversation

Marina Terkourafi
British School at Athens and University of Cambridge
A way of extending the applicability of Grices Co-operative Principle to
non-co-operative exchanges is suggested. The argument builds on the
premises of rationality and interlocutors face wants (Goffman 1967; Brown
and Levinson 1987), which yield different degrees of co-operation depending
on the cultural and situational context. More specifically, it is proposed that,
in cases of non-co-operation, the correct results are obtained by applying the
maxims not just to what is said, but also to what is implicated. What prompts
this extended application of the maxims is interlocutors reciprocal sensitivity
to face-wants. Rather than being independently stipulated, the CP now falls
out from rationality and interlocutors mutual face-wants, affording us with a
glimpse into interlocutors' reasons for abiding by this principle.

INTRODUCTION

Grice's (1975/1989a) theory of conversational implicature is built on the assumption that

interlocutors share some basic goals, captured under the rubric of the Co-operative Principle:
Make your conversational contribution such as is required, at the stage at which it occurs, by
the accepted purpose or direction of the talk exchange in which you are engaged (henceforth
CP; 1989a: 26). The CP is a pre-condition of linguistic communication: it is because this is
assumed to be in operation at a deeper level that inferring the speakers intended meaning
(which, as much recent work in pragmatics argues convincingly, may well include
understanding the proposition expressed by his/her utterance) is at all possible. This leaves us
at a loss as to how to begin to account for instances of conflictual communication, or even
instances when, in Grices words, the CP is opted out of, at the same time as research
increasingly draws attention to the pervasiveness of such instances in everyday
communication (Haviland 1997; Eelen 2001; Leezenberg 2003).
Yet Grices scheme has proved extremely inspiring for linguistic research. One of its
major advantages lies in having formulated the maxims of conversation which, by hinging on
linguistic aspects of the speakers utterance, provide an opportune tool with which to
explicate otherwise opaque inferential processes. Indeed, attempts at formulating alternatives
to the CP (e.g. the Principle of Relevance, Sperber and Wilson 1986/1995) may be criticised
exactly on account of not having provided us with a tool of similar predictive force. It would
then seem that the restrictive nature of the CP is a mixed blessing: it is only because it limits
its applicability to co-operative exchanges that it can make predictions of some validity.
However, this is not an inescapable conclusion.
In this paper, I suggest a way of extending the applicability of the CP to non-cooperative exchanges as well, while retaining its full predictive power. Informed by recent
research tracing the origins of co-operative behaviour back to the pursuance of self-interest or
status (Dessalles 1998, 2000), the argument put forward builds on the premises of rationality
and interlocutors face wants (Goffman 1967; Brown and Levinson 1987), which yield
different degrees of co-operation depending on the cultural and situational context. This move
takes into account variability in definitions of the self, and as to which aspect of face, defined
as wants of the self, is prioritised in context, thereby enabling us to account not only for
2005 by Marina Terkourafi
Faye Chalcraft and Efthymios Sipetzis (eds.)
Cambridge Occasional Papers in Linguistics 2: xx-xx.

Marina Terkourafi

instances where co-operation is preferred, but also for instances where it may be opted out of,
or otherwise not fully provided. Such instances are not restricted to particular cultural
contexts (Keenan 1976/1998; Harris 1996). They are also a matter of situational context
(Sperber and Wilson 1986). Discussing examples of non-co-operative behaviour, I suggest a
re-analysis in which the maxims operate not just on what is said, but also on what is
implicated. What prompts interlocutors to apply the maxims not only to what is said, but also
to what is implicated is their reciprocal sensitivity to face-wants. Thus, in addition to being
compatible with Grice's scheme, the proposed account provides us with a glimpse into
interlocutors' reasons for abiding by the CP. This now falls out from rationality and
interlocutors' mutual face-wants, and does not have to be independently stipulated. To the
extent that rationality motivates face-constituting directly and Gricean co-operation only
derivatively, face-constituting rather than the CP must be placed at the basis of a general
theory of communication as providing the situational order that interlocutors expect in
interaction.
2

PREVIOUS ACCOUNTS

Since Grice, the notion that people talk co-operatively has become deeply entrenched.
However, the scope of co-operation has been a matter of considerable debate. Whereas
Grice (1989a: 28) saw co-operation as governing non-linguistic exchanges as well, Bach and
Harnish (1979; cf. Harnish 1998 [1976]: 304, fn.31; Bach 1987) restrict its application to
linguistic communication. Closer to the spirit of Grices original formulation, who saw
talking as a special case of purposive, indeed rational behaviour (Grice 1989a: 28), Cohen
and Levesque attribute to co-operative agents the properties of being sincere and helpful
(1990: 22930), which they also define formally, while Thomason suggests an implicatureenabling notion of accommodation defined as acting to remove obstacles to the achievement
of desires or goals that we attribute to others (1990: 332).
Attempting to clarify matters, Attardo (1997: 756) distinguishes two levels of cooperation: Locutionary Co-operation (LC) refers to the amount of co-operation, based on the
CP, that two speakers must put into the text in order to encode and to decode its intended
meaning, while Perlocutionary Co-operation (PC) captures the amount of co-operation two
speakers must put into the text/situation to achieve the goals that the speaker (and/or the
hearer) wanted to achieve with the utterance. Illustrating this point with examples, Attardo
argues that the speaker can be LC co-operative (s/he can abide by the CP) without necessarily
being PC co-operative (taking into account the hearers goals in the situation). He then
proposes the following Perlocutionary Co-operative Principle (PCP):
(1)

