Eight challenges for interview researchers

38 EIGHT CHALLENGES FOR INTERVIEW RESEARCHERS ◆ T Jonathan Potter and Alexa Hepburn here is little need to provide further evidence here of the ubiquity of the openended research interview across the range of contemporary social sciences. Chapters in this volume and its predecessor (Gubrium & Holstein, 2001) make this point very effectively, as does a survey of the content of contemporary qualitative methods handbooks such as Denzin and Lincoln (2005) and Willig and Stainton-Rogers (2008). In some cases, the term interview is not even mentioned, as this method of eliciting material from participants has become well wired into the commonplaces of social science. A diligent reader need only read through the content of the past full year of mainstream journals in sociology, social psychology, geography, and anthropology to see that where qualitative research is conducted, it is overwhelmingly done using some forms of interviews. Cutting things up another way, the open-ended interview is the preeminent data generation technique in methodological traditions as disparate as ethnography, phenomenology (in its different forms), psychoanalysis, narrative psychology, grounded theory, and (much) discourse analysis. Our aim in this chapter is to make the case that interviewing has been too easy, too obvious, too little studied, and too open to providing a convenient launch pad for poor research. We will argue that interview research will be made better if it faces up to a series of eight challenges that arise in the design, conduct, analysis, and reporting of qualitative interviews. Some research studies already face up to some of these challenges; few studies face up to all of them. We will make our case strongly and bluntly with the aim of provoking debate where not enough has taken place. These challenges are overlapping, but we have separated them in the way we have for clarity. It is important to emphasize that our aim is not to criticize interviews but to make them better. There are two contexts for this chapter for us. First, the past 20 years have seen an extraordinary ◆ 555 556 ◆ PART VII. CRITICAL REFLECTIONS development of our understanding of what might be called the central motor of interviews, the questionand-answer pair. Profound work in the tradition of conversation analysis (Schegloff, 2007) has been done on the organization of questions and answers in institutional settings such as television news interviews (e.g., Clayman & Heritage, 2002), court rooms (e.g., Atkinson & Drew, 1979), police interrogations (e.g., Stokoe & Edwards, 2008), helplines (e.g., Hepburn & Potter, 2010), and medical examinations (e.g., Boyd & Heritage, 2006), as well as in mundane settings such as everyday phone calls and family mealtimes (Heritage & Raymond, in press; Stivers & Hayashi, 2010). Such work has started to unpack some of the basic design features of questions, how they embody preferences, manage neutralism, build presuppositions, and work to constrain the actions of the recipient in different ways (Clayman & Heritage, 2002; Raymond, 2003). Researchers are starting to turn the analytic searchlight from this tradition onto the operation of social science methods (Antaki & Rapley, 1996; HoutkoopSteenstra, 2000; Maynard, Houtkoop-Steenstra, Schaeffer, & van der Zouwen, 2002; Puchta & Potter, 1999). In this chapter, we will draw on this tradition and shine a bit of its light on the qualitative research interview. The second context is a more biographical one. Both of us started our research careers doing research with open-ended interviews. Both of us have published widely using interview research (e.g., Hepburn, 2000; Wetherell & Potter, 1992) and on the nature and role of interviews (Potter & Hepburn, 2005b; Potter & Mulkay, 1985; Potter & Wetherell, 1995); yet both of us became increasingly dissatisfied with what was possible, particularly as we have become more sophisticated in the analysis of interaction. In a way, this chapter is an attempt to make sense of that dissatisfaction and provide a more positive set of suggestions for interview research. The current chapter develops two earlier discussions of the status of research on open-ended interviews and highlights the general implications beyond the field of psychology (Potter & Hepburn, 2005a, 2005b, 2007). We have arranged the eight challenges into two groups. Four of them can be addressed by attending to the reporting of the interview in the research study. These challenges can be met by 1. improving the transparency of the interview setup, 2. more fully displaying the active role of the interviewer, 3. using representational forms that show the interactional production of interviews, and 4. tying analytic observations to specific interview elements. The aim will be to provide a set of suggestions for how any interview study can be reported, so as to support more comprehensive evaluation by readers. Currently, interview studies often provide only the most limited possibilities for auditing by other researchers. These four suggestions will extend those possibilities. The second set of challenges arises in the analysis of the interview. They require interview researchers to pay more attention to 5. how interviews are flooded with social science categories, assumptions, and research agendas; 6. the varying footing of interviewer and interviewee; 7. the orientations to stake and interest on the part of the interviewer and interviewee; and 8. the way cognitive, individualist assumptions about human actors are presupposed. These analytic issues are potentially highly consequential for how the talk of the interviewee is understood in relation to the claims of any research study. These are by no means the only analytic challenges, but all are potentially relevant for any researcher who wishes to make adequately grounded claims on the basis of an interview study. After describing these challenges, we will make two broader sets of suggestions. First, work on questions and interaction should feed back into the design, conduct, and analysis of interviews as well as into the training of interviewers and the strategic exploitation of interviewer conduct. In particular, open-ended interviews should be viewed as a range of rather different conversational occasions, and the research should be designed to build a particular kind of occasion as suitable for the specifics of the Chapter 38. Eight Challenges for Interview Researchers research. Second, we believe that there needs to be more consideration of whether research aims could be furthered more effectively by using other forms of data generation, including working with records of natural interaction. Reporting 1. MAKE THE INTERVIEW SETUP EXPLICIT One thing that is striking about much contemporary interview research is how little is said about how participants were recruited. The concern here is not with sampling; it is with the sort of potentially highly consequential interaction that goes on when interviewees are