Pragmatics and power

This paper sets out to ascertain on the basis of actual language behaviour whether 'universal pragmatics', especially Grice's maxims and Habermas's validity claims, can provide a sustainable conceptual framework for understanding the relationship between pragmatics and power. On the basis of a data base of magistrate/defendant and police/suspect discourse, it examines the nature of strategic discourse in settings where participants have conflicting goals. Three specific propositions are put forward: (1) that the asymmetrical distribution of speech acts as a mode of strategic communication prevents validity claims being raised or challenged except by institutional representatives; (2) that 'truth' comes to be defined pragmatically as what is accepted explicitly as 'shared knowledge'. Powerful institutional members move from the 'given' to the 'new', which is often 'disputable', by a variety of communicative strategies which the less powerful 'clients' find difficult to challenge; (3) that there is a need to re-define the widely used concept of communicative competence so that any challenge to authority or the right to speak is not merely labelled as 'inappropriate' and/or 'incompetent' language behaviour. The paper concludes with an assessment of the usefulness of the work of Grice and Habermas in relationship to strategic discourse and proposes a possible model for future work. 1. Universal pragmatics Recent theories of 'universal pragmatics ' are concerned with attempting to identify the fundamental and underlying principles o f interactive communicat ion. I f pragmatics is, broadly speaking, the study of language usage (Levinson, 1983), 'universal ' or ' fo rmal ' pragmatics seeks to define those principles which operate at a necessarily very abstract level, across diverse subject matter, contexts and even cultures. Though much of the impetus and many of its most important concepts derive specifically f rom the phi losophy of language, pragmatics has acquired a distinct and significant life of its own and, as Green (1989) points out, it now lies at the intersection o f a number of fields, including psychology, anthropology, social theory and, sociolinguistics. (See Watzlawick et al., 1967; Adams, 1985; Mey, 1985; Ver0378-2166/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSD1 0378-2166(94)00008-3 118 S. Harris /Journal of Pragmatics 23 (1995) 117-135 schueren and Bertucelli-Papi, 1987; Hickey, 1989; Wilson, 1990; Leech, 1983, and Brown and Levinson, 1987, for diverse versions of pragmatics.) Most influential on linguistics has been the work of Grice, who sets out to establish Cooperation as a general principle governing 'conversation', which is never explicitly defined but interpreted very generally by Grice as communication. The basis for the Cooperative Principle is ultimately, for Grice, what is reasonable (rational) and, indeed, he views 'talking' as " a special case of purposive, indeed rational behaviour" (1975: 47). Echoing Kant, Grice identifies well-known maxims Quantity (being informative), Quality (being truthful), Relations (being relevant) and Manner (being clear) which form the basis of his Cooperative Principle. It seems to me that the influence of Grice on linguistics and other disciplines has probably been disproportionate, given the rather limited scope and substance of his original paper, for several reasons. Firstly, the Cooperative Principle is presupposed to operate at a very general level which, if valid, would provide a very powerful determining principle underlying human communication; and, secondly, the maxims are presented in simple and 'commonsense' terms which make them both appealing and accessible, in contrast to much of the highly technical work that has been done in conjunction with the philosophy of language. But Grice's proposed Cooperative Principle raises a number of important and controversial issues for linguists interested in the social meanings of language, inspiring not only a great many commentaries (not all of them in textbooks on pragmatics), but also a number of interesting attempts to tighten up and make more explicit the maxims in diverse ways by linguists who approach Grice from very different directions (Kempson, 1975; Gazdar, 1979; Leech, 1983; Sperber and Wilson, 1986; Carston, 1988; Green, 1989; Schiffrin, 1990, etc.). However, it is questionable whether cooperation as an abstract concept provides sufficient grounding in itself for interaction. Grice's notion of 'conversation' is vague and undifferentiated, and it is arguable whether it is helpful to focus so exclusively on cooperation as the underlying principle of communication, even if as an abstract principle it makes a convenient and justifiable starting point. Grice's concept of rationality, with its prescriptive and moral overtones, is also open to serious question. Most important is his failure