Cambridge Catalogue  
  • Help
Home > Catalogue > Discursive Research in Practice
Discursive Research in Practice
Google Book Search

Search this book


  • 6 b/w illus.
  • Page extent: 332 pages
  • Size: 228 x 152 mm
  • Weight: 0.654 kg

Library of Congress

  • Dewey number: 150.19/8
  • Dewey version: 22
  • LC Classification: BF201.3 .D572 2007
  • LC Subject headings:
    • Discursive psychology

Library of Congress Record


 (ISBN-13: 9780521849296)

Discursive Research in Practice Cambridge University Press

1       Discursive research: themes and debates

          Alexa Hepburn and Sally Wiggins

There has been a quiet revolution in the social sciences. Over the past few decades new ways of working and new ways of conceiving the relation between people, practices and institutions have been developed. These have started to make possible an understanding of human conduct in complex situations that is distinct from the traditional conceptions offered by disciplines such as psychology and sociology. This distinctiveness is derived from the sophisticated analytic approach to social action that has been developed by conversation analysis combined with the fresh treatment of mind, cognition and personality developed in discursive psychology. Both of these approaches work with the displayed perspectives of participants in interaction, perspectives embodied in people’s constructions and orientations. In addition, this research has exploited the new recording technology and representational forms that enable it to engage more immediately with human practices; that is, to study ‘the world as it happens’ (Boden, 1990) instead of working through the mediation of interviews, questionnaires or ethnographic field notes. This work offers a sophisticated and theoretically nuanced empiricism that focuses on discourse as the central medium for action, psychology and understanding.

This book brings together researchers who have been doing discourse research in this new tradition. It features well-known contributors, some of them pioneers in their field, as well as exciting new researchers who are still early in their careers. Most come from the fields of discursive psychology and conversation analysis. It provides a range of analyses, which illustrate and exemplify new ways in which institutional and everyday settings can be researched and understood, as well as showing how key psychological topics can be reworked. All of the contributors work with direct records of interaction from various institutional and everyday settings. These are highly varied, and include: family conversations with young children; mundane telephone calls; therapeutic and medical sessions; psychological experiments; market research focus groups; sex offender therapy; political speeches and emails; relationship counselling; psychiatric assessment for gender reassignment; school group evaluation and school counselling sessions; therapy for autistic children; and a child protection helpline. Taken together, the chapters illustrate an approach to social science issues that cuts across the traditional disciplinary divisions to provide a rich participant-based understanding of action.

In this opening chapter we will set this work in context, outlining developments in ethnomethodology, conversation analysis and discursive psychology, and distinguishing those developments from traditions of discourse work that make different assumptions.

Talking organisations

Conversation analysis

Conversation analysis (occasionally CA) originated in the 1960s in the lectures of Harvey Sacks (now published as Sacks, 1992). It was refined and rethought with his colleagues Emanuel Schegloff and Gail Jefferson. Conversation analysis offers an approach to analysis that combines a focus on the systemic nature of conversation and the way it is heard and understood by its participants. CA highlights three key elements of conversation:

  1. Talk is a medium of action. Central to conversation analysis is the notion that any utterance can be examined for the action that it performs. This focus on action (and on the way actions are parts of broader practices) is in contrast to traditional linguistic approaches where talk is treated in terms of sentences or similar structures.
  2. Actions done in talk are both context-dependent and context-renewing. That is, any action is oriented to the immediately prior turns of talk, and the action done in the current turn of talk provides a context for what comes next. Hence CA has a particular focus on sequences of action, which are often organised around paired actions such as invitations and acceptances, questions and answers, or assessments and second assessments.
  3. In producing a next action, a speaker displays their understanding of the prior action. For example, if an invitation is issued by a speaker, the recipient not only accepts (or declines) the invitation, but in doing one of these acts (or relevant alternatives) shows that they have understood what has been issued as an invitation. This, in turn, provides further opportunities for the issuing speaker to acknowledge or initiate repair of a problem. This turn by turn display of understanding is crucial, both for participants, as it allows them to coordinate their actions, and for analysts, as it allows them to ground their claims in participants’ own understandings.