Co-operate in whatever goals the speaker may have in initiating a

conversational exchange, including any non-linguistic, practical goal. (Or in
other words, be a good Samaritan). (Attardo 1997: 766)

The PCP is more general than the CP, and takes precedence over it. Other principles
governing our actions, such as self-interest, may nevertheless override the PCP in cases of
conflict (ibid.: 777).
Attardos PCP goes some way toward accounting for common conversational
practices, but provides no clear insights as to why this principle should hold. Leech takes a
step in this direction, when he introduces the P[oliteness] P[rinciple] as a principle
functioning on a par with the CP:
(2)

The CP enables one participant in a conversation to communicate on the

assumption that the other participant is being co-operative. In this the CP has
the function of regulating what we say so that it contributes to some

Socialising Grice

illocutionary or discoursal goal(s). [] the P[oliteness] P[rinciple] has a higher

regulative role than this: to maintain the social equilibrium and the friendly
relations which enable us to assume that our interlocutors are being cooperative in the first place. To put matters at their most basic: unless you are
polite to your neighbour, the channel of communication will break down, and
you will no longer be able to borrow his mower. (Leech 1983: 82)
In other words, it is important to be polite (or PCP co-operative)1 because not doing so makes
pursuing ones aims a lot harder. In recent work, Attardo (1999, 2003) outlines practical
features of human rationality and complements his earlier proposal of the PCP with a number
of further principles (Politeness, Non-Co-operation, Rationality, Self-interest) interacting with
the CP in an optimality-theoretic way to yield the most likely interpretation.
The proliferation of principles explaining departures from the CP, nevertheless, fails to
answer (at least) one question, that of motivation: in other words, why should it be these
particular principles (Leechs PP, or Attardos hierarchy of principles) that underlie
communication and no others? So long as the principles proposed reflect observational facts
about conversation but lack deeper theoretical motivation, both the question of their number
(already five in Attardos latest proposals) and of their universality will remain open.
3

AN ALTERNATIVE ACCOUNT

Contrary to the proposal of new principles, drawing on the premises of interlocutors

rationality and claims to face provides a more promising route to dealing with departures from
the CP. This section begins by outlining the relevant notions of face and rationality.
Subsequently, it is shown how different degrees of co-operation can be derived from these
premises in particular cultural and situational contexts. Crucially, the proposed account
remains Gricean in spirit: application of the maxims correctly predicts the intended
implicatures, even in cases of non fully co-operative, or even openly conflictual
communication.
3.1

Face as wants of the self

The notion of face appealed to builds on, yet revises, that of Brown and Levinson (1987
[1978]: 61). Brown and Levinson define face as: the public self-image that every member
wants to claim for himself, consisting in two related aspects: (a) negative face: the basic claim
to [] freedom of action and freedom from imposition (b) positive face: the positive
consistent self-image or personality claimed by interactants, crucially including the desire
that this self-image be appreciated and approved of. First, taking heed of Goffmans
definition of face as an image of [a person] that is internally consistent, that is supported by
judgements and evidence conveyed by other participants, and that is confirmed by evidence
conveyed through impersonal agencies in the situation (1967: 67; italics mine), emphasis is
placed on the interpersonal and emergent aspects of face. Face is now seen as the societys
mark of approval, the positive sanctioning of particular behaviours, as it emerges in
interaction with another agent. To recognise and positively sanction specific behaviours,
interacting agents must partake of a common stock of closely related evaluative practices. For,
lacking such a common stock of practices, there exists no guarantee that they will mutually
either identify particular token behaviours as the intended type of behaviour, or deem these
worthy of a positive evaluation. In other words, face cannot be constituted in vacuo through
1

The two attitudes are equivalent inasmuch as they reflect the observation that people often co-operate
beyond what is necessary for understanding and being understood.

Marina Terkourafi

rational individual action alone (schematically represented in (3)), but requires a common
stock of socially ratified evaluative practices within which it can be constituted (schematically
represented in (4)). Such evaluative practices are per definitionem supra-individual, and thus
take precedence, temporally as well as ontologically, over individual agents wants and the
situated content of these wants.
(3)

Face as constituted through rational individual action alone

(4)