to build into his version of 'universal pragmatics' a social and political dimension, since it is far from certain whether what counts as true, sufficient, relevant and clear (Gricean maxims) can be meaningful apart from the social and political contexts of particular speech acts. As a social theorist rather than a linguist, Habermas attempts to embed the study of language within a more general theory of social action and supplies, in his work, the social and political dimension of 'universal pragmatics' which is missing in Grice. As Thompson (1984: 280) suggests, Habermas defines his task much more broadly than Grice as that of elaborating a theoretical account which will link communicative action to crucial aspects of social reality and so provide a means of comprehending the essential nature of modern industrial societies. Instead of conversational maxims, Habermas proposes a series of 'validity claims', which act, similarly, as a set of general principles on which all communication is based and which can be S. Harris /Journal of Pragrnatics 23 (1995) 117-135 119 called into question and ' redeemed' by interlocutors. ~ In addition to Comprehensibility, which relates to linguistic competence, Grice sets out 'validity claims' based on Truth (agreement based on a truth claim), Truthfulness (agreement based on a speaker's sincerity), and Rightness (agreement based on normative social rules and/or legitimately ordered interpersonal relationships). Like Grice's 'conversational maxims', these 'validity claims' presuppose consensus as the first premise of communication. Habermas also predicates, contrafactually but as a necessary ideal, a speech situation which is oriented to rational argumentation and where access to speech acts is symmetrically distributed, i.e., all participants have equal rights to state, to question, to request, to criticise, etc. For Habermas, the power differential which characterises many contemporary interactional contexts is manifested in linguistic strategies which distort communication by displacing the rationality which should be its centre. Thus, like Grice, but in a more sophisticated and politicised way, Habermas is also committed to 'rationality' as the underlying first premise of communication, which is at its most fundamental and basic level oriented towards reaching understanding. The power and inequality which distort such communication can be explicitly called into question and must be 'justified' with regard to the validity claim of a 'rightness' which is based on social norms. Habermas's commitment to truth goes beyond that of Grice and signifies the promise to attain a rational consensus, especially in the public sphere, in an attempt to link the political and the linguistic in a f undamen ta l way. By focusing o n communication as social action, Habermas foregrounds language not only as the primary means of understanding and consensus but also as the potential instrument of power and inequality. Indeed, in his most recent work, Habermas (1990) has put forward a concept of 'discourse ethics', which shifts the focus of his work on communication more specifically towards a process of moral argumentation. This process locates not only rationality but a "common core of morality in the normative pre-suppositions of communicative interaction" (1990: ix), since moral arguments are seen to be embedded in specific communicative contexts, and thus foregrounds the validity claim of 'rightness'. Habermas sees essential questions of 'justice' as underpinning normative justifications which are linked specifically to the rational argument of those subject to particular social norms. This process of moral argumentation is particularly relevant to situations where social conflicts arise and to institutional discourse. Thus, the concept of 'discourse ethics' allows for the norms of existing institutions to be called into question, and for the intrusion of moral arguments into the world of institutionally ordered relations ( 1990: 107). The purpose of this paper is to examine whether certain central aspects of 'universal pragmatics', specifically 'conversational maxims' and 'validity claims' as put forward respectively by Grice and Habermas, provide a sustainable conceptual framework for analysing strategic discourse recorded in real situations. Neither Grice nor Habermas make use of actual language as data, and only a few of the t For Habermas's own critique of Grice, see Habermas (1984:274-276 and 312). 