In the thirty years since the publication of the foundational turn-taking paper (Sacks, Schegloff and Jefferson, 1974) conversation analysis has resulted in a cumulative set of studies that map out some of the systemic features of the organisation of interaction – the very ‘structures of social action’ (Atkinson and Heritage, 1984) that provide the building blocks for social life. The power of these structures is that they are not brittle templates that must be followed; rather, they are normative. If an invitation is not followed by an acceptance or refusal (or some other normatively relevant action) this may occasion possible inferences (the recipient has not heard, is rude, is embarrassed or some other contextually relevant possibility).

Sacks worked on a range of materials from settings such as suicide-prevention lines and therapy groups. His focus was on the basic conversational phenomena – he was less concerned with how these conversations were refined in, or did the work of, institutions. The first major work that explicitly and systematically applied conversation analysis to institutional materials was Atkinson and Drew’s (1979) study of courtroom interaction. This considered the way the practices that make up the work of the courts are achieved, for example, how the constraints of cross-examination questions necessitate more than one turn to generate an accusation, and how witnesses produce different kinds of defences in the sequential positions before and after the blaming is completed.

Although a series of studies on various topic areas was produced in the intervening time it was not for another decade until conversation analytic work on institutional interaction came to major prominence in four collections (Duranti and Goodwin, 1992; Watson and Seiler, 1992; and particularly Boden and Zimmerman, 1991, and Drew and Heritage, 1992a). Taken together, and despite some differences, this work revolutionises the way what it is to be an institution can be understood. In particular, it offers an alternative to the more common ‘container’ view of institutions, which treats them as broad societal boxes within which interpersonal actions take place in a way that is somewhat determined by features of the institutional box. At the same time it offers a radically different treatment of the role of broad social categories such as race, class and gender that sociologists have often taken to be central to the asymmetries of social institutions. Such categories are often a major focus of Critical Discourse Analysis (discussed below).

Constructing context: relevance and procedural consequentiality

The four collections on institutional interaction offered different arguments against the ‘container’ view of institutions. To better comprehend the radical direction of these four collections, we can focus on the arguments presented by Schegloff (1992). Schegloff presents two key challenges that illustrate the subtlety and complexity of addressing institutions in this way, one focused on the issue of relevance and one on the issue of procedural consequentiality. The argument about relevance starts with the observation that there is a wide range of alternative categorizations available for persons and settings. Put at its simplest, even though the analyst may have some judgements about what categorizations are appropriate or correct the key interactional issue is what categorizations are treated as relevant by the participants. Thus, whether a person is female, or Scottish, or a teacher is not a sufficient warrant for the analysts to invoke that person’s membership of these categories (or any of the many other categories that the person could potentially occupy) to explain their utterances. The key issue is not abstract descriptive adequacy, but practical relevance to the interactional business at hand.

Schegloff suggests that there are two broad approaches to the problem of relevance. One can be described as positivist and requires that the success of a particular categorisation be assessed by statistical or historical methods, perhaps supplemented by interpretation on the basis of the appropriate theory. This approach works independently of participants’ use of and/or orientation to the terms. The other approach is central to the conceptualisation of institutions and categories in conversation analysis. In this approach the social science categorisations are grounded in the conduct of the participants and in particular in the categorisations and orientations that they are themselves using. As Schegloff (1992) puts it, it is not just that social scientists find people ‘to be characterizable as “president/assistant,” “chicano/black,” as “professor/student,” etc. but that for them, at that moment, those are terms relevant for producing and interpreting conduct in the interaction’ (p. 109).

The point, then, for Schegloff is not that these categories do not matter – they do. The problem is showing analytically that some features of the structure or some categories are what the participants themselves are orienting to. This will involve showing ‘how the parties are embodying for one another the relevancies of the interaction and are thereby producing the social structure’ (p. 110, italics in original).