Face as socially constituted

Society

In addition to (re-)conceptualising face as a social attribute of interactants dynamically

constituted in interaction, the current proposal does not grant priority to one aspect of face
over the other a priori. Brown and Levinson award priority to negative face over its positive
counterpart on the basis of arguments such as that: it is safer to assume that [the] H[earer]
prefers his peace and self-determination than that he prefers your expressions of regard,
unless you are certain of the contrary (1987: 734). Such statements suffer from a certain
degree of ethnocentrism that turns out to be unjustified both with respect to their intellectual
sources,2 and data from a range of languages,3 in which considerations of in-group
membership and/or social hierarchy replace the emphasis on individualism characteristic of
western societies.4 The variable claims advanced in the literature as to interpersonal
supportiveness (Arndt and Janney 1985: 282) or non-imposition (Ide 1982: 382) taking
precedence suggest that decisions as to the priority of one aspect of face over the other are
arbitrary. Consequently, while I take the need for both aspects, as well as their paradoxical
inter-relation (Scollon and Scollon 1995:36), to be well established theoretically in Goffmans
and Durkheims writings, I maintain that decisions to prioritise one over the other can only be
taken following consideration of the culture and the situation at hand.
2

On this topic, Durkheim and Goffman; Brown and Levinson acknowledge an intellectual debt to both
scholars on p.1 of the1987 re-edition of their essay. Durkheims and Goffmans shying away from similar
generalisations is seen in the following two excerpts:
Whatever the importance of the negative cult may be, [...], it does not contain its reason for existence
in itself; it introduces one to the religious life, but it supposes this more than it constitutes it (Durkheim
1976[1915]: 326).
If we examine what it is one participant is ready to see that other participants might read into a
situation and what it is that will cause him to provide ritual remedies of various sorts [...], then we find ourselves
directed back again to the core moral traditions of Western culture. And since remedial ritual is a constant
feature of public life, occurring among all the citizenry in all social situations, we must see that the historical
centre and the contemporary periphery are linked more closely than anyone these days seems to want to credit
(Goffman 1971: 184-5).
3

These include: Japanese, Chinese, Ojibwa, Igbo, Polish and Greek. For references, see: Matsumoto
1988: 405; Gu 1990: 241-2; Rhodes 1989: 254-257; Nwoye 1992: 324; Wierzbicka 1991; Sifianou 1992.
4

Even in western societies, importance attributed to negative face varies, as attested by differences
between West-coast and East-coast Americans (Brown and Levinson 1987: 245; Tannen 1981: 229fn.4).

Socialising Grice

In what follows, positive and negative face will be seen as wants (rather than
properties) of the self at once dynamically arising and finding fulfilment in interaction (cf.
Arundale 1999). This definition remains valid across situations and cultures, while variability
in the specific contents of face each time can be attributed to variability in conceptualisations
of the self during interaction. Recent research suggests that notions of the self emerge early on
in life (Butterworth 2000) as a result of interaction with the environment (ibid.; Gallagher and
Marcel 1999). The emergence of intentionality itselfa cornerstone of the reflexive notion of
selfhas been linked to dyadic interaction, initially realised as mother-infant interaction
(Gibbs 2001: 120121; Brinck 2001: 268ff.). These findings support the present argument in
two ways. First, since notions of self arise through reference to non-self (Gallagher and
Marcel 1999: 290), face as wants of the self inevitably arises and finds fulfilment in a dyad,
which accounts for its omni-relevance to communication. Second, the fact that ecological
information is crucial to distinguishing self from non self (ibid.) provides some initial
scientific backing to the claim that definitions of the self are contextually variable, a claim
previously advanced on semi-folk observational grounds (cf., e.g., Hsu 1983 quoted in
Scollon and Scollon 1995: 1323). Thus, by incorporating variation at the level of
conceptualisations of the self, we are afforded with a more flexible notion of face: the
universal validity and theoretical usefulness of this notion are retained, while its specific
contents are fixed inhence, vary acrossparticular situational and cultural contexts.
3.2

Rationality

With our revised notion of face in place, we may now begin to account for the observed
variability in applications of the CP. Kashers Rationality Principle (R) provides a useful, if
unwitting, starting-point in this respect: Given a desired end, one is to choose that action
which most effectively, and at least cost, attains that end, ceteris paribus (1976/1998: 188).
Reasoning in accordance with R yields the conclusion that coordination must be preferred to
realise ends that are intimately related to the actions of others, in that such ends can only be
attainedor attained at a lesser costthrough coordination (ibid.: 191).
In attempting to answer the question why it is rational for people to talk cooperatively,5 Kasher has in fact spelled out how face considerations can both motivate, and
deter, co-operation between parties. Understanding considerations of face as desired ends
motivating specific plans, we may view co-operation as a plan in itself, whose preferability
is judged every time relative to the aspect of face (positive or negative) prioritised in context.
This may initially create the impression that only positive face is concerned: after all, cooperation will constitute the preferred planin the sense current within Conversation
Analysisonly if considerations of positive face are given precedence.
Nevertheless, negative face is not completely out of the picture: on considerations of
negative face alone (avoiding imposition), co-operation constitutes a dispreferred plan, one
that will be purposefully avoided. It would then be quite irrational to engage in co-operative
activities that do not serve any practical ends unless some gain (e.g. in positive face) can be
posited as an additional benefit of such activities. That people commonly do engage in such
activities constitutes prima facie evidence of their carrying some such additional benefit.6 In
5