120 S. Harris / Journal of Pragmatics 23 (1995) 117-135 many studies which have elaborated, interpreted and made reference to the work of both writers have attempted to assess the conceptual usefulness of either 'conversational maxims' or 'validity claims' in a consistent way (see Sarangi and Slembrouck, 1992), involving real speech acts. Indeed, Habermas himself comments that in comparison with the kind of empirical pragmatics which has recently been developed with contributions not only from linguistics but also anthropology, sociology and psychology, "formal (universal) pragmatics which in its reconstructive intention ... is directed to the universal presuppositions of communicative action seems to be hopelessly removed from actual language use." (1984: 328) A number of other researchers in pragmatics (e.g. Leech, 1983: 231; Levinson, 1983: 38; Wilson, 1991) have also maintained that universal pragmatic hypotheses should be tested against a corpus of empirical data and have highlighted the advantages of applying pragmatic theory systematically to real language behaviour, particularly in strategic contexts. The analysis which follows will focus primarily on 'strategic' discourse produced in institutional contexts for the following reasons: (1) The distinction between 'communicative' and 'strategic' language is a crucial one for Habermas. His concept of power and inequality as a 'distortion' of communicative action is one which can be fruitfully examined in conjunction with actual language data which is both accessible and socially significant; (2) Grice has often been criticised for constructing a theory of universal pragmatics which cannot handle unequal encounters which are clearly not cooperative and where the goals of participants conflict in quite obvious ways; (3) Conflicting goals in the institutional contexts selected are power-laden and related to knowledge claims which must be negotiated linguistically in ways which are directly relevant to 'conversational maxims' and 'validity claims'. Moreover, the dimensions of power and inequality in those contexts are clearly visible in both the wider social and political structures and in the immediate situation, and these can be explicitly related to the linguistic choices of the participants involved; (4) Institutional sites are in a mediating position between individual action and social structure (Thompson, 1983) and can provide a useful way of bridging the gap, methodologically, between case study and abstract theory (Agar, 1985; Fisher and Todd, 1986). (5) Habermas' more recent work on 'discourse ethics' is particularly relevant to institutional discourse. The language of authority, in legal and medical discourse especially, is the powerful instrument of the professional, developed over a long period of time and deeply rooted in the social and political structures of a particular society (see Foucault, 1972, 1989; Fisher and Todd, 1986; Goodrich, 1987). Concepts such as 'justice' and 'rights', which are crucial to the legitimation of the legal process, are not only subject to specific definition in legal disS. Harris / Journal of Pragmatics 23 (1995) 117-135 121 course but may be called into question and become the object of moral argumentation. 2. Universal pragmatics and natural language data Using selected examples from a recorded data base of magistrate/defendant and police/suspect discourse, the remainder of this paper will attempt to assess the usefulness of 'conversat ional max ims ' and 'validity claims ' when confronted with empirical data. Three particular propositions are put forward as a means of clarifying the relationship between pragmatics and power that emerges f rom working natural language data and which relate directly to the work of both Grice and Habermas. Each of these propositions will be considered in turn and in some detail. (1) That the asymmetrical distribution o f speech acts as a mode of strategic communication acts to prevent validity claims being raised or challenged except by institutional representatives. Habermas attempts to make a crucial distinction between 'strategic discourse', which he maintains is oriented to success, and 'communicative discourse', which is oriented to reaching an understanding. Strategic discourse is basically instrumental in mode, power-laden and often located in institutional sites. However, as Habermas makes clear, he is not attempting to characterise behaviour dispositions empirically but to define structural properties of reaching understanding in such a way as to enable him to derive the most general pragmatic presuppositions of communicative action. Hence, the distinction between the two types of discourse rests solely on the pretheoretical knowledge of competent speakers, who are intuitively able to distinguish situations in which they are exerting an influence on others from those in which they are attempting to come to a mutual understanding. Habermas clearly asserts the primacy of communicative discourse, linked as it is to a higher level of consensual rationality and to the force of argumentation which individual participants must recognise and accept. "We call an action oriented to success strategic when we consider it under the aspect of following rules of rational choice and assess the efficacy of influencing the decisions