The problem of procedural consequentiality comes into play when some category or social structure has been shown to be relevant. The point is that even if some category or structure is oriented to as relevant, that does not mean that it is procedurally consequential for the unfolding interaction. Thus if a classroom context, say, is relevant for the parties to an interaction, that does not mean that it has specific consequences for the content or trajectory or character of the interaction. What needs to be established is the mechanism by which the context (as understood) is consequential for the ongoing interaction.

For example, if it is thought that some style of question asking is central to classroom teaching, the analytic challenge is to show how this style is produced institutionally rather than being a questioning style that is common elsewhere and which has simply been drawn in the classroom setting. What this challenge encourages is careful comparative work. It is easy to assume that some interactional practices in an institutional setting are a product of that setting when a broader study might show that these practices are more generic.

Talk at work

Although Schegloff’s discussion of social organisation can seem more negative than positive it paves the way for a broader conversation analytic approach to institutions. Drew and Heritage (1992b) highlighted three features of institutions that would provide a framework for understanding the contribution of conversation analysis.

  1. Goal orientation. In institutional settings at least some of the participants are oriented to basic goals or tasks. These can be clear cut and relatively consensual (such as in calls to emergency services) or ill defined and fluid (such as health visits to new mothers). There are often differences between the orientations of lay and institutional participants (the patient and doctor, for example).
  2. Interactional constraints. Different institutional settings generate formal and informal constraints on conduct. Note that these can be inhibiting or promoting. In doctor–patient settings some talk is discouraged and some is encouraged.
  3. Inferential frameworks. In institutional settings the ‘inferential’ properties of actions may be different. For example, withholding an expression of sympathy might be treated as disaffiliative in a mundane setting such as a phone call between friends but not treated as such in a television news interview (e.g. Clayman and Heritage, 2002); in contrast, inconsequent-seeming remarks in a chat between friends might be treated as critical or threatening in some institutional contexts (e.g. Heritage and Sefi, 1992).

Overall, they noted that interaction in institutional settings often involves both a restriction of what happens elsewhere and also a refinement. Certain kinds of activities and certain sorts of responses drop out, but basic institutional activities such as courtroom cross-examination, medical consultation or news interviewing involve a refinement of more basic mundane practices (cf. Drew, 1992; Peräkylä, 1995; Clayman and Heritage, 2002).

In the time since the publication of the Drew and Heritage (1992a) volume there has been a large amount of work on interaction in institutional settings. This has increasingly refined the alternative to the container view of institutions, highlighting a range of different levels at which the operation of institutionality can be revealed. Many of the themes in this work will be picked up in the chapters collected here.

Talking cognition

Discursive psychology

At around the same time that conversation analysts were offering this reworking of the nature of social institutions discursive psychology (sometimes DP) offered what can now be seen as a parallel reworking of the nature of psychology and, in particular, the status of cognition. Just as CA moved researchers away from the idea that institutions are separate entities that have some kind of determinist effect on interaction, so DP moved researchers away from the idea that cognition is a separate mental space that has a determinate effect on action (Potter, 1998). It criticised the assumptions of the kind of cognitivism which assumes that the explanation of human conduct is dependent on the understanding of prior and underlying cognitive processes and entities. In these traditions of work action is treated in a more constitutive than dependent relationship to either the institution or the cognition. Indeed, both perspectives provide a critical stance in relation to the reified and solidified versions of institutions and cognitions.

Edwards and Potter’s (1992) volume Discursive psychology set out the foundations of the discursive psychological programme of work by way of a series of studies that reworked classic studies in cognitive psychology. For example, they examined the way Ulrich Neisser had used the testimony of John Dean to the Senate Committee investigating the Watergate break (ultimately crucial in the impeachment of President Nixon) as a basis for developing a theory of memory. They argued that by treating Dean’s testimony as a product of cognitive processes (different kinds of memory) Neisser was failing to appreciate the practical role of different versions of what had happened as Dean dealt with cross-examination. They show how he imposes a cognitivist construction on Dean’s testimony by treating it as determined by his memory and failing to attend to conversational and institutional pragmatics. In discursive psychology cognition is not the thing that explains interaction; rather, in a discursive psychological analysis we can see how versions of mind (memories, traits and attitudes) come to be produced for the purposes of action.

Core principles of discursive psychology

Discursive psychology works with three fundamental principles in its approach to discourse (Potter and Edwards, 2001).

  1. Action orientation. As in CA, discourse is treated as primary means through which actions are done and interaction is coordinated. Actions are seen as typically embedded in broader practices. DP does not assume that there will be a one-to-one relationship between discrete acts and discrete verbs. Rather, DP has had a particular focus on the way actions are done indirectly through different kinds of descriptions.
  2. Situation. DP treats discourse as situated in three complementary senses. First, it is organised sequentially in the way emphasised by conversation analysis, such that the primary environment for any utterance is the immediately prior utterance, and the new utterance sets up (although does not determine) what comes next. Second, discourse is situated institutionally, such that institutional identities (therapist and patient, perhaps) and tasks (managing problems, offering advice) will be relevant to what takes place. Third, discourse is situated rhetorically, such that any description can be inspected for how it counters relevant alternative descriptions (often, but by no means always, from the immediately prior talk).
  3. Construction. DP treats discourse as both constructed and constructive. Discourse is constructed from a variety of different resources (words, categories, rhetorical commonplaces, interpretative repertoires). Discourse is constructive of different versions of the world, including versions of actions, events, histories, social structures and organisations, psychological characteristics and phenomenological experiences. DP studies both the actions done with these construc-tions (how a person uses a version of the traffic on the motorway to account for missing a meeting) and the way these constructions are built to be stable, objective and independent of the speaker.

There is a considerable overlap of these DP principles and the basic assumptions of conversation analysis. This is not surprising as discursive psychology was itself heavily influenced by work in conversation analysis. Moreover, some of the key alternative sources for the development of DP (Gilbert and Mulkay, 1984; Potter and Wetherell, 1987) were themselves somewhat influenced by CA. And all of this work was somewhat influenced by ethnomethodology which played a powerful role in the inception of conversation analysis.

One of the achievements of discursive psychology has been to move away from the individualist and cognitivist assumptions of recent psychology. For discursive psychologists, what people say is not taken to represent the contents of their mind (what they are really thinking) or reality (what really happened); rather things such as mind and reality are seen as first and foremost resources for participants in dialogue – which also makes them a useful resource for the analyst. DP has therefore focused on broadly ‘psychological’ topics such as cognition and emotions (Edwards, 1997, 1999; Locke and Edwards, 2003; Potter and Hepburn, 2003), attitudes and evaluations (Potter, 1996, 1998a; Puchta and Potter, 2002; Wiggins and Potter, 2003), racism and prejudice (Buttny, 1999; Edwards, 2005; LeCouteur, Rapley and Augoustinos, 2001), and memories and motives (e.g. Edwards and Potter, 1992). In doing so, it has offered an alternative to traditional psychological approaches to these topics, and also to how we theorise ‘psychology’. Psychological concepts are treated in DP not as something we have or we are, but as resources for action. Psychology becomes more interactionally focused, dynamic and culturally specific as a result.

Themes in discursive psychology

Work in discursive psychology has developed around a number of different themes. These are cross-cutting, but it is useful to highlight some of their differences. Edwards (2004) picks out three themes.

  1. Respecification and critique. Respecification involves the reworking of topics from cognitive psychology and social cognition from a discourse perspective. For example, the core social cognitive notion of script has been reworked in DP by considering the way descriptions of actions and events can produce them as standard and orderly (Edwards, 1994). Note that respecification is not intended in DP to just provide a different version of the same objects – rather it changes the whole perspective from a cognitive one to a constructed and action-oriented one. In many cases the coherence of the cognitive object will simply dissolve.
  2. The psychological thesaurus. One of the key aims of DP is to explore the working of common sense, including the use of categories that would conventionally be treated as psychological. This includes the range of words for emotional and mental states such as remember, think, upset, angry and so on as well as the huge range of available idiomatic and metaphorical constructions – bear in mind, boiling, ragged and so on. Several of the studies collected in this volume address terms from the psychological thesaurus.
  3. Management of psychological business. A major topic in DP since its inception has been the often implicit management of psychological themes. How are matters such as agency, doubt, prejudice and emotional investment displayed, built up or undermined through descriptions of actions, events or circumstances?