introduced to the study. In particular, there are two elements of the interview setup that are likely to be important. The first is the issue of the category that participants have been explicitly recruited under. For example, were they recruited as “adolescent recreational drug users,” “young adult unemployed,” or “active waste recyclers”? Researchers typically recruit participants grouped into such categories, which are themselves key parts of the research arguments that are being pursued. Such recruitment is itself an active part of the research process, not unlike the social identity salience manipulations in social identity theory research studies (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), and it may figure in the introduction to the research that the participant is given, the ethics procedures, the administrative arrangements, and so on. Note that this issue is separate from the issue of sampling. It is not a matter of the category embedded in the abstract research design; it is a matter of the processes through which particular individuals are actually recruited and told about their recruitment and role in the research. It is during these processes that particular category memberships can in various ways be made central to the research. Thus, a study might be concerned with the beliefs of “depressed students” about the causes of their problems. When such students are recruited, the category “depressed student,” and potentially a range of related categories, is likely to be made central, formulated, and reworked. Indeed, recruitment itself often has some of the characteristics of an interview. ◆ 557 During recruitment, a particular understanding of how each participant is relevant to the research may be delivered to them; and such an understanding is, at least potentially, highly consequential for their conduct during the interview. The second issue that becomes live in the interview setup is the understanding of the task that participants are given. This involves issues such as what the interview will be about (healthy and unhealthy eating, say), what the research will be used for (influencing government health policy), and what tasks the interviewees will be performing in the interview (e.g., explaining how their eating preferences have changed as they grew up). In many ways, it is not surprising that these matters are rarely included in research studies. Full records are often not kept, and when asked for full transcripts and sound files, researchers typically cite tensions between generating ethical consent and collecting full records. Adequate consent procedures may depend on prior description of the interview topics and the rationale for recruitment. Nevertheless, given how potentially consequential for the research outcomes such category membership ascriptions and topic and task formulations are, it is important to attempt to capture these features of recruitment. One possibility is including textual materials relating to recruitment and attempting a descriptive overview of the early interaction between participant and researcher. Crucially, when the actual interview takes place, the recording should be started as early as possible, so that the researcher’s formulation of the nature of the interview and its goals is captured and made available for scrutiny. This issue has received almost no attention in previous discussions of methodology. Our aim is to signal both its neglect and its importance and to start to indicate some ways in which the situation can be improved. 2. DISPLAY THE ACTIVE ROLE OF THE INTERVIEWER Contemporary interview studies regularly emphasize that what goes on in interviews is interactional. Yet this is rarely followed through in the research practice. We will try and illustrate this point without picking on individual studies; it would be invidious to pick any one paper out of the thousands published every year. 558 ◆ PART VII. CRITICAL REFLECTIONS But we have no doubt that readers will find what we are illustrating here instantly recognizable. Anyone still not sure should sit down with the past few years of any mainstream social science journal from sociology, human geography, nursing, or qualitative psychology and count the number of extracts that include both interviewers and interviewees in relation to the number that simply include the latter. Expect a ratio of 1 to 10 or more! We will use an example from our own materials so that readers can refer to the audio record if they wish to and to spare the blushes of others. The following quotation comes from a corpus of interviews conducted by Alexa Hepburn as part of a project looking at the different ways bullying appears in school settings (published papers using this corpus include Hepburn, 1997a, 1997b; Hepburn & Brown, 2001). The extract has been rendered in the style typical of contemporary qualitative research. Extract 38.1 I think all teachers are stressed. Because they’re stressed they may react inappropriately in certain situations, because they are near the edge themselves. If you’re tired and stressed you’re not always in the best situation to make good judgements. The children I think at least are slightly more aware of this than they used to be in the past. Let us make some observations about the form as presented here. First, note what is absent. The interviewer’s question has not been reproduced. The extract comes only from the talk of the interviewee, the teacher. And we are not told if this is an entire turn of talk that is responsive to a question or if it is a fragment from a longer turn. This has the effect of framing the talk as an abstract pronouncement on the nature of teachers and the effects of stress; it is not framed as a specific answer to a specific question put by a specific interviewer. Second, note that the extract uses a conventional orthographic representation of talk. It is rendered as a form of play script; it has been turned into mostly grammatical sentences and uses conventional punctuation marks. Some of what is missing can be clarified by simply extending the extract using the same orthographic form of representation. Material absent from Extract 38.1 is rendered in bold to show the kind of contextual and interviewer talk that is commonly omitted. Extract 38.2 Interviewer: So do you feel then that the constraints on teachers’ time and the resources that are available to you actually err constrain your ability to do your job well to deal effectively with kids Teacher: yes I think all teachers are stressed err because they’re stressed they may react um inappropriately in certain situations because they are near the edge themselves (Int: yes yes) erm if you’re tired and stressed you’re not always in the best situation to make good judgements (Int: oh yeah yeah) the children I think at least are slightly more aware of this than they used to be in the past (Int: mm mm) but yes I would say it can affect it With the introduction of the interviewer’s question and various contributions during the teacher’s