of a rational opponent .... By contrast, I shall speak of communicative action whenever the actions of the agents involved are coordinated not through egocentric calculations of success but through acts of reaching understanding. In communicative actions, participants pursue their individual goals under the condition that they harmonise their plans of action on the basis of common situational definitions." (1984: 285-286) Communicative action, for Habermas, is thus distorted by power and inequality, and this is primarily made manifest in discourse through speakers' differing rights to particular speech acts. Strategic discourse is essentially goal directed, bringing power to bear through the systems of linguistic constraints which operate in most institutional contexts. 122 S. Harris l Journal of Pragmatics 23 (1995) 117-135 In reality, institutional discourse is much more complex than this, and it is difficult to argue except at the most abstract theoretical level for a clear distinction between 'instrumental' and 'communicative' action. Most linguistic contexts contain elements of the 'strategic', even those which might be considered the prototype of discourse oriented to reaching an understanding, e.g. academic seminars. Moreover, the power inherent in strategic action, and especially in institutional contexts, is not limited to the differential access to speech acts of various participants, i.e. the power to question, command, criticise, contest, etc. It is not, arguably, even primarily linguistic power. Nevertheless, most institutional discourse does involve restrictions on speakers' rights. The significance of these restrictions is not confined to what Habermas calls 'institutionally bound' speech acts (or Searle's 'declarations'), which are, after all, relatively rare. It is the differential access to such common speech acts such as questions, directives, accusations, assertions which is crucial to understanding how pragmatics and power interact. This can been seen by examining selective instances of courtroom discourse as a type of case study. The data cited here was recorded in the Arrears and Maintenance Court of the Nottinghamshire County Magistrates Courts, the lowest tier of courts in the United Kingdom system. The recordings involved twenty-six cases, with five magistrates. In addition to the clear social class contrasts in this context between magistrates and defendants, all cases have a common goal, and there is a high degree of consistency in the purpose and structure of court sessions which lessens the likelihood of making gross errors of interpretation. The prototypical pattern of speech acts in court is the question/answer sequence, which is indeed common to much institutional discourse (see Agar, 1985). A considerable number of writers (O'Barr, 1982; Philips, 1984; Danet, 1984; Harris, 1984, 1989; Woodbury, 1984; Adelsward et al., 1987) have now suggested that questions in court appear to be used by institutional representatives (i.e. judges, lawyers, magistrates) as a mode of control, making it difficult for non-institutional participants (i.e. defendants, witnesses) to put forward propositions of their own. A defendant or witness who is limited to the role of respondent may find it difficult to raise or challenge validity claims for the following reasons. Firstly, questions sit much less easily with Habermas' concept of a 'validity claim' than do assertions, promises or even directives. It is not merely incidental that his taxonomy excludes them. A question can put forward a truth claim, but only inasmuch as it contains a presupposition or a proposition which is embedded in its interrogative structure. The following questions (all magistrate or clerk utterances addressed to a defendant) provide relevant examples: (a) would you like to explain to the magistrates why since then you've chosen to ignore the order of the court, please (presupposition) (b) who (someone) wrote this letter for you (presupposition) (c) everybody else seems to have done something but you, don't they you didn't choose to pay it off then, did you (proposition + tag) s. Harris ! Journal of Pragrnatics 23 (1995) 117-135 123 Although it is possible for a defendant to raise a validity claim with regard to the above presuppositions/propositions, in effect, interactively, it is extremely difficult; and few defendants in the data base challenge the 'truth claim' of presuppositions or propositions which are embedded in conducive syntactic forms. Secondly, questions in court are often multi-functional. Accusations which a defendant might want to challenge explicitly on the basis of a truth claim are often posed as questions. The following sequence follows an account given by the magistrate addressed to the court, including the defendant, concerning the ' facts ' of the latter 's case, i.e. that he has two fines on which he still owes the court