In addition to these three basic DP themes, some of the contributors to the current collection will pick up a newly emerging topic which is the discursive psychology of institutions (Edwards and Potter, 2001; Potter, 2005). Social psychology has traditionally had little interest in the specifics of social institutions, being focused instead on the operation of generic trans-historical social processes (Gergen, 1982). In contrast, DP has started to ask the question of how particular institutions and organisations – therapy sessions, classroom teaching, police interrogation – are done through the use of specific ‘psychological’ business. DP studies with this focus may ask how particular psychological notions and orientations are drawn on to do the work of the institution.

This emerging theme in DP builds on the ground-breaking work done by conversation analysts and adds a particular focus on the organisation and refinement of practices or issues that have more traditionally been understood as psychological. Issues of knowledge, stance, understanding, blame, guilt and responsibility are threaded through everyday situations and are at the core of many institutions. DP focuses on the way these issues are practically managed in interaction.

Differences between conversation analysis and discursive psychology

As we have already indicated discursive psychology has drawn heavily on the theoretical ideas and analytic approach of conversation analysis. Sacks’ (1992) early work not only laid the foundations for conversation analysis it also developed a sophisticated interactional approach to the relationship between utterances and psychological states which is, to some extent, a forerunner of modern discursive psychology. However, it is worth briefly considering three areas of potential tension between DP and CA.

First, DP developed a systematic approach to the relation between the way descriptions are assembled and the actions they are involved in. This drew as much on developments in the sociology of science and broader constructionist ideas as specifically CA work (see Potter, 1996, for an overview). For example, Edwards (1995, 1997) studied the way constructions of anger in relationship counselling can play a role in assigning problems to one of the partners, nominating them as the person who needs the therapy. Constructions of this kind are mutually inferential – people construct versions of their memories, feelings and cognitive states as part of establishing the nature of events or settings; and they construct versions of events or settings as part of establishing the nature of feelings or cognitive states. In addition, as we have noted DP draws on the rhetorical tradition of Billig (1996). This shows how descriptions are put together to counter actual or potential alternatives. DP is distinctive from other constructionist traditions (and closer to CA) in its focus on the business of constructing versions in talk and texts, and its emphasis on the way constructions are parts of situated practices. Conversation analysts have been less focused on constructionist themes of this kind (although they are not necessarily inconsistent with CA work).

A second area of potential contrast involves the way cognition is conceptualised. DP is a systematically non-cognitivist approach. It puts aside questions of the existence of cognitive entities and processes (technical or everyday) in favour of a focus on how cognitive entities are constructed in and for interactional practices. It can study ‘upset’ in a therapy session, for example, without trying to answer the question of whether the word ‘upset’ has an inner referent, and without trying to assess whether a ‘display’ of upset is ‘authentic’ or ‘invented’. Nor is such a study required to decide on the reality of distinctions between, say, surface and depth psychology prior to analysis. These things can be treated as topics for study in their own right. For the most part CA has also been an enterprise that avoids cognitivist assumptions. However, at times CA researchers have had a more ambivalent approach to cognition. Sometimes this has involved an attempt to connect interactional phenomena to putative mental objects or at least to suggest the coherence of such a programme (Drew, 2005; Kitzinger, 2006; Schegloff, 2006). For an overview of these issues see the different papers in te Molder and Potter (2005) and the debate between Coulter (1999) and Potter and Edwards (2003b).

© Cambridge University Press
printer iconPrinter friendly version AddThis