turn of talk, we get more of a sense of what is going on here as an interaction. Crucially, we can start to see how the form of the answer may be occasioned by the form of the question. Nevertheless, the talk is still rendered as a play script. This makes it hard to identify the precise actions going on here—it is left to the reader to animate the talk as it must have sounded. One crucial thing we are missing here is information about the delivery of the talk—the prosody, the delay and overlap, the emphasis and volume, the tempo and various features of voice quality—everything that turns it back into normal human interaction. It might be observed at this point that interview data analysis and representation are always partial and incomplete (Andrews, 2008). The key question is what kinds of information may be consequential for understanding what is going on in the interview. Information about what color of socks each party is wearing is unlikely to be relevant to how the interview unfolds. However, information about whether the interviewer receipted an interviewee turn with “oh,” whether interviewee intonation is closing or questioning at some point, or where the parties are speaking in overlap is much more likely to be consequential. If we are going to take seriously the idea that the interview is an interactional occasion, surely this kind of information should be available in some form. It is ironic that as we claim to be social scientists, rather than students of literature, we have been wiping out the embodied and voiced nature of talk. Chapter 38. Eight Challenges for Interview Researchers 3. REPRESENT TALK IN A WAY THAT CAPTURES ACTION ◆ by a transcription service (usually untrained speed typists); and then the recordings are boxed away, and the researcher works with these impoverished transcripts. It is important to consider what is missed by doing this. The transcription conventions for representing talk as delivered were developed by the late Gail Jefferson (2004; for a broader discussion, see Hepburn & Bolden, in press; a brief summary is given in the appendix to this chapter, Table 38.1). Extract 38.3 shows the same research interview exchange now represented using Jeffersonian transcription. For nearly half a century, interaction analysts, and particularly conversation analysts, have found that if they are to understand the actions that are being done in talk, they need to pay close attention to actual talk. This means listening closely to recordings, watching the video, if one exists, and combining this with a representation of talk that captures what is interactionally relevant. It is common in interview studies to get the recordings done Extract 38.3a 01 Int: So d’you feel then that the constrai:nts on teachers’ 02 ti::me and the resources that are available to you 03 actually .hh (0.2) c- er constrain your ability to 04 do your job ↑well to deal effectively with- (0.2) 05 °with kids: an° (0.2) [((inaudible))] 06 Tch: [U::m: ] (0.9) 07 ((swallows)) Ye:s, (0.7) I think all teachers are 08 stressed 09 (0.2) 10 Int: Mm:. 11 Tch: Er because they’re stressed (.) they may react (0.5) 12 13 u::m inappropriately, Int: 14 Mhm, (0.2) 15 Tch: 16 Int: 17 in certain situatio[ns,] [ M ]hm. (0.4) 18 Tch: 19 Int: 20 Because they (.) are near (.) the edge themse:[lves, ] [Yeah.] (0.4) 21 Int: °Yeah. 22 Tch: Er::m (0.9) if you’re ti:red, (.) an stressed, (.) erm 23 you’re not always in the best: situation to make 24 good judge[ments. ] 25 Int: 26 Tch: [↑Oh ye]ah. °yeah. [°Mm.] [Er: ]m (0.4) the CHILdren 27 I think at least are slightly more aware of this 28 [than they used to] be in [the pa: ]st. 29 Int: [ Mm:::. ] 559 [M m: :.] (Continued) 560 ◆ PART VII. CRITICAL REFLECTIONS Extract 38.3a (Continued) 30 Int: 31 (0.2) 32 Tch: 33 Int: 34 35 Er::m:= =Mm. (0.9) Int: 36 37 Mm:. °Mm. (1.5) Tch: BUT (huh)YEs I would sayhh (0.2) er it can affect it. (The audio record of this extract is available online—search for Loughborough Discourse and Rhetoric Group) This is a brief extract. Yet it shows up a wide range of hearable, and therefore potentially interactionally consequential, features of the talk compared with the play script version. Note the overlaps, closing intonation, latching of turns to one another, rising and falling intonation, raised volume, stretched vowel sounds, and different kinds of breaths and laugh particles. In contemporary social science articles, there is a pervasive failure to capture features of delivery such as these. The fuller transcript allows the identification of a number of potentially consequential interviewer actions. For example, note the “acknowledgement tokens” (Clayman & Heritage, 2002; Jefferson, 1985) on Lines 10, 13, and 16. Such tokens have a range of jobs, including displaying attentiveness and passing the opportunity to take a turn. Note also the news receipt and agreeing second assessment (Heritage, 2002; Pomerantz, 1984) on Line 25. As we will see, such objects are particularly interesting in research interviews. Our general argument is that such elements are universally treated as relevant by the parties to the interaction (Hepburn & Bolden, in press) and so will be consequential for the conduct of the interview and how it is understood. They, therefore, ought to be represented in a way that makes them available to readers. 4. TIE ANALYTIC OBSERVATIONS TO SPECIFIC FEATURES OF INTERVIEWS In a recent critical piece about the qualitative analysis of interviews, and particularly the uses of discourse analysis, Antaki, Billig, Edwards, and Potter (2007) argue that researchers pervasively underanalyze their materials. They highlight six different forms of underanalysis. These include underanalysis through summary, underanalysis by taking sides, underanalysis through overquotation or through isolated quotation, the identification of discourses from mental constructs (and vice versa), overgeneralizing claims, and ad hoc feature spotting. Our aim here is not to repeat this argument, which is fully documented with examples of interview analysis in the original piece. Instead, our focus will be on the issue of how links between specific analytic claims and specific sections or elements of interviews are made plain to the reader. One of the consequences of conventional orthographic representation is that it often makes it unclear what specific elements of the interviewee’s talk are being referred to. In part, this is because this form of transcript collapses together groups of different conversational elements. More technically speaking, the interviewee’s answer (and the interviewer’s question) can be built from a number of different practices that are not easy to separate because of the representational form used. For example, Extract 38.1 allows for a highly restricted consideration of the way different aspects of the interviewee’s answer are built and what elements of the interviewer’s talk they are responsive to. Even the simple procedure of putting the interviewer’s and the interviewee’s talk on separate lines allows for a much clearer understanding of what is going on. The additional, and equally simple, procedure of adding line numbers allows for much more precision in referring to segments of talk. It is common in contemporary interview studies to find large blocks of text combined with analytic observations that are hard to clearly link to specific elements in the talk of the interviewee. Chapter 38. Eight Challenges for Interview Researchers Challenges in Reporting Interviews In proposing these ways of improving the quality of representational practice with interview research, we are mindful that there are difficult tensions between practices that support the thorough academic auditing of research claims and practices that simplify the process of interview research. We are asking the question rather than offering a template. Transcription, in particular, both is time-consuming to conduct and requires training and effort (Hepburn & Bolden, in press). Where transcription has been addressed by interview researchers, it has often been with an argument for “Jefferson Lite”—transcription that captures words and some of the grosser elements of stress and intonation but leaves pauses untimed and does not attempt to capture more subtle elements such as closing and continuing intonation, latching, and so on. Poland (2001) offers a relatively sophisticated version of this position. There is an argument that attention to what seem to be merely micro aspects of talk will detract from broader themes or ideological organization. However, an alternative argument is that broader themes and organizations are in practice understood through working with the concrete specifics across a number of examples of talk; therefore, analysis will benefit from an engagement with a form of transcript that goes beyond the reconstructed, simplified, and distorted version of interaction that comes from many transcription services. A further argument for attending to the specifics is that this is what participants do, pervasively and thoroughly, in the course of their own interaction. The full Jeffersonian transcript drags into the open the jointly constructed, socially engaged nature of what is going on in interviews, including the close dependence of what the interviewee says on the interviewer’s question design and delivery in all its specifics. Even if the interview researcher is happy to work with a reduced form of representation, there is still the issue of how the research is to be adequately evaluated by others. Insofar as the evaluation of the research is a communal endeavor for journal referees and readers, there is a strong argument that the researchers should provide a form of transcription that will offer readers a much fuller understanding of what is going on and, moreover, an understanding less likely to have already embedded their own theoretical assumptions within it. Over the years, we have refereed many articles using interviews in different social science areas. There are often important cases where the interviewee’s talk is ◆ 561 constrained by the actions of the interviewer, and yes, this is obscured by the form of representation. Our feeling is that some moves in this direction are overdue and need to be encouraged by journal editors. Jeffersonian transcription is a slow process—done well it involves a recording to transcription time ratio of at least 1 to 20 hours. However, the aim would not necessarily be to do a full transcript of the complete interview corpus; rather, the full transcript would be reserved for the instances of some analytically identified theme. For example, in conversation-analytic studies working with large corpuses of institutional talk, it is common to work with collections of fully transcribed extracts that relate to particular questions. It may be that some of what we are suggesting can be provided by making audio or video records available via the web. These could be comprehensive, illustrative, or just provide the recording of the extracts that are reproduced in the papers. Ethics procedures would need to be adjusted accordingly. The provision of audio and video materials would help address a further problem with the representation of interview material, which is that transcript may be faulty. Sometimes, the transcribed form is such that it “sounds” impossible to a trained ear, or the formulations offered seem pragmatically odd. On those occasions when such doubts have led us to ask if we can check the recording ourselves, we have been told that ethics preclude this or the recordings are unavailable. While this may be true, it does not promote a social research practice that is publically accountable and transparent. We do not underestimate the extra effort involved in moving in this direction. However, we have made strong arguments for the value of that effort. It is up to others to show that it is unnecessary and these arguments are wrong. A defense of the current practice will need to do more than say that the researcher is not interested in interaction, as Morgan (2010) has argued; researchers defending the status quo will need to show that the close dependence of both the form and the content of the “answer” on the design and delivery of the “question” is not of general consequence. Analysis The first part of this chapter has focused on the issues that arise in the representation of interviews in research publications. In the second part, we will focus on some challenges that face researchers when they are analyzing interviews. Again, these are 562 ◆ PART VII. CRITICAL REFLECTIONS challenges that are brought into focus by a familiarity with interaction research. These highlight features of research interviews that are pervasive and yet hard to successfully manage analytically. Although we are describing them as challenges, we do not mean that they are typically seen in this way by interview analysts; indeed, for the most part they are overlooked. The point is that they are consequential for understanding how the interview should be interpreted. We will work through them in turn. 5. FLOODING Research interviews are flooded with social science agendas of different kinds. The notion of a “social science agenda” is meant to capture the loose set of concerns and orientations that are central to the researcher who is conducting social research. Researchers have spent years developing particular views of social organizations and structures, the nature and competence of human subjects, and so on. They often work with a “factors and variables” picture of how different elements of sociality affect one another. Such pictures are often taken for granted to the point of invisibility. Agendas of this kind are expressed in various ways, including the research recruitment briefings, the kinds of questions asked and how they position recipients, the categories they use, and the way questions are organized into narratives that build particular researcher concerns. It is the apparent invisibility of such agendas that makes their influence on the actions of research participants hard to identify. Sometimes, clearly technical or quasi-technical terms such as internalized homophobia (Giorgi & Giorgi, 2003) mark the way the participant is being recruited into a particular social science tradition. However, the relevant social science agenda is often developed much less explicitly. Let us look more closely at the initial question from Extract 38.3 above. Extract 38.3b 01 Int: So d’you feel then that the constrai:nts on teachers’ 02 ti::me and the resources that are available to you 03 actually .hh (0.2) c- er constrain your ability to 04 do your job ↑well to deal effectively with- (0.2) 05 °with kids: an° (0.2) [((inaudible))] This question has the feel of something put together on the spot. The delay, hitches, and an “er” on Lines 3 to 5 give a sense of the interviewer building the question locally rather than rehearsing something preformed. The trailing off to quiet on 5 and then the overlapped inaudible element contribute to this sense. Nevertheless, this is a highly recognizable and recurrent form of social research question. Indeed, Puchta and Potter (2004) suggest that the display of informality here is part of what constitutes it as a standard question from an openended interview (as opposed to a systematic survey question). These features help in the production of informality and manage potentially problematic epistemic asymmetries between interviewer and interviewee. When we drill down further into what is going on here, however, things are a bit more complex. This informality is combined with a question form that is highly constraining. A polar question, or yes/no interrogative, advances a candidate proposition and sets the terms from within which participants’ responses are to be interpreted (Heritage & Raymond, in press). Let us consider this candidate proposition a bit more closely. It does not draw on terms from one of the obvious social science dictionaries (“internalized homophobia”). Nevertheless, it is built with references to abstract objects (“teachers’ time”—Line 2) and processes (“constraints”—Lines 1 and 3). Moreover, the category is the generic “teachers” rather than a reference to specific teachers at a particular school. The point here is not that such questions are poorly designed or use illegitimate constructions but rather that such types of question can subtly coach the participant in a relevant social science agenda. This is not in itself a bad thing; the challenge is how to work with it in a way that avoids circularity. Although interview questions are an obvious place to look for the development of a social science agenda, it can also be seen in other kinds of interviewer contribution to the interaction. The following comes a little later in Extract 38.3: Chapter 38. Eight Challenges for Interview Researchers ◆ 563 Extract 38.3c 18 Tch: 19 Int: 20 Because they (.) are near (.) the edge themse:[lves, ] [Yeah.] (0.4) 21 Int: °Yeah. 22 Tch: Er::m (0.9) if you’re ti:red, (.) an stressed, (.) erm 23 you’re not always in the best: situation to make 24 good judge[ments. ] 25 [↑Oh ye]ah. °yeah. [°Mm.] Int: The point we want to highlight here is the difference between the interviewer’s contributions on Lines 19, 21, and 25. While the interviewee’s turn on Line 18 receives only an acknowledgment token (Jefferson, 1985)—a conversational object that specifically holds off doing a number of potential actions—her contribution on Lines 22–24 is responded to with an “oh-prefaced” agreement (Heritage, 2002). These interviewer responses are markedly different in what they display about the status of the propositions advanced. We will say more about that later in relation to footing. For the moment, just note the way such interviewer turns are a vehicle for the broader social science agenda. They potentially lead the interviewee down a desired path. Crucially, the challenge is to avoid the interview taking the form of a circular rediscovery of preexisting social science ideas and assumptions. How can the analysis cut down the possibility that it is chasing its own tail, offering back a refined or filtered form of the ideas and intuitions that went into the building of the original schedule and briefing of participants? another, they do so from a range of different positions or “footings.” Interaction has a “participation framework,” where a speaker may be the origin of ideas, for example, or may represent the ideas of another speaker (compare a national president and his or her speechwriters) and recipients may be directly addressed, say, or be in a position to overhear the conversation. Interview talk can be considered in this way, focusing in particular on the different bases on which participants are speaking. Interviewees are typically recruited as members of particular categories (lesbian parents, teachers, multiple sclerosis sufferers, etc.). But when they are speaking, they can be speaking as an individual with her or his own unique beliefs and preferences; as a category member, where the answers are intended to reflect the category as a whole; or, more likely, as some complicated mix of the two. There are subtle but important differences between the two. Moreover, if they are speaking as a category member, what precisely is the relevant category? We will illustrate the analytic significance of footing with examples from Extract 38.3. Consider first the question design and its use of categories: 6. FOOTING One of the key observations made by Erving Goffman (1981) was that when people speak to one Extract 38.3d 01 Int: So d’you feel then that the constrai:nts on teachers’ 02 ti::me and the resources that are available to you 03 actually .hh (0.2) c- er constrain your ability to 04 do your job ↑well to deal effectively with- (0.2) 05 °with kids: an° (0.2) [((inaudible))] 06 Tch: 07 [U::m: 08 stressed_ 09 (0.2) 10 ] (0.9) ((swallows)) Ye:s, (0.7) I think all teachers are Int: Mm:. 564 ◆ PART VII. CRITICAL REFLECTIONS The interviewee has been recruited as a category member (a teacher), and she is addressed in a way that invokes her category membership (“teachers’ time,” “your job”). Yet she is addressed in direct personal terms (“you”—Line 1) in a way that separates out her stance on the category that she is, here, relevantly, a member of. As we noted above, in virtue of its yes/no interrogative form, the question restricts the response options. Anything other than a yes/no response will be type-nonconforming (Raymond, 2003) and therefore liable to inspection by the questioner for what it is doing in departing from the terms of the question. Thus, the respondent could build an answer that separates her from the category—most teachers say X, but I find Y—but this would involve unpicking the terms of the question. Instead, she offers (after an extended um and delay, which may be occasioned by the trail-off in the question but more likely relates to the potentially challenging nature of the question) a standard response with the type-conforming “yes” and the expansion “I think all teachers are stressed.” The general point is that participants can answer questions in a “personal” or “institutional” capacity, and without careful analytic attention to this, researchers risk making misleading inferences from their material. Here, the challenging nature of the question makes answering in a more generalized way a more useful resource. The footing of interviewers is also complicated. For example, are they treated by the interviewee as the addressed recipient? Or is their participant status as a conduit or reporter? Take television news interviews. Here, both interviewer and interviewee treat the overhearing audience as the relevant recipient of the talk. One interactional feature that displays this is the absence of “oh” receipts. There is no interactional need for news anchors to mark a “change of state.” Indeed, they may be asking questions to which they already know the answer or in which they have no personal interest. The issue, then, is not the news interviewer’s change of knowledge state but how informed the audience is (Clayman & Heritage, 2002; Heritage & Greatbatch, 1991). Again, we can consider the issue of footing as it is played out practically in the conduct of the interviewer and the interviewee. Interviewer turns can display different footing positions. Compare the simple acknowledgment token “mm.” in Line 10 of Extract 38.3d with the turns on Lines 25 and 29 below: Extract 38.3e 22 Tch: Er::m (0.9) if you’re ti:red, (.) an stressed, (.) erm 23 you’re not always in the best: situation to make 24 good judge[ments. ] 25 Int: 26 Tch: [↑Oh ye]ah. °yeah. [°Mm.] [Er: ]m (0.4) the CHILdren 27 I think at least are slightly more aware of this 28 [than they used to] be in [the pa: ]st. 29 Int: [ Mm:::. 30 Int: Mm:. ] [M m: :.] In these turns. the interviewer is displaying full recipiency, with oh-prefaced agreement on Line 10 and stretched acknowledgment tokens on Line 29. By indexing existing relevant knowledge and views in this way, the interviewer builds herself or himself as a fully active participant (Heritage, 2002) rather than a neutral, disinterested recorder of what is said by the interviewee. This form of engagement can be part of a process where the interview productively becomes “an interventionist and confrontative arena” (Potter & Wetherell, 1987, p. 164). However, note that what we have here is not something consistent through the whole extract, let alone the whole interview. The different interviewer footing appears selectively in different positions in relation to different interviewee turns. Footing itself is a somewhat limited notion (see the debate in Leudar & Antaki, 1996; Potter, 1996), and there is a range of further possibilities for how footing can be displayed in interviewee talk (Ensink, 2003; Lee & Roth, 2004). Our point is that these are subtle, complex, and yet potentially consequential matters that up to now have been barely considered by analysts of interviews. Chapter 38. Eight Challenges for Interview Researchers 7. STAKE AND INTEREST A central feature of interaction is that people orient to issues of potential stake and interest (Edwards & Potter, 1992). They may respond to what other people say as a product of interests and manage the potential interestedness of their own talk. This can play out in complex ways in open-ended interviews. Consider the interviewee. She or he will typically have been recruited as a member of a social category; and she or he is likely to be treated as having a stake in that category. At the same time, interviewees are typically treated as being broadly neutral informants of their own practices. This combination of being more or less neutral and having a stake in a category is an analytic commonplace of more discoursefocused work on interviews (e.g., Edley, 2001), ◆ 565 although that does not make the analysis easy to do. And such issues are more or less ignored in most qualitative interview analyses. Now take the interviewer. Much qualitative research is conducted on doctoral programs, and here it is overwhelmingly the case that the interviewer and the researcher are the same person. Moreover, it is not just common but expected for those researchers to care deeply about the topic they are studying. So there is a profound issue of how this potentially live stake is managed. It is interesting to compare this with market research settings, where there is often an explicit focus on the potential interestedness of the interviewer or moderator. The following is the kind of construction that is recurrent at the start of market research focus groups: Extract 38.3f 01 Mod: As I say I <↑don’t make> the adver↓tising, 02 (0.5) I don’t sell cars.=I don’t work for 03 ei:ther company that doe:s:. .hhh s:o: er:: 04 ↑whilst (0.3) >this research has clearly 05 been commissioned< by: (.) er a >company 06 that does< both. An >you’ll see (as we go 07 through) who ‘tis.< 08 (.) 09 10 11 Mod: ↑I don’t have a vested interest. (0.2) Mod: °Right, >so I don’t really mind what you< say:. (Puchta & Potter, 2004) When the moderator introduces herself to the group like this, she emphasizes her independence from the firm that has commissioned the focus group and how she was not involved in producing the advertising for its products (something that participants might expect). Note the very explicit denial of having an interest (Line 9) and the assertion of the implication of this for the conduct of the participants (Line 11). Note that the converse implication here is that the moderator otherwise might have had an interest in what is being discussed and might mind it being criticized! In academic qualitative social science work with interviews, this kind of asserted separation between the interests of the researcher and the topic is much less common. The researcher is unlikely to be able to plausibly and honestly assert his or her disinterest. Researchers do on occasion introduce the issue of their own stake in the topic at hand—although this is probably more common offstage during the interview setup. An important area of future metastudy will be the way such avowals of concern (or relevant membership) play out in the trajectory of the interview. As with footing, issues of stake and interest vary through the trajectory of the interview. They appear in different places and in different ways. We have already noted an example. Consider the following extract again. Note again here the interviewer’s displays of investment in the topic in Lines 25 and 29. The teacher is being interviewed about the potentially delicate topic of punishment of children (and how some of the schoolchildren interviewed have suggested the phenomenon of teacher bullying). The teacher is building an account for “inappropriate” 566 ◆ PART VII. CRITICAL REFLECTIONS Extract 38.3e 22 Tch: Er::m (0.9) if you’re ti:red, (.) an stressed, (.) erm 23 you’re not always in the best: situation to make 24 good judge[ments. ] 25 Int: 26 Tch: [↑Oh ye]ah. °yeah. [°Mm.] [Er: ]m (0.4) the CHILdren 27 I think at least are slightly more aware of this 28 [than they used to] be in [the pa: ]st. 29 Int: [ Mm:::. 