money. M: yes what are you going to do about it D: well uh I would like to ask your worships if it would be all right if I had fourteen days to pay the three pound one and uh and still have to pay the fifty pence on the other your worship M: why is it you 've suddenly become flush with money D: well as I say uh they 've put an increase on and uh but I daren' t draw out again until Monday (they= Social Security) M: how much money have you got on you D: I haven ' t got any on me your worship M: how'd you get here D: I uh got a lift part way here Though on the surface these questions request information, their underlying function is as accusations that the defendant has money which he is wilfully using for purposes other than to pay his fines. While the defendant's responses indicate that he recognises the accusation, he does not directly challenge the validity of the presupposition in the magistrate's second question or the implied accusations which underlie the final two questions. Indeed, it is also interesting to note that these questions function as accusations only in relationship to a specific asymmetrical context and in a questioning sequence. It would be quite possible to pose such questions ( " H o w much money have you got on y ou?" , " H o w ' d you get here?") in a different context without any accusatory intent. Questions used as indirect accusations tend to prevent defendant validity claims (denial of truth) which a direct accusation would make possible. Thirdly, it is illuminating to match Habermas 's validity claim based on ' t ightness' , on normative social rules, with instances of actual discourse. For example, what kind of 'validity claim' is being raised in the following exchange between a clerk of the court and a defendant? C: In May of this year you were before this court for motoring offences when you were fined a total o f thirty-five pounds and for costs fifteen pounds making a total o f fifty and uh you were ordered to pay within twenty-eight days you haven ' t as yet paid anything or been in touch with the court can you explain why you haven ' t paid and what your situation is at the present time D: yes the reason I haven ' t paid is A that I haven ' t got the fifty pounds and also secondly that I feel I was totally unfairly judged to be guilty at the time 124 S. Harris / Journal of Pragmatics 23 (1995) 117-135 Clearly, the validity claim being raised by the defendant in this instance does not relate to the ' t ruth ' of the clerk 's previous assertions. Both participants accept the circumstances and the amount o f the fine and that the defendant has not as yet paid any money. Propositional truth is not in question. Nor is the ' t ruthfulness ' or sincerity o f the clerk being called into question. The validity claim which is raised would intuitively seem to be concerned in some way with ' r ightness ' but not in the sense of the appropriacy of 'social norms ' . The defendant is not challenging the ' r ight ' of the clerk as representative of the court to impose a fine on him but the ' r ightness ' of the action itself on other grounds (unfairness). In other words, he is not calling into question the validity of the clerk 's statements on the basis of a normative rightness which depends on legitimately ordered interpersonal relations but on a matter of ' justice ' . This would seem an example of a defendant explicitly 'moral is ing ' the world of institutionally ordered relations, attempting to engage in moral argumentation (in Habermas ' s terms) from a position o f inequality in terms of power. D: even even the prosecutor at the time at the recess when the two magistrates went out of whom I think your companion was one stated at the time that the case should never have been brought up that he agreed entirely with my action M: hmm we can ' t re-try the case I ' m afraid D: so I ' v e got to pay it and accept injustice M: yes well it 's up to you I say [the alternative could D: [which says a lot for M: very well be prison D: which says a lot for British justice (utterances left out) D: so I just have to accept injustice M: u m um well it 's up to you Mr H uh uh I ' m putting it to you again are you u m are you going to make an offer uh uh to discharge this debt The magistrate here deflects rather than redeems the defendant 's validity claim by a threat (of prison) and by re-defining what is being talked about not as an act o f ' injustice ' but as a 'debt ' to be 'd ischarged ' . Such cases, when the defendant directly challenges the ' justness ' of the legal system are relatively rare, but they do occur. Instances where a magistrate raises a validity claim of ' r ightness ' can be more easily interpreted within Habermas ' s conceptual f ramework and these are more common. Interestingly, such instances often tend to focus on disputes over speakers ' rights to particular speech acts, at least in a court context: