INTRODUCTION
Fairclough 2000 notes that modern social theory has frequently highlighted the influence of language in the process of social change and reproduction, yet it has rarely drawn upon actual language resources or linguistic theory. Equally, while sociolinguistics situates linguistic practice within the domain of the social, it has been accused of only “weakly” making the link with social theory (Bell et al. 1997; Cameron 1990; Cameron et al. 1992; Erickson 2004; Fairclough 1989, 1992, 1995, 2000). Sociolinguists have been accused of borrowing concepts from social theory (or sociology in general) such as class, race, ethnicity, and gender and treating them as if they were “real things,” and hence “reifying social categories” (Poynton 1997:16).
Fairclough's response, operating within the framework of Critical Discourse Analysis (see Wodak et al. 1999, Wodak & Meyer 2001), is to call for a “transdisciplinary” relationship whereby linguistic resources are brought to bear upon issues of power and domination as they operate in sociohistorical contexts (Fairclough 2000). This approach has been criticized, however, from a number of angles. Erickson 2004 suggests that too much attention has been given to broader “global” issues, at the expense of “local” issues. On a similar tack, Schegloff 1997 accuses CDA of failing to work from actual data and their context of production, preferring instead to operate with what Scheuer (2003:143) calls an “overt political agenda” (although see Mayr 2004 below).
Without criticizing CDA directly, the question is whether sociolinguistics needed to develop a specific paradigm for dealing with the issues arising from social theory (Wilson 2001). Despite criticisms of sociolinguistics' interface with social theory, there is a sense in which the “social” has always been important in certain areas of sociolinguistics, and within the work of certain scholars. Gumperz's (1982) work on interactional sociolinguistics, for example, tackled the issue of social inequalities and their reproduction through forms of contextual marking and interpretation (see also Huspek 1991, 1993). Similarly, Hymes's original case for ethnography of speaking castigated the obsession with abstraction and lack of engagement with the role language plays in the structural organization of cultures and societies (Hymes 1974). Although in these cases, and most specifically within the research agenda of Gumperz, there has been a sense that although minorities have been freed from a responsibility for their own social situation, there is a still a sense in which this has “been too concerned with language as an instrument for assimilation to the demands of a capitalist bureaucracy” (Singh et al. 1988:45, cited in Rampton 1995; see also Roberts et al. 1992).
This is an issue for CDA, sociolinguistics, and social theory. Classical social theorists not only constructed “grand narratives” of social structure, but they did so in a determinist manner. Hence, it has not always been easy to see how local actors' discourse could disrupt the power of a dominant order, since they themselves are involved in reproducing that order. Even more open postmodernist theories such as those of Foucault (1972) or Bourdieu (1977, 1990, 1994, 1998) are said to carry forward some level of determinism. Giddens 1993 suggests that Foucault's distribution of power resources within discursive practices is so all-encompassing that it is hard to see how individuals caught up in power vortexes can escape.
However, both within social theory and sociolinguistics we have examples of how this is possible. Scott (1985:xvi, cited in Erickson 2004:136: see also Blommaert et al. 2000) developed the concept of “hidden” as opposed to “public” transcripts to explain one of the ways in which subordinates may express resistance. These transcripts reflect the unspoken or whispered criticism; as Scott puts it, “the foot dragging, dissimulations, desertion, false compliance, pilfering, feigned ignorance, slander, arson, and sabotage” (cited in Erickson 2004:136). Alternatively, the principle is encapsulated in an old Ethiopian proverb: “When the great Lord passes the wise peasant bows deeply, and silently farts” (Scott 1990). Scott offers no linguistic evidence of how this actually operates, however. But we may see something of this in a number of recent sociolinguistic studies (e.g. Gal 1993; Heller 1995, 2001; Mayr 2004; Rampton 1995, 2001; see also chapters in Coupland et al. 2001). Gal notes the way in which dominant language perceptions are undermined through reinterpretive processes at the local level, while Heller and Rampton show the way in which language choices may be used for specific social practices of local support and institutional resistance. Indeed, Heller (2001:213) has gone so far as to suggest that “sociolinguistics is a form of social theory.”
Much work at this level has tended to concentrate on linguistic choice between various different languages and dialects. There is no reason to assume, however, that local resistance could not operate within a single language or dialect. Mayr 2004, for example, applies functional linguistics from a CDA perspective to analyze how discourse in British prisons reveals forms of local resistance. She notes that resistance operates by challenging linguistic choices, by adopting anti languages, and by giving values to what the prison authorities would see as devalued practices. Gal 1993 suggests that in such resistance, devalued practices may be seen as embodying alternate social models of the world. One need not create an anti language, different language (dialect), or code to achieve this however. By “blocking off” institutional or other dominant or state forms of interpretation, and then reinterpreting all messages within an alternate social world, social groups can actually take the state message and turn it against the state, using the same basic language of the state but not the same interpretation. That is, the institutional or state world is simply blocked out, and all actions are reinterpreted and justified within an alternative local model of the world.
In this article we want explore this potential extension of Gal's description of resistance, and we want to do this by placing alongside each other both a close contextual, rhetorical and pragmatic of analysis of data, and the utilization of relevant social theoretic insights that assist in understanding the specific discourse practices under scrutiny. Of course, no one should appropriate theory, in either direction, for its own sake. It is only where sociolinguistic materials or theory complement and enhance social theory that they should be employed, and equally, only where social theory adds to or augments sociolinguistic claims that it should be appropriated for sociolinguistics. In this sense we offer a case study that, we believe, will contribute to the continuing debate on the relationship between sociolinguists and social theory.
THE PRESENT STUDY: LOCAL RESISTANCE IN CONTEXTS OF CONFLICT
In both individual and group senses, one would expect to see resistance operating at a number of levels, and one place they would be particularly highlighted would be in places of social conflict. Specifically interesting here are cases of conflict that arise when the gap between the acts and obligation of authority – what Bourdieu 1994 calls “noblesse oblige” – become so wide that the “habitus” of those under such authority is disrupted and the compliance of both local action and local discourse is no longer guaranteed. This has been seen in several different ways throughout over 30 years of conflict in Northern Ireland (NI), and here we want to look at one way in which local resistance operates in the context of policing and policing change. We are thinking specifically here of the nationalist community's reaction to policing changes that have been instigated by the UK government as part of the Belfast Agreement of 1998. These changes are, in many respects, responses to criticisms from the nationalist community; consequently one could see this as an instance of social change mediated from below. As we will show, however, not all members of the nationalist community are willing to accept such changes; indeed, as we will see, they challenge the assumption that they are changes at all. They do this, we argue, by “pragmatically blocking” state-based sets of beliefs and assumptions, and replacing these with alternate interpretations and models of police behavior. In explaining this, we will specifically consider how the discursive processes of resistance can be understood in terms of Bourdieu's concept of habitus, specifically in the claim that habitus contributes to “determining what determines it” (Bourdieu 1994:104–5; see below), and the further claim that habitus carries within it both personal and collective history.