30 Int: Mm:. ] [M m: :.] actions, prompted by a question that indexes certain constraints on “dealing effectively with kids.” By employing an “oh-prefaced” agreement, the interviewer treats its contents as “already known,” thereby more strongly aligning with this account and encouraging the teacher to be less guarded. Agreements (and disagreements, of course) can display broader alignments and interest in topics (see Koole, 2003). For the moment, we will just note again that if we take this seriously, it makes the process of interview analysis considerably more complicated than is often presented. And it highlights the powerful role of interviewer activities. Our point is not that such activities should be driven out of interviews. Agreements show attentive listening and engagement (often glossed as “rapport” in methods texts). However, they do more than this, and their specific placement deserves analytic attention. Our general point is that the interactional organization of interviews is complex and consequential, yet it is rarely explicitly addressed. To take it into account during analysis is a major challenge (for a research example that highlights the subtlety of the challenge, see Edwards, 2003). However, to fail to take it into account risks interview conclusions that are based more on researchers’ prior expectations than on analysis of actual interview conduct. 8. COGNITIVISM AND INDIVIDUALISM One of the pervasive but almost unnoticed byproducts of interview research is the reproduction of a kind of cognitive individualism. The point is not that such a perspective is necessarily wrong; it is that there is no way of testing it when it is reproduced in this way. It is a vision of human life presupposed rather than discovered in most interview research. There is something particularly ironic about this given that many researchers who draw on qualitative interviews espouse a critical, nonindividualist perspective on social life. We will focus on two ways in which cognitive individualism is reproduced: (1) the privileging of conceptual meditation over action and (2) the treatment of cognitive language as referential. Privileging Conceptual Rumination Qualitative interviews recruit participants to report on events, actions, social processes, and various kinds of cognitive objects (attitudes, beliefs, etc.). This kind of explicit conceptual meditation is treated as providing a way into participants’ minds or social organizations. Participants orient to this in their answers, offering up their reflexive thoughts as skilled participant theorists. Note the syllogistic pattern in the following, and the way causal relationships are adduced: Extract 38.3g 06 Tch: [ U : : m : ] (0.9) 07 ((swallows)) Ye:s, (0.7) I think all teachers are 08 stressed_ 09 10 (0.2) Int: Mm:. Chapter 38. Eight Challenges for Interview Researchers 11 Tch: 12 13 Er because they’re stressed (.) they may react (0.5) Int: Mhm, (0.2) Tch: in certain situatio[ns,] The teacher here is being asked as a teacher not to be a teacher but to reflexively formulate aspects of the lives of teachers and to provide causal observations. Looked at in one way, this is a restatement of the basic rationale for doing interviews in the first place. People are asked about what they do and what they think, and they helpfully tell you about these things. However, looked at another way, what is going on here is that people are being treated as being in a special epistemic position with respect to their own conduct—not just actions and events but also causal and developmental relationships, intrapsychic processes, and so on. The interview depends on a range of ambitious cognitive judgments and feats of memory and analysis. Cognitive Language The contemplative picture of interview talk goes hand in hand with the kinds of referential understanding of specifically cognitive language that were criticized by Wittgenstein, whose critiques were refined in different ways by discursive psychology (Edwards, 1997) and conversation analysis (Sacks, 1992). Recent studies have started to consider the role of cognitive language in social research settings. Myers (2004) and Puchta and Potter (2002, 2004; Potter & Puchta, 2007) looked at the role of what focus group researchers call POBA terminology (perceptions, opinions, beliefs, and attitudes). For example, Puchta and Potter (2004) note that questions constructed in POBA terms make it harder for participants to produce “don’t know” and slow responses (people are treated as having immediate and privileged access to their own opinions and attitudes). And they also show how the interactional organization of social research can be used to generate POBAs as objects within individuals (Puchta & Potter, 2002). Consider the interviewer’s question that starts our illustrative extract: 01 Int: So d’you ↑feel then that the constrai:nts 02 567 u::m inappropriately, 14 15 ◆ on teachers’ ti:me and the resources >that are Note the interviewer’s use of the POBA term feel at the start of this topic’s initial question. Asking for a “feeling” softens the epistemic demands of the question while heightening the interactional demands. Your “feeling” may not be subject to the same scrutiny as a factual claim; however, you are expected to know and be able to report on your own feelings. Similar sorts of issues arise with the interviewee’s use of psychological or cognitive terminology. To take one example, the interviewee uses the term stressed on three occasions in this sequence (Lines 8, 11, and 22). Whatever referential role this term has, careful analysis will be needed to consider what it is being used to do in this sequence. For example, Hepburn and Brown (2001), in an analysis of these interviews, highlight some of the practical uses of stress talk in managing accountability and linking individual actions with broader institutional roles and relationships. Our point is that to fully understand what is going on in qualitative interviews researchers will need to be attentive to the practical role of cognitive and psychological language in the talk of both interviewer and interviewee. Conversely, we will need to be cautious when treating such talk as a way of referring to inner psychological objects of some kind. The Conduct and Analysis of Interviews Recent interaction work in the tradition of ethnomethodology, conversation analysis, and discursive psychology has revolutionized our understanding of questions as social actions. This has two kinds of implication for the way qualitative open-ended interviews have been used in social research. First, it throws up a series of challenges. Once we recognize the subtlety and complexity of interaction in social research interviews, we are faced with the importance of representing that interaction in ways that allow fuller inspection. Not recording the potentially 568 ◆ PART VII. CRITICAL REFLECTIONS consequential work of the recruitment and setup makes