Kasher derives the conversational maxims directly from R, without appealing to the CP. He also
distinguishes limited co-operation, which is warranted by R, from the CP, which is too strong, and therefore
wrong and needless (1998: 192).
6

Trudgills (1983[1974]: 13) example of two Englishmen who have never met before and who, upon
coming face to face with one another in a train compartment, start talking about the weather, is perhaps the
paradigmatic case of such an activity (but only under his first explanation, that it can often be quite
embarrassing to be alone in the company of someone you are not acquainted with and not speak to them; cf.
Leech 1983: 141). While suggesting that it is indeed legitimate to view considerations of face as desired ends

Marina Terkourafi

this respect, two levels of rational behaviour may be distinguished: face-constituting is

rational at a deeper level, on which it orientates participants to maintain social equilibrium.
For this reason, face-constituting can, at surface level, become an end in itself. In this sense, it
constitutes perhaps the prototypical case of an end intimately related to the actions of others
(Kasher 1998: 191).
3.3

Cross-cultural and intra-cultural degrees of co-operation

Incorporating considerations of face into interlocutors desired ends enables us to account not
only for cases where co-operation is preferred, but also for instances where it may be opted
out of, or otherwise not fully provided. The Malagasys systematic7 reluctance to provide
personal information when asked, even when they could truthfully do so (Keenan 1998
[1976]), is often cited as an instance of less than full co-operation. However, rather than
constituting downright refutation of the Gricean maxims, Keenans findings have been
repeatedly interpreted as highlighting the need to constrain their application with reference to
the operative cultural norms (Harnish 1978/1998: 304fn.29; Brown and Levinson 1987: 288
9fn.27; G.Green 1989/1996: 1001; Mey 1993: 74). In a similar vein, apparent violations of
the maxim of Quality during everyday interaction in an Egyptian village have led Harris
(1996) to discoverunderneath what, at first sight, constitutes practically no co-operation at
alla complex system of face considerations. Her proposal is that, during utterance
interpretation, social norms constrain the operation of the CP, as well as providing, together
with the situational context, the input to the operation of the maxims of Relation and of
Manner (1996: 49).
Instances when people co-operate less than fully are not restricted to particular cultural
contexts. They are also a matter of situational contexteven outside institutional discourse.
Discussing Grices (1989a: 32) example
(5)

A: Where does C live?

B: Somewhere in the South of France

Sperber and Wilson (1995 [1986]: 273ff.) point out that, in addition to Grices analysis, on
which Bs utterance flouts the first sub-maxim of Quantity giving rise to the implicature that
B does not know exactly where C lives, Bs utterance may also be taken as implicating that B
does not want to reveal exactly where C lives.
For instance, A and B may be planning a trip to France. A wants to know where C
lives so they may pay C a visitonly B cannot stand the sight of C, and is therefore not
inclined to give out the requisite information that would allow A to go ahead with his/her
plan. Now, if A has reason to believe that, contrary to appearances, B is in fact in possession
of this information (say, because B has said so in the past), A may well interpret Bs utterance
as indicative of Bs reluctance to reveal this information. In this case, only if A reasons on the
assumption that B is not being co-operative (in the sense of not sharing As extra-linguistic
goal), will A infer Bs reluctance based on Bs utterance. What is more, the implicature that B
does not want to say exactly where C lives arises contrary to Gricean prediction, since the CP
is obviously not operative at a deeper level. Based on this discussion, Sperber and Wilson
conclude that the CP is sometimes too strongpredicting that interlocutors always share
that people aim to fulfil through rational cost-and-effect accounting, this example at the same time challenges
Brown and Levinsons assumption that negative face generally takes precedence over positive face (Brown and
Levinson 1987: 73-4), showing that even in an Anglo-Saxon context this is not always the case.
7

Systematic points to the fact that, in Malagasy society, the amount of information provided in
conversation is regulated by situational constraints (Keenan 1998: 224ff.).

Socialising Grice

some common extra-linguistic goals and will co-operate toward their achievementand
therefore neither always at work, nor always presumed to be at work (1995: 274).
3.4

A Gricean account of non co-operative exchanges

Sperber and Wilsons analysis of example (5) in the previous section is however not the only
possible one. Another possibility is to assume that B is opting out of the CP (Grice 1989a:
30). However, this entails that the CP must be operative to start with, for A to derive from Bs
reply the implicature B does not know exactly where C lives, then contrast this with a piece
of background knowledge (that A knows that B knows exactly where C lives), and based on
the resulting contradiction infer that B has in fact opted out of the CP. This is problematic,
since the CP is first assumed to be operative, and subsequently not to be. Also, in this case,
Bs opting out is inferred, while Grice (ibid.) only predicts that the CP may be opted out of by
say[ing], indicat[ing], or allow[ing] it to become plain that [one] is unwilling to co-operate in
the way the maxim requires. [One] may say, for example, I cannot say more; my lips are
sealed.
A further possibility is to assume that the CP remains operative throughout. Two lines
of argument are now possible. First, one may argue that, in view of the blatancy of Bs
performance, B is trying to mislead A (Grice 1989a: 30). Again, this will not work if A
already knows that B has the requisite information, and B knows that A knows that.
Alternatively, we may allow the maxims to operate not just on what is said, but also on what
is implicated. Quantity operates first and gets A from Bs reply to the implicature that B does
not know where C lives. Then Quality enters the picture, and when the implicature that B
does not know is contrasted with the piece of background knowledge that B in fact knows,
the implicature that B does not want to tell is derived. The resulting two-stage application of
the maxims is diagrammatically represented in (6) (where +> stands for conversationally
implicates) .
(6)

Two-stage application of the maxims to what is said and to what is implicated

A: Where does C live?

Quantity +>
B doesnt know where C lives
B: Somewhere in the South
of France

Quality +> +>

B doesnt want to tell...
cf. B knows where C lives
(background information)

This time, rather than Bs utterance, it is the implicature that B does not know that breaches
Quality, since B has now implicated something which s/he believes to be false. What could
prompt A to apply the maxims not just to what is said, but also to what is implicated, so as to
derive this further implicature? Interlocutors reciprocal sensitivity to face-wants suggests
itself here. As Brown and Levinson (1987: 99) remark, respect for face involves mutual
orientation, so that each participant attempts to foresee what the other participant is
attempting to foresee. A may derive the further implicature that B does not want to reveal
precisely where C lives only if A is sensitive to the fact that B, by not openly expressing
disregard for As wishes, may be avoiding to threaten As positive face.
It is difficult to tell whether Grice actually had in mind such operation of the maxims
on two levels. Discussing example (5) above, he only derives the dont know implicature
(1989a: 323). On the other hand, the proposed operation of the maxims on two levels is
distinct from the possibility, which he acknowledges, that conversational implicatures are

Marina Terkourafi

indeterminate, and may on occasion take the form of a disjunction of [] specific

explanations (1989a: 40).
The proposed account is nevertheless compatible with Grices scheme, which in fact it
may be viewed as strengthening, by spelling it out more fully. The inferential process is still
guided by the maxims on the assumption that the CP is operative. And the CP is operative
because interlocutors, whether they share any particular goals relating to the exchange at hand
or not, always share the over-arching goal of ensuring that their face wants are fulfilled in
interaction. Thus, the CP does not have to be abandoned in the case of conflictual or
otherwise not fully co-operative interaction.8
Further examples of conversational behaviour not obviously falling under the umbrella
of the CP can be explained in this way. Consider the following real-life exchange:
(7)

(A and B are academics discussing the vacancy of a post previously held by C)

A: Where is C gone?
B: Im sorry, I have forgotten.

On the classical Gricean account, Bs reply is akin to opting out from the CP, hence cannot
give rise to any implicatures. Nevertheless, given that, throughout the preceding conversation,
B had given several indications of being well informed about the situation, A did infer on this
occasion that B does not want to tell to which University C is gonea proposition which,
of course, qua implicated, remains cancellable.9 The problem, in this case, is how to account
for As interpretation of Bs reply. Again, this is only possible if their reciprocal sensitivity to
face is taken into consideration. By claiming inability to comply with As request, B, on the
one hand, succeeds in avoiding threat to As face, while A, on the other hand, realises this
only on condition of being aware of Bs face-saving intention. Experience of common
conversational practices includes several examples of such social white lies, termed white
exactly because they are underlain by a face-saving intention, with reference to which they are
also recoverable (and often recovered).
Other examples not falling under the umbrella of the CP include instances of openly
conflictual communication. In one such incident, A was a guest staying in Cs London flat,
and B was Cs neighbour. Despairing with Bs noisy lifestyle, one day upon running into B on
the stairs, A reprimanded B in As native language, being fully aware that B does not
understand this. Subsequently, B modified her behaviour, leading a less noisy life.
At first sight, one may feel inclined to deny that communication has taken place on
this occasion, claiming that this is not a case of meaningNN at all: As angry tone of voice
could well mean anger in the same (natural) way that spots mean measles in Grices
(1957/1989c) original example. Indeed, a natural meaning component must in all likelihood
be acknowledged as part of As utterance.
Nevertheless, I wish to argue that this acknowledgement would not be sufficient. For
Bs subsequent peace-preserving behaviour compels us to consider the possibility that
something was indeed communicated by As utterance. What was communicated and how are
the questions we are called to answer. First, let us note that As utterance (also) meets Grices
8

Speaking of co-operation as a gradable notion may become clearer if, paraphrasing Ide (1989: 225), we
point out that: [co-operation] is a neutral term. Just as height does not refer to the state of being high, a
theory of co-operative action ought to concern itself with the full range of possible attitudes toward cooperativity, from complete co-operation to non-co-operation.

It may seem that this example, by not involving the kind of two-level application of the maxims
outlined above, constitutes a classical case of Gricean implicature. This is, however, not so, since B in (2) cannot
be claimed to have wanted A to recover the proposition that B does not want to tell, whereas in the classical
Gricean scheme, implicated propositions are m-intended by their utterers (Grice 1969/1989b).