We are interested, therefore, in the local discursive dimensions of one particular historical slice of context, specifically the changing role of the police and policing practice as part of the ongoing peace process in NI. We want to look at this at a local level by exploring the discourse of policing within Catholic nationalist communities in Belfast. It is not that a related and/or different discourse does not exist in Protestant communities (something we take up elsewhere; Stapleton & Wilson in preparation). Rather we begin here with nationalist communities because it is these communities that have been most active in calls for police reforms, and it is they who would argue that past policing in NI was carried out by a mainly Protestant force representing British state repression. Consequently, their reaction to police reforms is of particular interest not only in itself but also in terms of what Erickson refers to as the local possibilities of social change.
POLICING IN NORTHERN IRELAND: AN OVERVIEW
Policing has long been a contentious issue in NI. In essence, the police force has lacked legitimacy, a problem that has been “closely connected with the absence of political legitimacy for Northern Ireland” (McGarry 2000:176). In many ways, this lack of legitimacy also reflects the challenges inherent in policing a divided society, and, further, the challenges of being closely identified with one sector of that society, as has been the case in NI.
Following the partition of Ireland (1920), the 26 southern counties of Ireland became the Free State in 1921, while, in 1922, Northern Ireland was established as a constituent part of the United Kingdom (see Brewer et al. 1996). From 1922 until 1972, NI was a Protestant state, with Unionists in control of local government. The disadvantaged position of Catholics (Mulholland 2002) fueled existing nationalist opposition to the partition of Ireland on ideological and political grounds. Consequently, the newly formed police force, named the Royal Ulster Constabulary (RUC), was charged with policing a divided society within a contested state, the political legitimacy of which was challenged by a sizable minority of the population (Ellison & Smyth 2000). From its inception, the RUC was primarily an instrument of the state rather than a crime prevention force, with the central function of defending the state against nationalist unrest (Smyth 2002a, 2002b). Not surprisingly, the relationship between the RUC and the NI nationalist community has always been troubled (Ellison 1999), contrasting with its relatively happy relationship with the Protestant/Unionist majority at this time.
During what became known as the Troubles (1969–1998), the RUC struggled to deal with the anarchy and disorder that resulted from extensive violence on the part of the Irish Republican Army (IRA) as well as a number of Protestant paramilitary groups. In 1969, the British Army was sent to NI in an effort control the violence, while direct rule from Westminster was imposed on NI in 1972. The Troubles also saw large casualties among both RUC and army officers, as both forces were increasingly targeted and killed by the IRA. Indeed, even when the worst years of the Troubles had passed, the RUC continued to be seen by many as the enemy of the Catholic/nationalist people, and they continued to be targeted and killed in Republican attacks throughout the 1980s and 1990s.
Since the 1998 signing of the Belfast Agreement, Northern Ireland has been officially engaged in a peace process, which has seen a number of paramilitary groups (including the IRA) declare ceasefires. Policing reform has been a central, if contentious, issue in the peace process. The Independent Policing Commission, chaired by Lord Patten, delivered a report (commonly known as “the Patten Report”) in September 1999, which aimed to establish a new basis for police consent and legitimacy in NI. The core theme of the report is the importance of “community policing” and of establishing policing as a “collective responsibility” within the public arena (Smyth 2002a). The report also proposes a number of recommendations to achieve its principles, such as neutrality, accountability, and compositional representation of the community. Many of these recommendations have subsequently been implemented. For example, the name and symbolism of the force have been changed; in November 2000, it became the Police Service of Northern Ireland (PSNI). Among other reforms, Policing and District Policing Partnership Boards (DPPBs) have been established; and a policy of 50/50 recruitment of Protestants and Catholics has been implemented.
However, the issue of policing remains highly contentious. Many unionists are angry that the name of the RUC has been changed, seeing this as an insult to the memory of the many officers who lost their lives in the fight against terrorism. They are also wary of placing too much power in the hands of community boards, fearing that this could allow terrorist involvement in the police service. Many are also concerned that Protestant applicants are being treated unfairly through the operation of blatant positive discrimination in recruitment. Republicans, on the other hand, are angry that the British government, in its Police Bill and subsequent Implementation Plan, introduced a number of amendments to the original Patten Report, for example the stipulation that anybody convicted of a terrorist offense should be excluded from independent membership of DPPBs. Republicans also claim that the implemented reforms fall well short of the Patten recommendations for community policing, and many continue to see the PSNI as a “political police force.” To date, Sinn Fein, the largest republican political party, refuses to participate in the DPPBs, and republicans generally remain hostile to the PSNI. Hence, policing continues to be a problematic issue in Northern Ireland (Stapleton & Wilson in preparation). Indeed, “in the supercharged political atmosphere … police reform has become the terrain of political struggle” (Smyth 2002b:307).
THE STUDY DATA
The data discussed here are drawn from a year-long study, “Devolution and identity in Northern Ireland,” which was funded by the Economic and Social Research Council. The aim of this study was to examine the way in which individual and community identities are negotiated in response to ongoing sociopolitical change in post-devolution Northern Ireland. Participants were drawn from “core” urban communities in East and West Belfast – local Protestant and Catholic groups, respectively. They were recruited both formally through community groups and informally through networking within the community.
Once established, each group of 5 to 10 people met at roughly monthly intervals over a period of nine months to engage in discussions about the political process and its impacts on their own lives. While these discussions were organized by a researcher from the University of Ulster and were loosely focused on particular sociopolitical developments, the study aimed to gather data approximating as closely as possible “real-life” conversation. This aim was facilitated by the fact that the members of each (community/network-based) group all knew each other very well prior to the study, and the fact that all meetings took place in informal settings (e.g. pubs, community centers). Furthermore, apart from a few “topic-starters,” the researcher made little attempt to control the trajectory of the conversation, but rather allowed it to be guided by participants themselves as they discussed their personal experiences and understandings of the ongoing political and peace processes. This open-ended approach allows us to examine how the newly evolving Northern Irish political discourses are being used at the community (as opposed to the political/institutional) level, as well as how the community members are negotiating their own location within this rapidly changing context (see also Stapleton & Wilson 2004; Wilson & Stapleton 2005, 2006). In the face of shifting dichotomies and certainties, such as those that, in the past, characterized the political terrain of Northern Ireland, the participants must realign existing community-based identities within new discursive contexts. Clearly, the micro-linguistic strategies and practices through which this is effected will reflect macro-level ideologies, discourses, and narratives (Lemke 1995, Wetherell 1998). In this article, however, we are also focally concerned with the ways in which individual and community voices and discursive practice can influence “upward” to the extent that they can actively accept, reconfigure, and/or reject sociopolitical change, with very real implications for the process itself.
The data used in this article are drawn from a discussion by the Catholic (West Belfast) participants concerning the policing situation after police reform (see above). This recording took place in spring 2004, as part of the larger research project, and therefore reflects many of the themes and issues that were recurrent throughout the dataset as a whole (see Wilson & Stapleton 2005, 2006). During this discussion, the participants discussed their own memories and experiences of policing in the past, their current attitudes toward the police, and their perceptions of any changes that may have been effected through the recent reform of Northern Ireland policing.
ANALYSIS
The analysis is based on a number of consecutive, extended extracts drawn from the discussion described above. These extended extracts are presented in the Appendix, which also contains a key to transcription symbols. The main part of the analysis concentrates on the strategy that we are calling “pragmatic blocking.” This works not only to negate, at a conceptual level, the notion of policing change, but also, and repeatedly, as an interactional attempt to reverse the researcher's assumption that change has in fact taken place. Before moving on to the main analysis, however, we will briefly present some of the recurrent rhetorical themes and structures of the participants' discussion. This first part of the analysis is based on a number of shorter extracts drawn from the extended extracts in the Appendix. It is not our aim to analyze these thematics in detail, but rather to demonstrate how they work together to bolster the pragmatic blocking of policing change within a general discourse of resistance.
Thematic and rhetorical structures
The general thematics of the transcribed data above are primarily a discourse of resistance and rejection. Participants are presenting cases that are rhetorically constructed using general and specific topics (topos), inductive and deductive formats, antecedent/consequent clauses (“if X does/does not do Y then Z”), and practical and historical examples from experience. Here are some examples:
In this extract, F3 categorically states that the police are not, and never will be, accepted in her community. She justifies this community stance in terms of the general ineffectiveness of the police, illustrated in the lenient manner in which they deal with the problem of joy-riding. The current lack of community trust, then, is here attributed to an inability on the part of the police to deal with mundane, practical issues; and in this sense, it is to be understood as unremarkable. However, there are also numerous claims throughout the discussion that the lack of community trust is based on something more than practical policing issues – that is, on the historical relationship between the police and the nationalist community. It is in this context that the notion of change is rejected, as in the following extracts.
Here, the women jointly construct a sequence in which the ostensible changes to the police service are minimized (the only difference; nothing else) and presented as superficial – symbolic changes such as vehicles, uniforms, and names. It is perhaps worth noting that symbolic and emblematic displays are invested with huge importance in Northern Ireland (Brown & MacGinty 2003, Wilson & Stapleton 2005). Here, however, the symbolic changes accompanying police reform (which were bitterly contested by unionists) are dismissed as superficial and meaningless. Moreover, the participants present such changes as laughable (it's just funny, their wee white jeep), thereby underlining not only a rejection of these “changes” but also a sense of contempt that they should be presented as such.
In this extract, F1 concedes that some change has, in fact, occurred; for example, the police's manner when dealing with the (nationalist) community members. This is in contrast to the categorical denial of any meaningful change, discussed above. However, these changes are rejected almost as soon as they are acknowledged, on the basis that they come across as contrived, and therefore, not “real.” F1 then draws a contrast between the present attitude of the police and their behavior in the past, the starkness of which gives weight to her claim that their present demeanor cannot be a reflection of how they genuinely feel. Once again then, the concept of policing change – this time an acknowledged and observable change – is dismissed and rejected.
Here, the basis of the researcher's question regarding nationalist community representation is challenged on the grounds that the new service does not, in fact, represent the whole of this community. The fact that certain people (e.g. ex-prisoners) are still excluded from joining the service underlines the claim that even in the face of seeming progress, the community as a whole is not properly represented. While the exclusion of those with a criminal record from police service may seem reasonable in most peaceful societies, it is seen by these participants as a barrier to fair and inclusive policing. Thus, the researcher's concept of change is once again challenged and rejected by the speakers.
Like the example just discussed, this extract also shows a rejection of reforms aimed at proportional community representation. While in the previous extract the speakers highlighted the exclusion of certain members of the nationalist/Catholic community, they argue here that those who have joined the new service are “not real Catholics” but really more akin to unionists. Hence, even if their community representation has increased statistically, the speakers reject the idea that this is, in fact, a meaningful development. Indeed, F4 explicitly states that because these (unionist-minded) Catholics would have joined the “old” police force, then, in effect, nothing's changed as far as community representation is concerned.
This extract encapsulates a number of the key themes of the group's discourse: the historical lack of trust between the nationalist community and the police; the ineffectiveness of apparent reforms, which are seen as superficial and ultimately meaningless; the repeated claim that the police service will never be accepted; and the need to replace conventional policing with an alternative form of community patrol (i.e. community policing). Within this discursive framework there are both an explicit denial that policing changes have occurred, and a call for an entirely “new” form of policing. The latter, of course, is what many perceive (for good or for ill) as having been delivered by the Patten reforms.
Summary of thematics
Reviewing these general rhetorical and argument structures, a number of main themes emerge, which may be summarized as follows:
(a) the ineffectiveness of present policing
(b) the failure of proposed changes to policing
(c) the need to look to an alternative solution
The logic of this seems straightforward. If the police are ineffective, and proposed changes don't work (and in that sense are not changes at all), we need to look for another solution. However, the obvious logical problem with this argument is that it is very similar to that which led to the production of the Patten Report in the first place, and to the implementation of a series of changes to the police force of the kind noted above. The difficulty is that what are seen as changes by Patten, or by the government, or indeed by the Protestant loyalist/unionist community, many of whom protest against these very changes, are not changes for the nationalist community. How, then, do they maintain this position? They do this, as above, by rhetorically reformulating the argument of change, but they also do this in a very specific way. First, they “pragmatically block” off discursive access to the concept of change in one modality, the one understood by the researcher, and this process is then reinforced by the constant reinterpretation and reassessment of “actual” changes as not changes at all. When examples from any set of changes are selected by the interviewer, the informants' rhetorical move is to reassess and reinterpret these as non-changes, argumentatively culminating in a shifted or alternate world view in which there has not been any change. This “blocking” process operates in a number of ways, which we will now explore in more detail.
Blocking: Pragmatic and interactional analysis
Consider F3's response to the researcher's initial question in Extract 1.
Focusing directly on the response itself as an interactional resource, the first thing worth noting is that the response There virtually isn't one, where one is a pro form anaphorically linked to the phrase policing situation, seems to reject the premise of the question – that there is a temporally marked (current, at the moment) policing situation. It is accepted that the adverbial marking of virtually acts as a “hedge.” Hedges come in many shapes or forms, but their job, according to Lakoff 1973, is to make things fuzzy or less fuzzy; that is, they relate to degrees of speaker commitment. Hyland (2000a, 2000b) notes that hedges might be best understood as reflecting an attitude or attitudes not only toward a proposition and its content, but also toward readers or listeners. In F3's turn, virtually modifies the contracted form isn't one, which, as noted, anaphorically stands for police situation. This can then be seen as “There is virtually no police situation.” We want to suggest that, since this negative statement stands in opposition to the interviewer's question, then virtually may be acting as a politeness marker (see Myers 1989) and not, or not centrally, as reflecting a limited commitment to the negative statement itself. After all, F3 is essentially disagreeing with the interviewer's assumptions (and possibly more than that; see below), which is a “face threatening act” (Brown & Levinson 1987); hence one would expect to see a politeness marker in this context (see Schiffrin 2001). Further, given F3's following comments, and other statements throughout the transcript, she does not seem to have an ambivalent attitude to the context of “policing”; indeed, she has a very clear attitude to the context of policing, as elaborated below.
Accepting this argument for the moment, we want to focus on the core inference from the claim that there is limited, if not, in fact, no, “police situation.” A statement such as “There is no police situation” could be considered within a context of negation. The study of negation has a long and complicated history focused mainly on formal and abstract rules for the operation of negation (see Horn 1989, Neale 2000, Van der Sant 1988). A detailed consideration of this would take us well beyond our present concerns. However, Carston (2002:266) summarizes two main themes within the literature on negation: (i) the scope distinction and (ii) the representation distinction. The scope distinction relates to whether the negative operator should be seen as having a “wide” or “narrow” scope reading – that is, whether it might operate over the whole sentence or another smaller structural element, as illustrated in the following example.
Under (i), if there is no linguistics professor, then the color of his hair is irrelevant. Under (ii), what is negated is only the supposed color of the professor's hair; red vs. black. The representation distinction separates two possible types. The following examples are taken from Carston 2002:27.
In the (a) sentences what is negated (the predicate within its scope) is descriptively representative of some aspect of the world, and the following statement is consistent with this reading (Not P; Q). In the (b) sentences a similar analysis would lead to contradiction; that is, P is negated and then P is affirmed (Not P; P). In this case what is happening is that an objection is being made about a “non-descriptive” aspect of the prior utterance; in one case (2) the inappropriate plural, and in the other (3) that a weaker adjective has been used.
In the case of “there is no police situation,” this is a response to the interviewer's comments and acts to negate their claim. In this sense, the issue seems to be one of representation.
But this is (P;Not P) as opposed to (Not P;Q). As such, therefore, it is simply a negation of the affirmative claim made by the interviewer. So both turns combined give us a contradiction:
But the question then is: What is being objected to? And what is being corrected? At best, we can speculate on what “police situation” means for both the interviewer and F3. But the case is much less clear-cut than that found in many of the constructed examples from the formal debate on negation. While we might consider that “police situation” has different meanings for the interviewer and for F3, there would still be something odd here. F3 says there is not a “police situation” but that the police are not accepted. However, a normative interpretation of either a context where a community has problems with the police or a context where the police have problems with the community would seem quite logically to be a “situation.”
To try to feel our way a bit further toward a possible solution, let us move with the data themselves. F3 has implied that there is no police situation and then has gone on to note not only that the police are not accepted, but that they have never been accepted (proved themselves) and, in the respondent's view, they are unlikely to ever be accepted. Consequently, if the police are rejected, have always been rejected, and will always be rejected, then there is “no police situation.” In this case the important thing is that nothing has changed. For the interviewer, “police situation” is seen as related to “police change”; for F3 there has been no “police change,” therefore there is no situation. Hence we could have:
Yet, once again, when we look at the rest of the data, time is spent marshaling arguments against forms of descriptive change within the police that are accepted by the informants, but that are reassessed against some normative view of “real” change and are found not to match. Are we caught up here in a circle of the semantics of the words “change” and “situation,” or is there something more to this? Perhaps, as Gal's (1993) mode of resistance implies, we may be trying to assess the construction of an alternative model of the world as a form of resistance within the researcher's social and different (perhaps state) view of the world. In the next section we explore this possibility further when we look at the interaction under the guise of a question/answer interaction.
Questions, answers, and accounts
Returning to the data, consider F3's comments in terms of their positioning as a response to the researcher's question. As already noted, the respondent seems to reject the premise of the question. This has both formal and interactional consequences. First, questions are normally asked when the speaker assumes that the hearer has access to information that will allow him or her to answer the question (Dillon 1990, Hargie et al. 1994). There are exceptions to this, of course, with specific, contextually realized types of questions as found in classroom teaching, doctor/patient interactions, courtrooms, parliamentary interactions, and so on (see Wilson 1990, chapter 6; also Clayman 2001 on evasions). In general, however, we ask questions to get information, and we want that information for a purpose of some kind. There are, of course, instances in which the question can misfire, as in the case where the recipient does not have access to the information, or even where the recipient rejects the question by challenging the question, and hence potentially challenging the questioner himself.
F3 does not seem to be rejecting the question in this way; she seems to suggest, with the rest of her response, that the assumption underlying the question – that there is a developing or changing police situation – is basically false; hence it would be difficult to answer the question. But there may be more going on here at a formal level. Consider the following examples:
In these constructed examples, the response seems, once again, to reject an underlying premise of the question, specifically that there is news in the newspapers, there are movies at the cinema, or there is food available. Looking at elliptical expansion, we can see that what is claimed is:
However, this cannot be correct. It is extremely doubtful that:
So what is going on? There are at least two possible, and interrelated, ways of explaining the issue. The first considers issues of pragmatic processing and relevance (see Carston 2002), and the second focuses more directly on general conversational structure. In the pragmatic case, one could invoke a Gricean or neo-Gricean perspective, which looks at issues of communicative relevance (Carston 2005; Grice 1975, 1989; Levinson 2000; Sperber & Wilson 1986). Grice's well-known theory of meaning distinguishes between “what is said” and “what is meant.” He suggests that when we communicate we do so within a general, overarching Cooperative Principle:
Make your contribution such as is required at the stage at which it occurs, by the accepted purpose and direction of the talk exchange in which you are engaged.
This principle in turn has a number of sub-maxims:
- Quality: Speak truthfully.
- Quantity: Make your contribution as informative as required.
- Relevance: Make your contributions relevant.
- Manner: Be brief, clear, orderly, and unambiguous.
Grice did not claim that we follow these rules all the time; indeed, it is when we flout the rules that things get interesting. Consider the following:
Assuming that the respondent answers the question within a generally Gricean frame, she has not given a direct answer and has flouted one or more than one of the maxims. In such contexts, Grice tells us, we should look for inferences that he refers to as “conversational implicatures.” These are forms of meaning we can work out from the context and the maxims; in this case, since the answer does not refer to any significant aspect of “good teaching,” the respondent means that she does not think that Bill is good teacher. In the sample cases, it is not clear what the relevance of “nothing” is in relation to the aims of the question. Consequently, questioners could infer that they should seek out a contextual analysis that would allow them to derive a conversational implicature. Here, if we assume that the speakers have a motive in asking the questions – for example, that they would like to discuss today's news, go to the cinema, or have dinner – then here the answer might indicate, from the hearers' viewpoint, that they don't want to do any of these things.
An alternative but related way of considering the same issue would be through Sperber & Wilson's (1986) Relevance Theoretic approach. This develops the general insights of Grice within an overall processing-based view of communication. In this model, communicators are expected to maintain a theory of relevance where a balance is achieved between the amount of processing effort and the information gained. The aim is to minimize processing cost to obtain maximal informational gain. Certainly, in the Gricean analysis above we do have to do a lot of work to find out that the hearer does not want to go to the cinema and so on. Employing relevance theory, we could first suggest that “nothing,” as a subsentential form, may be enriched by such things as syntactic ellipsis (see above) to give us what are called “explicatures.” Explicatures were developed in analogy with Gricean implicatures, only in this case the term refers to a development of linguistically encoded information in a specifically limited contextual set. As Carston (2005:2) suggests of explicature:
It is a term belonging to a theory of communication and interpretation and it is distinguished from most uses of the term “what is said”, in that it involves a considerable component of pragmatically derived meaning, in addition to linguistically encoded meaning. A key feature in the derivation of an explicature is that it may require “free” enrichment, that is, the incorporation of conceptual material that is wholly pragmatically inferred.
Such inferences are normally reached by a pragmatic expansion of logical form. So in the case of 10a, below, the general explicature would be that the relative amount of time to be taken will be significant, and in the case of 10b, that the speaker will have eaten breakfast that day (see Carston 2002).
Explicatures have been developed to account for pragmatically formed information directed through linguistic enrichment, and it has been suggested that basic forms such as “nothing” may be further linguistically developed by processes such as “pragmatic narrowing” and “pragmatic expansion.” Consider the following example from Carston (2005:23), which is a real example taken from the O. J. Simpson trial.
As Carston notes, this statement is linguistically a contradiction on any formal analysis, but it was clearly the case that the speaker neither intended it nor understood it as a contradiction. To explain this, Carston argues that the same lexical form upset was being used differently at each occurrence (hence the asterisk distinction). She goes on to suggest that the lexical concept UPSET requires a pragmatic narrowing to produce implications from the second occurrence that are not carried by the first. To quote her directly:
The two instances of the word “upset” were interpreted as communicating two different concepts of upsetness (as indicated by the asterisks), at least one, but most involving a pragmatic narrowing of the lexical concept UPSET; the second of the two concepts carries certain implications (e.g. that he was in a murderous state of mind) that the first one does not, implications whose applicability to Simpson Kato wants to deny. The proposition explicitly expressed here is true just in case O. J. Simpson had one sort of property at the time in question, but lacked another, related but stronger, property.
The upshot of all this, claims Carston, is that a single lexical item could communicate a wide range of concepts.
Returning then to the case of “nothing” in the examples above, what could it be communicating? The answer that seems plausible in such contexts is that the response means there is nothing within the papers that the respondent wishes to talk about, no movies at the cinema he wishes to see, and, of the food which might be available, there is none that he wishes to eat. In all cases “nothing” blocks off any future direction of talk related to the original concepts of the question – news, cinema, and food. But how does it do this? According to Relevance Theory, processing the response to balance minimal effort with maximal information leads to a clash between an absurd claim and a relevant response. This clash is resolved by enriching the lexical concept “nothing” in the appropriate contexts, such that it acts as blocking any news, movies, or food. The communicated information is that the respondent does not want to explore/discuss the news, view any available movies, or eat any food in the house. Although this is not a dissimilar conclusion to that provided by our Gricean analysis, it has a parsimony and logic that need not appeal initially to the assumed motives of the questioners themselves (see Carston 2002, chapter 2 for a detailed discussion of this issue).
There is another and related way of looking at this issue, however. Sacks 1998 notes that it is generally normative that following an answer, the person who asks the question has the right to talk again. As Sacks concedes, there are exceptions to this latter case, as in press conferences or classrooms, but we understand the exception by the very way in which the context constrains the operation of the rule. Given both the expectation of a response (answer) to a question and the expectation that the questioner may talk again, with a response such as “nothing,” it is hard to see what the questioner would talk about without shifting topic altogether, or challenging the response itself. Further in his work, Sacks (1998, Lecture 3: drawn to our attention by referee comments) also focuses on responses that seem to behave in a manner similar to what we view as “pragmatic blocking,” what he calls an “account apparently appropriate, negativer” (Sacks 1998:23). The example given by Sacks emerges from a call to a suicide prevention center in which it is revealed that the caller has a gun in the house. The desk attempts to get an account of why there is a gun in the house, and the following response occurs: “Everyone does don't they.” Sacks suggests that we have here a member of a set of things that “cuts off the basis for the search for an account.” This is discussed further by Pomerantz 1986 where she suggests that in specific adversarial contexts, participants can make use of what she calls “extreme case formulations” “when they anticipate or expect the co-interactants to undermine their claims.” Clearly, “extreme case formulations” (ECFs) do seem to block off the direction of future talk, and they may do this for a variety of purposes (see e.g. Sidnell 2004). Such ECFs, however, differ in a number of ways from pragmatic blocking as defined in the present article. First, ECFs, as noted by Pomerantz, arise in adversarial contexts. The assumption is that the speaker is being criticized or attacked. In the case of the QA examples given above, there is no suggestion of an adversarial base to the interaction: Asking someone what is on at the cinema could hardly be seen as an attempt to undermine anyone's claims. Indeed, if anything, it is the responses that are challenging, rather than the questions.
There is also a more important distinction to be made between ECFs and pragmatic blocking. Sacks's and Pomerantz's claims are that ECFs cut off the basis for providing an account. In our data F3's comment generates, or is followed by, a lengthy interaction in which there is a co-construction of why the police have not really changed. Hence, it is not the case that F3 is cutting off an account or accounts; rather, she is correcting or adjusting the assumptions explicit/implicit in the interviewer's talk. Here we have a classic case of linguistic “metarepresentation” (Sperber 2000; see above), in which some aspect of prior talk is presented and commented upon. When applied to contexts of negation of the type we have been discussing, the term “metalinguistic negation” is used (and several of the negation examples discussed previously would fall within this category). If we look at metalinguistic negation, Horn (1985:121) notes that one can negate almost any level of linguistic structure from phonology to pragmatics: “It is a device for objecting to a previous utterance on any grounds whatsoever – including its conventional or conversational implicature, its morphology, its style or register, or its phonetic realisation.”
Example (12a) objects to a specific pronunciation, (12b) objects to a specific quantity implicature (see Gazdar 1979) and (12c) objects to the presupposition of the first clause, that there is a king of France.
Viewed in this way, F3's comment is blocking assumptions that it objects to, specifically, not claiming that there has not been change, but rather claiming that there has not been the right kind of change, as explained in the rest of the talk. Seen thus, one could suggest that ECFs may also be a type of metarepresentational behaviour, whereby the undermining assumptions of prior talk are dismissed.
Pragmatic blocking is a process, then, of challenging explicit/implicit assumptions in prior talk, and of offering an account for that challenge. This account need not itself be explicit. In the case of the QA samples above, the formulaic structure of the responses allows for an enrichment analysis of logical form and context that provides an account: There is nothing of interest to X in the newspaper or cinema; there is no food presently available that is to X's liking. So is there virtually isn't one meant as blocking? In one sense, the rest of the turn would support this assessment, since the implication of the rest of the response is that nothing has changed with the police so there is no “at the moment” context within which to consider the policing situation. This works despite the use of virtually as a hedge. Similar hedges could be used in our constructed examples 6–8 above: “virtually nothing,” “hardly anything to eat,” “not really anything on at the cinema.” In these cases, there would still be a clear indication that the respondent is not keen on going to the cinema, discussing any news, or eating any food.
But isn't this odd, given that not only does F3 go on to talk about the police, but so do all the other respondents? Well, no, and now we can see why not. To simply challenge the interviewer's assumptions that there has been change would be to create a context of (P; Not P). And this tells us nothing. It is only where any inferences derived from the use of “change” (situation) suggest fundamental or core differences that are not accepted within the alternate world of the informants that there is a requirement for (Not P) to be expanded with an account. These accounts explain that the interviewer's original assertion (P) is, in fact, wrong. Hence, what is being discussed is not a contradiction (P; Not P), but simply (Not P): there has been no policing change, and this is why (P) is blocked. The blocking move plus the hedge given by F3 allows the informants to talk about the police (and hence to do the opposite of what is expected following extreme case formulations), but we have been given a very clear indication of what is to come by initial use of a blocking turn. If we treat “police situation” as meaning the present changes, then it is the view of F3 and, as we see, all the others, that there are no “real” changes, and hence there is no situation to discuss. What is then expanded upon is not a police situation per se, but a reflexive consideration of why the blocking occurs.
Finally, consider the way in which F3 takes her own lead and provides in her turn continuation the following:
Here, F3 gives us a first example of the rhetorical reassessment that underlies the “blocking.” Notice here the use of the focus particle even (see Boguslavsky 2001, Handley & Feeney 2004, Horn 1989, Karttunen & Peters 1979, Lycan 1991). This form is used in very specific circumstances, normally to mark a member of a set of possibilities as one of surprise. To take one example:
It is claimed in the literature that even picks out an extreme position on a pragmatic probability scale (see Fauconnier 1975). Even directs the listener's attention to a set of related statements that are more probable than the one expressed – X likes Bill, Y likes Bill, Z likes Bill – which are higher on the scale than John likes Bill, but taken all together this directs the listener to the inference that Bill is very likeable. This may be represented, following Boguslavsky (2001:31), as follows:
In less formal terms, the expectation that John likes Bill goes against what we would predict in the case of John; which further strengthens what we wish to express about Bill, i.e. his likeability (see Francescotti 1995).
Now in the case of our example above, this statement occurs following F3's blocking of the set of assumptions about police change that might normally have gone through. Her use of even draws our attention to some expectation that will be contravened. But what is this? Given that we know the police catch the joy-riders, it is not that; it must be rather that they do nothing about this, they are ineffective. If there were a set of issues more complex than joy-riding, say murder, rape, burglary, theft, or sectarian violence, then, on a scale, joy-riding would or could be seen as a simple issue, one that we might expect the police to deal with, particularly since they can catch the joy-riders. Hence the expectation contravened is that the police could do something about joy-riding, but this is not the case:
Even plays its part in F3's emerging account of the failures of the police, and these are built on by adding incrementally in each case the weakness of the police and offering alternative interpretations for any action that might look like change. Discussion of the police is allowed to proceed, but only in a framework where there has been no change, and where any mistaken view of surface change is corrected. The bottom line is that there has been no change, and discussion to the contrary is blocked.
Sequential blocking
Consider now the following list of examples, some drawn from the data discussed above; some drawn from elsewhere in the transcript.
There are two things to note. The first is that in some of these cases the blocking is quite straightforward and explicit. In example (A.i), for instance, the response no, simply and baldly “on record,” closes off certain claims and assumptions. It does this by quite openly rejecting these. This could be seen as the most basic (or explicit) form of blocking. When the researcher recycles her turn and tries again to achieve some recognition that there has been change, this gets the response that it's [just] funny. Efforts at change are “laughable” and certainly not to be taken seriously, or even discussed as changes. Later in the same transcript (A.ii), the researcher tries again to suggest that change may have taken place, and here again we have a bald on-record rejection, a blocking that is of the possibility of change and the discussion of change, at least as it might be understood by the researcher.
The second issue of interest is the way in which there is a repeated pattern of agreement that arises actively and enthusiastically following statements of the failure of the police and that they will never be accepted (examples B–D). This agreement is marked by forms sometimes referred to as “non-transactional” (McCarthy 2003), those elements of “small talk” (J. Coupland 2000) such as mm, huh, well, yeah, oh, and no. While in early research these forms were often treated as simply providing “back-channel” information (Yngve 1970), Schegloff 1982 has argued that these response tokens have a multifunctionality and may be doing several things at once. They may play a role in the turn organization of the interaction; they can acknowledge a prior turn and offer a receipt; they may provide closure; and they may signal understanding and agreement. It is here, in the area of agreement, that the forms seem to function most clearly in the above examples – projecting agreement that there has been no change, but also providing a point of group convergence.
This is particularly the case when a second-turn agreement is reinforced by a third-turn matching agreement. The pattern is mantra-like, where a core police strap-line – the police will never be accepted in this community – is a view held by everybody, and reinforced by everybody. The agreements have a “relational function” in the talk (Candlin 2000:xv); that is, these repeated and supporting moves are structurally limited and function to offer support for the previous move of a member of the group. Given that the core alternative model of the world adopted by the group is more or less accepted by all, these agreement markers, we would suggest, are mainly group ties and are not built on any core content analysis of the previous turn. As Sapir (1921; cited in Sacks, 1998:429) noted some time ago, a sentence such as I had breakfast this morning might simply be a reflection from memory, a “habitual expression,” but it need not suggest that someone is in “the throes of laborious thought.” He notes that “each element in the sentence defines a separate concept or conceptual relation or both combined, but the sentence as a whole has no conceptual significance whatsoever.”
Sacks 1998 criticizes this assumption, suggesting that Sapir is cutting off an avenue of analysis: the structural role of such turns. There is some truth in this, but what Sapir says need not be incompatible with Sacks's thinking. As noted above, Schegloff's multifunctionality allows for more than one role for such an utterance, and it allows the utterance to operate at different levels. If the agreements above function purely interactionally and not fully intentionally, for example, then their core content may indeed – as Sapir predicts – have no “significance” in this particular context of expression (see also Coulter 1999, 2005 for a radical account of language without thought). However, as we have noted, we don't have to settle for an either/or conclusion; multifunctionality allows an analysis on several levels (not always simultaneously, of course). What we have to consider is what it buys us, in explanatory terms, to think of the agreement markers as reactive rather than necessarily intentionally reflective.
In the case of the agreements, we see these as behavioral reactions to specific group “triggers”: as rule-following where everybody agrees that there has been no change, and that the police are rejected. Any statement to that effect gets support, and the support gets support. This is social interaction within the alternative social model of the local group (see Gal's remarks quoted above). In much they same way as silence within conversation often triggers general comments such as isn't it warm in here or your shoes are nice, or similar forms such as we find in “phatic” interaction at the bus stop or in a lift, the above agreement tokens act centrally as group support and may be more easily understood in this way than in terms of intentional content. However, this is not to argue that they cannot have intentional content. Rather, in agreement with Weinreich (as cited in Sacks 1998:429), we believe it is possible that in selected interactional contexts speech can become “desemantized to a formidable extent.”
Returning to the data, there is little doubt that the patterns of agreement above reflect participants' orientations to one another's actions. These orientations are, we would argue, framed mainly within the talk; that is, they function centrally as interaction markers. It is in this sense that we wish to see them in opposition to the reflexive accounts that also occur in the data. The reflexive accounts are contextually realized representational analyses of a social context of community operation. The patterned mantras of resistance and rejection are forms of interactionally reactive and binding language (at one level), community responses that are rule-governed and meaningful, but not necessarily transactional – that is, not requiring unpacking or interpretive responses.
One finds similar patterns in various contexts of resistance: striking miners' chants of “Maggie, Maggie, Maggie, out, out, out”; some unionists' reactions to the signing of the Belfast agreement, “Belfast says no”; or as articulated by the DUP leader Ian Paisley, “No never, never, never.” These forms are individual strap-lines for particular political positions. What we are witnessing above is the coarticulation of a community strap-line of rejection, “no police (RUC/PSNI) here.” Underlying any strap-line is an unpacked argument of some kind. This unpacked argument, in the case of the nationalist community's discourse, is what we have in the reflexive and reflective content of the strap-line position. Here one simply needs to understand how to position oneself structurally at an appropriate moment in relation to others to be part of the same community articulation of meaning. Here we have a linguistic construction of habitus, an unconscious form of linguistic behavior reflecting participant understanding of local practice.
SOCIAL THEORY AND SOCIOLINGUISTICS
We have argued above that the configuration of discourse patterns in focus-group discussions of policing reflects a subset of nationalist thinking. The reflexive and reflective level of argumentation underpins a set of dispositions that orient toward a rejection of the police. When this rejection is brought to consciousness in a focus-group context, informants utilize various rhetorical devices and logical argument structures to justify their dispositional attitudes in the context of policing. We want to look, now, at how this pattern might be explained or understood within social theory, and to do this we will draw directly on the social theoretic model of habitus as developed by Bourdieu 1994.
Habitus theorizes social identity as “socialised subjectivity” (Bourdieu & Wacquant 1992:126). This is realized in a set of behavioral dispositions operating in formats of speech, dress, movement, education, sport, politics, and so on. Habitus is, then, a form of practical action within a commonsense reasoning that allows individuals to position themselves within social life worlds, or “fields,” as Bourdieu refers to them. Habitus describes a collective relation for individuals in which each is “the individual trace of an entire collective history” (Bourdieu 1990:56). This does not mean a strict form of determinism, however. Bourdieu rejects an interpretation of a mindless behavioral response. Social practice is at once rule-bound and open-ended, a “regularised improvisation” that is a “conditioned and conditional freedom” (Bourdieu 1977:95).
Change occurs as an outcome of the competition for various forms of capital – for example, symbolic, political, social, or cultural capital. This competition may be worked out in a variety of ways, but the one that concerns us is the competitive context of access to human rights and rights to identity. Clearly, sections of the nationalist community believe that this is what they have been fighting for throughout the Troubles (see Wilson & Stapleton 2005, 2006). Honneth (2003:131) argues that individuals and groups in society have normative expectations of how social order should operate, and that “what subjects expect of society is above all recognition of their identity claims.” When a society fails to achieve this, or more starkly, attempts to institutionalize asymmetric control of and access to resources for some groups as opposed to others, without any clear rational justification, then it is here that subjects are motivated to act for justice. This description is not dissimilar to the way in which Bourdieu describes the rupture that can occur between the habitus of the dominated and the behavior of the dominant. Where the expectations of the behavior of the dominant fail to operate as they should, the habitus of the dominated may become shifted. Thus habitus may become a site of both struggle and creative potential as “identities are continually being re-produced as responses to social positionings” (Skeggs 1997:2).
The Troubles in Northern Ireland arose, in part, as the result of the dislocation of habitus relations in the competition for identity and rights (see above). The perceived Protestant hegemony of NI allowed the dominant to ignore normative habitus relations with the dominated nationalist community. Eventually, as Bourdieu predicts, this led to a rupture of the relationship between such habituses. In such an outcome the dominated then reject the behaviors of the dominant and the behaviors they themselves have been assigned in the dominant/dominated relationship. In so doing, they seek out alternative dispositions and orientations, and derive an alternative habitus – in this case, a habitus of resistance that is reflected in the individual and collective behaviors of the community.
This seems to be what is taking or has taken place for some of the nationalist community. As we noted earlier, the police in NI were (and, in some communities, still are) seen as anti-Catholic and anti-nationalist. For nationalists, the previous behavior (and, in the view of some, the present behavior) of the police reflected a clear bias against one section of the NI community. Consequently, the habitus of sections of the nationalist community, as it operated in the field of interaction with the police, was brought into conflict. This conflict arose from the failure of the police to perform to normative expectations, specifically that all communities and individuals should be treated equally under the law. Normal communities and community members, regardless of their feelings toward the police, recognize the behavioral contract between both sides. The police are there to protect and serve the community without fear or favor. In return, the community is expected to support the police through generally law-abiding activities.
In the context of NI, however, the nationalist assessment of the history of policing suggests that police behaviors (police habitus) shifted to reflect bias; hence the compliance behavioral contract the police had with the community was broken. As a result, community habitus was displaced, and there was a need to shift to a different set of dispositions that reflected a habitus of conflict and resistance. This is what we are seeing at the level of discourse as revealed in the sociolinguistic patterns analyzed above. The reflexive analysis generated by the focus-group context underpins a subconscious disposition of community members, which becomes reflected in a sequential coordination of agreement in rejection. The habitus of certain sections of the nationalist community is anti-police, and this is presently passed on to each new generation. Here we see in action what Bourdieu calls the “creativity” of habitus. In a context of disruption, new dispositions may emerge. However, they emerge only from the understanding of a conflict within habitus, and they are hence born of what previous dispositions have creatively reconfigured for the present and future circumstances of interaction.
So what can the police or government do? How does one reorient the nationalist communities' habitus in the field of police interaction? As we have seen, attempts to present changes to policing within the Patten Report are being reinterpreted as cosmetic and as not as dealing with the real problem, the police's institutional attitude to the nationalist community. It is for this reason that the informants call for their own community police, since this will give them people they know and can trust. Consequently, there is perhaps no single thing the police or government can do at present. The move toward community Policing Boards is helpful, but at present, the main republican political party, Sinn Fein, refuses to sit on such boards. Such a disposition is to be expected while the habitus of resistance and rejection remains in the community. After all, Sinn Fein represents the largest section of the nationalist/republican community. This raises the question of ordering of events: Are the politics of the community the politics of Sinn Fein; or is it that the politics of Sinn Fein are the politics of the community? Clearly, both are intertwined, and if the government could convince Sinn Fein to take part in Policing Boards, it would be a major step forward, and for individuals in the community, it would begin to challenge the present habitus of resistance and rejection (and this is now beginning to take place).
We noted at the outset of this essay that theories should not be appropriated on an ad hoc basis, so what is the balance of benefit for sociolinguistics and social theory? First, the patterns that we have highlighted within the data remain descriptions without a socially theorized account of such sociolinguistic behaviors. We can see from the utilization of habitus, within contextualized historical circumstances, that certain discourse patterns within the nationalist community in Northern Ireland operate to realize specific dispositions in the dominated/dominant relationship. These dispositions are such that they resist and reject normative societal standards in the field of policing (and beyond). The dispositions have emerged from the fissure that derived from the conflict over identity rights within society in general and within the police specifically. The behavioral shifts occurred at a number of levels, including street rioting, refusal to recognize the police as a legitimate force, and, of course, sociolinguistic behaviors at subconscious and conscious levels. From a social theory perspective, there is also a payoff. Consider, for example, that Bourdieu's theory, despite his protestations, has been criticized as being deterministic. Further, he has been criticized for not giving clear accounts of habitus in action. In our study we show how the community has sociolinguistically adapted to the creative potential provided by habitus, and at once, therefore, provided a real-world example of both habitus itself and the flexibility of habitus.
But what of “change” itself? What is social policy practice to make of “conditioned and conditional freedom?” If, as we have explained, state representations of change become reformulated as false change, or no change at all, how can we progress from a position of conflict and resistance to one of tolerance and understanding? This is the seeming paradox of the creative potential of habitus. Change is both material and linguistic, although these may not operate together at the same time; hence the disruption we are discussing. The Patten Report has so far produced mainly material changes, and less change in terms of the discourse of policing. Of course, it is not so simple to effect the latter where several audiences are concerned. Nevertheless, as we have seen, change must not only be practical but practiced, and it is here that the community rejects policy and policing. In simple terms, the discourse of policing has not changed; there is no community voice to articulate the alternate world they inhabit. Progress involves linguistic reformulations negotiated out of the different interpretive models, and this requires negotiation. It is here that change will be eventually coarticulated. But first there must be “talk,” and there must be agreement to “talk.” We are not quite there yet, but when that point is reached we predict a struggle over the resource of language to articulate an agreed position – a position, that is, where a shifted set of dispositions emerge, within a changed habitus reflecting an alternative set of dispositions articulated in a different discourse of policing.
CONCLUSION
One response to the criticism of sociolinguistic theory – that it reifies or pays insufficient attention to social categories – has been the emergence of a strong and growing literature linking language and social theory. Much of this is top-down rather than bottom-up, however, and most work begins with social theoretical concerns first and then indicates the ways in which these may be enhanced using selected forms of sociolinguistic analysis. Within the broad remit of interactional sociolinguistics, there has always been some level of consideration of the potential analytic synergy between social theory and sociolinguistics. This article continues to develop this approach by arguing first for an analysis of sociolinguistic materials, and then considering how findings may be enhanced (interactively) by social theoretic considerations. Looking specifically at the discursive dimensions of community views on policing in Northern Ireland, we have attempted to provide an example of how this synergy might operate. We have come at the issue from the position of the data first, and attempted to show how particular discursive patterns may be further understood by reflecting on their broader social theoretic implications.
Working from what we have called “pragmatic blocking” within the data, we indicate how this maps with the concept of dispositional attitudes found within Bourdieu's theory of habitus. “Pragmatic blocking” becomes a discourse manifestation of the “determined and determining” nature of habitus, while at the same time reflecting, through the structural flexibility of “blocking,” opportunities for creativity under constrained conditions. One such condition arises from the disruptive context found when the dominated/dominant contract of police/state and community is broken. Here we may find the dominated rejecting material and linguistic forms of domination and setting up alternative discursive forms of resistance. Here, sociolinguistic choices constrain the way in which particular social worlds are represented, and indeed which social worlds are legitimated for representation. We have provided evidence of this in the context of policing and policing change in Northern Ireland. The nationalist community has adopted a habitus of resistance and rejection, and this may be seen to operate at a subconscious level of behavioral dispositions toward the police. This is reflected in the way in which certain discourse formations are rejected or reinterpreted within limiting frames of interpretation of police action. This occurs within both the broad reflexive formulation of rhetorical accounts of resistance, and through the use of pragmatic and conversational features of interactional structure.
HISTORICAL FOOTNOTE
Since this paper was written, there has been a major shift in the stance of political republicans to policing. In January 2007, Sinn Fein decided to conditionally endorse the PSNI, in part fulfillment of the 2006 St. Andrew's Agreement (aimed at restoring the suspended NI Assembly). In light of our analysis above, this development raises interesting questions: in particular, the extent to which ‘top-down’ acceptance of the PSNI by politicians can work to produce a new form of habitus within the republican community, or, conversely, to cause new forms of schism and resistance.
APPENDIX 1: EXTENDED EXTRACTS
Extract 1
Extract 2
Extract 3
Extract 4
APPENDIX 2: KEY TO TRANSCRIPTION SYMBOLS