Socialising Grice

criteria for meaningNN: by means of this utterance, A intended to produce a certain effect on
B, and A intended to produce this effect on B by way of Bs recognition of As intention to
produce it. Crucially, as in Grices examples, it is this audience-directed intentionrather
than anything inherent in the form of the utterancethat endows As utterance with
meaningNN. As utterance was intended to be recognised by B, not by just anyone who
happened to be present; while the combined knowledge that it was A and not any other
speaker who uttered this utterance, and that he addressed it to B and not to just any other
addressee, is equally important to Bs recognising the content of As intention, despite not
understanding Bs words. As utterance may, then, be viewed as an extreme violation of the
maxim of Manner, and in particular of its first sub-maxim, Avoid obscurity of expression.
On this view, it would be As violation of Manner that would prompt B to engage in
inferencing based on As utterance.
However, the absence of any semantic clues on which to base the inferential process
would make this process seemingly disproportionately costly, perhaps even present B with an
insurmountable obstacle in this respect.10 Any attempts on behalf of B to infer As meaning
(other than As anger, a matter of natural meaning as outlined above) would have to be made
on extremely shaky grounds and, were it a matter of rationality alonei.e., face
considerations notwithstandingB would be better off avoiding the effort that would lead to
such uncertain gain. But perhaps it is not a matter of rationality alone, after all. Put
differently, perhaps face-considerations are part of the expected effects in a process of rational
cost-and-effect accounting: if B did expend the extra effort and did indeed engage in
inferencing based on As utteranceas may be hypothesised based on her subsequent
compliancethis can only be attributed to sensitivity to face wants. Risky as clarifying As
goal in threatening Bs face might be, the risk entailed by ignoring the threat to her face
would have been even greater. This is then an opportune example of how face-work may be
directed to own [sic] and/or to others [sic] face, and the effects on face of a face-work
attempt include not only maintenance and restoration, but also enhancement and damage
(Muntigl and Turnbull 1998:243). Many other instances of conflictual behaviour can be
similarly shown to draw on the Gricean inventory of maxims, exemplifying how face
considerations can motivate (and be taken as motivating) the whole range of degrees of cooperation.
4

FURTHER AFIELD: EVIDENCE FROM LANGUAGE PATHOLOGY AND LANGUAGE

EVOLUTION

The account of non co-operative exchanges outlined in the previous section provides us with a
glimpse into interlocutors reasons for abiding by the CP. This principle now falls out from
rationality and interlocutors mutual face-wants, and does not have to be independently
stipulated. Rather than the CP, it is these two premises that are assumed by default. It is
important (read: rational) to co-operate, not just to avoid damage to anothers face, but also to
ensure that ones own face will not be damaged. Given variability regarding definitions of the
self, and which aspect of facere-conceptualised as wants of the selfis prioritised in
context, this move enables us to account for the cultural and situational variation in degrees of
co-operation reported on in numerous studies (see section 3.3).
Evidence from language pathology and language evolution supports mutual
orientation to face as regulating co-operation, by showing how face considerations are
responsible for promoting and/or enhancing socially embedded behaviour in general. Recent
research on motor performance of brain damaged subjects (reported in Gallagher and Marcel
1999: 276ff.) showed that best performance was obtained in actions that were personally
10

Language-determination is a pragmatic prerequisite for the semantic process of content-determination

to begin (Rcanati 2001).

10 Marina Terkourafi
significant or that derived their signification from the social and cultural system (ibid.: 277).
Compared to disembedded, or even contextualised action, patients performance improved to
near-normal or normal, when performing the same actions as part of a cultural practice which
contributes to the constitution of the social competence and self-esteem of the agent (ibid.),
in other words, when guided by considerations of face as defined in 3.1 above.
In a similar vein, research into the origins of co-operative communication has
suggested an underlyingly non-altruistic motivation for seemingly altruistic behaviour. More
specifically, Dessalles (1998, 2000) has argued that providing relevant information is guided
by the quest for status within a group. Providing relevant information is altruistic, hence
should be counter-indicated on Darwinian grounds. Nevertheless, if norms of relevance (such
as are arguably captured by the CP) have evolved in human communities, this is because of a
continuous trade-off between relevance and status. Being granted status within a group
enhances ones chances of evolutionary survival, hence is worth sharing relevant information
for. Moreover, the more relevant the information, the more status one will be granted; that is,
competition for status produces an increase in relevance, while at the same time listeners
checking procedures ensure the accuracy of the information provided (since providing
inaccurate information will slowly but surely lead to loss of status).
The notion of status appealed to by Dessalles bears some close affinities to the revised
notion of face outlined in 3.1 above. Thus, status is defined as an emergent attribute resulting
from a complex combination of attitudes adopted by other individualsattitudes such as
respect, esteem, deference, loyalty allegiance, admiration, honour, homage or worship
Dessalles (2000: 71; emphasis added). This definition contains several of the elements
identified earlier as determinants of face as wants of the self: face is not static, but emerges in
interaction, and as a result of others evaluation of selfs behaviour; moreover, it consists in
two related aspects, positive (appreciation and approval of ones desires; cf. admiration)
and negative (freedom from imposition; cf. deference). The identification of such a notion
of status as a force motivating co-operative communication thus lends support to extending
the Gricean scheme by including face considerations among the goals that interlocutors
pursue in interaction, an extension proposed in 3.4 above on purely synchronic grounds.
Ongoing research on the evolutionary importance of empathy (Preston and de Waal 2002) and
joint attention (Kumashiro et al. 2003), two phenomena ontogenetically linked to the
emergence of notions of the self in 3.1 above, is likely to suggest further points of contact
between co-operation and face as wants of the self. An interesting line of research in this
respect would concern observing how increasing degrees of self awareness in primates, as
developed through, e.g. encouraging joint attention, may correlate with degrees of inferential
ability and the development of ToM (Theory of Mind).
The two independent lines of research briefly reviewed in this section thus concur in
placing face-considerations, rather than the CP, at the basis of a general theory of
communication as providing the situational order that interlocutors expect in interaction.
Acknowledging that rationality motivates face-constituting directly and Gricean co-operation
only derivatively allows us to accord priority to the social aspect (Geis 1995: 13ff.; Harris
1996: 94). We can thus begin to account, not only for instances of conflictual and otherwise
not full co-operation, but also for examples where face is constituted by speaking in
accordance with the CP, and not by means of deviations from it (Arundale and Ahston 1992;
Terkourafi 2001).
REFERENCES
Arndt, Horst and Richard Janney (1985) Politeness revisited: cross-modal supportive
strategies. International Review of Applied Linguistics 23: 281300.
Arundale, Robert (1999) An alternative model and ideology of communication for an
alternative to politeness theory. Pragmatics 9.1:119153.

Socialising Grice

11

Arundale, Robert and Ashton, Deborah (1992) Is face ever ignored? The case of Brown and
Levinsons bald, on record utterances. Paper presented at the Annual Conference of
the International Communication Association, Miami, Florida, 2125 May 1992.
Attardo, Salvatore (1997) Locutionary and perlocutionary co-operation: the perlocutionary
principle. Journal of Pragmatics 27: 753779.
Attardo, Salvatore (1999) The place of co-operation in cognition. In Sebastiano Bagnara
(ed.) ECCS99. European Conference on Cognitive Science 1999, Siena, Italy.
Siena/CNR, Roma: Universit di Siena, pp. 459464.
Attardo, Salvatore (2003) On the nature of rationality in (Neo-Gricean) pragmatics.
International Journal of Pragmatics 14: 320.
Bach. Kent (1987) Thought and reference. Oxford: Clarendon Press.
Bach, Kent and Robert M. Harnish (1979) Linguistic communication and speech acts.
Cambridge, MA: MIT Press.
Brinck, Ingar (2001) Attention and the evolution of intentional communication. Pragmatics
and Cognition 9.2: 259277.
Brown, Penelope and Stephen Levinson (1987 [1978]) Politeness: Some universals in
language usage. Cambridge: Cambridge University Press.
Butterworth, George (2000) An ecological perspective on the self and its development. In
Dan Zahavi (ed.) Exploring the self: philosophical and psychopathological
perspectives on self-experience. Amsterdam: John Benjamins, pp. 1938.
Cohen, Philip and Hector Levesque (1990) Rational interaction as the basis for
communication. In Philip Cohen, Jerry Morgan and Martha Pollack (eds.) Intentions
in communication. Cambridge, MA: MIT Press, pp. 221255.
Dessalles, Jean-Louis (1998) Altruism, status, and the origins of language. In James Hurford,
Michael Studdert-Kennedy and Chris Knight (eds.) Approaches to the evolution of
language: social and cognitive bases. Cambridge: Cambridge University Press, pp.
130147.
Dessalles, Jean-Louis (2000) Language and hominid politics. In Chris Knight, James
Hurford and Michael Studdert-Kennedy (eds.) The evolutionary emergence of
language: social function and the origins of linguistic form. Cambridge: Cambridge
University Press, pp. 6280.
Durkheim, Emile (1976 [1915]) The elementary forms of the religious life. London: George
Allen and Unwin. Second edition.
Eelen, Gino (2001) A critique of politeness theories. Manchester: St. Jeromes Press.
Gallagher, Shaun and Anthony Marcel (1999) The self in contextualised action. In Shaun
Gallagher and Jonathan Shear (eds.) Models of the self. Thorverton: Imprint
Academic, pp. 273299.
Geis, Mike (1995) Speech acts and conversational interaction. Cambridge: Cambridge
University Press.
Gibbs, Raymond (2001) Intentions as emergent products of social interactions. In Bertram
Malle, Louis Moses and Dare Baldwin (eds.) Intentions and intentionality:
foundations of social cognition. Cambridge, MA: MIT Press, pp. 106122.
Goffman, Erving (1967) Interaction ritual: essays in face-to-face behaviour. Garden City,
NY: Doubleday.
Goffman, Erving (1971) Relations in public: microstudies in the public order. London:
Penguin.
Green, Georgia (1996 [1989]) Pragmatics and natural language understanding. New Jersey:
Lawrence Erlbaum. Second edition.
Grice, Herbert Paul (1989a [1975]) Logic and conversation. In Herbert Paul Grice (1989)
Studies in the way of words. Cambridge, MA: Harvard University Press, pp. 2240.

12 Marina Terkourafi
Grice, Herbert Paul (1989b [1969]) Utterers meaning and intentions. In Herbert Paul Grice
(1989) Studies in the way of words. Cambridge, MA: Harvard University Press, pp.
86116.
Grice, Herbert Paul (1989c [1957]) Meaning. In Herbert Paul Grice (1989) Studies in the
way of words. Cambridge, MA: Harvard University Press, pp. 213223.
Gu, Yuego (1990) Politeness phenomena in Modern Chinese. Journal of Pragmatics 14:
23757.
Harnish, Robert M. (1998 [1976]) Logical form and implicature. In Asa Kasher (ed.)
Pragmatics: some critical concepts. Vol. IV. London: Routledge, pp. 230313.
Harris, Rachel (1996) Truthfulness, conversational maxims and interaction in an Egyptian
village. Transactions of the Philological Society 94: 3155.
Haviland, John (1997) Shouts, shrieks, and shots: Unruly political conversations in
indigenous Chiapas. Pragmatics 7: 547573.
Ide, Sachiko (1982) Japanese sociolinguistics: politeness and womens language. Lingua 57:
357385.
Ide, Sachiko (1989) Formal forms and discernment: two neglected aspects of universals of
linguistic politeness. Multilingua 8: 223248.
Kasher, Asa (1998 [1976]) Conversational maxims and rationality. In Asa Kasher (ed.)
Pragmatics: some critical concepts. Vol. IV. London: Routledge, pp. 181213.
Keenan, Eleanor Ochs (1998 [1976]) The universality of conversational postulates. In Asa
Kasher (ed.) Pragmatics: some critical concepts. Vol. IV. London: Routledge, pp.
215229.
Kumashiro Mari, Hidetoshi Ishibashi , Yukari Uchiyama , Shoji Itakura , Akira Murata and
Atsushi Iriki (2003) Natural imitation induced by joint attention in Japanese
monkeys. International Journal of Psychophysiology 50: 8199.
Leech, Geoffrey (1983) Principles of Pragmatics. London: Longman.
Leezenberg, Michiel (2003) Communication as social practice: the interface between the
cognitive and the social sciences. In Lszl Komlsi, Peter Houtlosser and Michiel
Leezenberg (eds.) Communication and Culture: Argumentative, Cognitive and
Linguistic Perspectives. Amsterdam: SicSat, pp. 6977.
Matsumoto, Yoshiko (1988) Reexamination of the universality of face: politeness
phenomena in Japanese. Journal of Pragmatics 12: 403426.
Mey, Jacob (1993) Pragmatics: an introduction. Oxford: Blackwell.
Nwoye, Onuigbo (1992) Linguistic politeness and socio-cultural variations of the notion of
face. Journal of Pragmatics 18: 309328.
Preston, Stephanie and Frans de Waal (2002) Empathy: its ultimate and proximate bases.
Behavioral and Brain Sciences 25: 172.
Rcanati, Franois (2001) Open Quotation. Mind 110: 637687.
Rhodes, Richard (1989) We are going to go there: positive politeness in Ojibwa.
Multilingua 8: 249258.
Scollon, Ron and Scollon, Suzanne Wong (1995) Intercultural communication: a discourse
approach. Oxford: Blackwell.
Sifianou, Maria (1992) The use of diminutives in expressing politeness: Modern Greek
versus English. Journal of Pragmatics 17: 15573.
Sperber, Dan and Wilson, Deirdre (1995 [1986]) Relevance: communication and cognition.
Oxford: Blackwell. Second edition.
Tannen, Deborah (1981) Indirectness in discourse: ethnicity as conversational style.
Discourse Processes 4: 22138.
Terkourafi, Marina (2001) Politeness in Cypriot Greek: A frame-based approach. Ph.D.
dissertation. University of Cambridge.

Socialising Grice

13

Thomason, Richmond (1990) Accommodation, meaning, and implicature: interdisciplinary

foundations for pragmatics. In Philip Cohen, Jerry Morgan and Martha Pollack (eds.)
Intentions in communication. Cambridge, MA: MIT Press, pp. 325363.
Trudgill, Peter (1983 [1974]) Sociolinguistics: an introduction to language and society.
Harmondsworth: Penguin. Second edition.
Wierzbicka, Anna (1991) Cross-cultural pragmatics: the semantics of human interaction.
Berlin: Mouton de Gruyter.

Marina Terkourafi
Linguistics Department
Sidgwick Avenue
University of Cambridge
Cambridge
CB3 9DA
United Kingdom
mt217@cam.ac.uk
http: //www.cus.cam.ac.uk/~mt217