it harder to appreciate the basis on which the interviewee is talking. The representation of the actual interaction in the interview needs to allow the reader to appreciate the active role of the interviewer and, most crucial, the way the interview questions are built. The pervasive failure to reproduce the interviewer’s questions is simply not acceptable now that we know the impact of different question designs and the way questions can embody different social science agendas. The second set of challenges is to do with the way the interview is analyzed and how far that analysis is able to take seriously the way social science agendas are hardwired into the interview, the way footing and stake and interest are managed, and the tendency of the very interview setup to analyze everything in terms of individual cognitive objects. Second, it throws up a set of opportunities. This literature can feed into the training of interviewers and the way interview schedules are developed. Work on news interviewers shows the way a range of subtle and challenging question forms can be effectively deployed. Social researchers have the opportunity to learn from this sophistication and improve their own practices. There is the potential for more control over the unfolding conduct of the interview than has currently been claimed in how-to methods books, with the possibility of managing issues of argument and displays of stake and interest to generate particular kinds of challenges and to provide a particular context to answering. The irony is that qualitative interviews are massively overused but their potential has been massively restricted. This kind of research also allows us to move on to a subtler position on issues of bias, neutrality, and leading questions. The way questions “prefer” particular kinds of answers is now better understood (Clayman & Heritage, 2002). Once we have the resources to identify such preferences analytically, we can move beyond bias as a problem toward considering how questions with different kinds of pressures can be used to open up issues for the interviewee. Despite the sophistication that is now possible, more systematic work on the social institution of the research interview is long overdue. Very few studies carefully describe the institution of interviewing and the different practices that make up that institution. This contrasts with the massive reliance of modern social research on the qualitative interview. We started with the observation that interviews have often been used on the basis that they are the obvious way of doing any qualitative social research, with few alternatives. Part of this increased sophistication in the conduct and analysis of interviews puts us in a position where we can make a more informed judgment on the choice of research approach. And the recognition of this choice starts to put the onus on researchers who are using interviews to fully justify that choice. Appendix Table 38.1 Basic Transcription Conventions [ ] Square brackets mark the start and end of overlapping speech. ↓↑ Vertical arrows precede marked pitch movement, over and above normal rhythms of speech. Underlining Underlining signals vocal emphasis. CAPITALS Capitals mark speech that is obviously louder than surrounding speech. °practical° “Degree” signs enclose obviously quieter speech. (0.4) Numbers in round brackets measure pauses in seconds (in this case, four tenths of a second). she wa::nted Colons show elongation of the prior sound: the more colons, the more elongation. bu-u- Hyphens mark a cutoff from the preceding sound. solid.= “Equals” signs mark the immediate “latching” of successive talk, whether of one or more speakers, with no interval. =We had hhh Aspiration (out-breaths): the more the longer. .hhh Inspiration (in-breaths): the more the longer. Chapter 38. Eight Challenges for Interview Researchers ◆ Yeh, The comma marks continuation, the speaker has not finished; intonationally it is a fall–rise or weak rising intonation. y’know? Question marks signal stronger, “questioning” intonation, irrespective of grammar. Yeh. Periods (full stops) mark falling, stopping intonation (“final contour”), irrespective of grammar. 569 Note: See Hepburn and Bolden (in press) for further details. References Andrews, M. (2008). Never the last word: Revisiting data. In M. Andrews, C. Squire, & M. Tamboukou (Eds.), Doing narrative research (pp. 86–101). London, England: Sage. Antaki, C., Billig, M., Edwards, D., & Potter, J. (2007). Discourse analysis means doing analysis: A critique of six analytic shortcomings. In J. Potter (Ed.), Discourse and psychology: Theory and method (Vol. 1, pp. 331–347). London, England: Sage. Antaki, C., & Rapley, M. (1996). “Quality of life” talk: The liberal paradox of psychological testing. Discourse and Society, 7, 293–316. Atkinson, J. M., & Drew, P. (1979). Order in court: The organisation of verbal interaction in judicial settings. London, England: Macmillan. Boyd, E. A., & Heritage, J. (2006). Taking the patient’s medical history: Questioning during comprehensive history-taking. In J. Heritage & D. Maynard (Eds.), Communication in medical care: Interactions between primary care physicians and patients (pp. 151—84). Cambridge, England: Cambridge University Press. Clayman, S., & Heritage, J. C. (2002). The news interview: Journalists and public figures on the air. Cambridge, MA: Cambridge University Press. Denzin, N. K., & Lincoln, Y. S. (Eds.). (2005). Handbook of qualitative research (3rd ed.). London, England: Sage. Edley, N. (2001). Analysing masculinity: Interpretative repertoires, ideological dilemmas and subject positions. In M. Wetherell, S. Taylor, & S. J. Yates (Eds.), Discourse as data: A guide to analysis (pp. 198–228). London, England: Sage. Edwards, D. (1997). Discourse and cognition. London, England: Sage. Edwards, D. (2003). Analysing racial discourse: The discursive psychology of mind-world relationships. In H. van den Berg, M. Wetherell, & H. HoutkoopSteenstra (Eds.), Analysing race talk: Multidisciplinary approaches to the interview (pp. 31–48). Cambridge, England: Cambridge University Press. Edwards, D., & Potter, J. (1992). Discursive psychology. London, England: Sage. Ensink, T. (2003). The frame analysis of research interviews: Social categorization and footing in interview discourse. In H. van den Berg, M. Wetherell, & H. Houtkoop-Steenstra (Eds.), Analyzing race talk: Multidisciplinary approaches to the inter view (pp. 156–177). Cambridge, England: Cambridge University Press. Giorgi, A. P., & Giorgi, B. M. (2003). The descriptive phenomenological method. In P. M. Camic, J. E. Rhodes, & L. Yardley (Eds.), Qualitative research in psychology: Expanding perspectives in methodology and design (pp. 243–274). Washington, DC: American Psychological Association. Gubrium, J. F., & Holstein, J. A. (Eds.). (2001). Handbook of interview research: Context and method. London, England: Sage. Goffman, E. (1981). Forms of talk. Oxford, England: Basil Blackwell. Hepburn, A. (1997a). Discursive strategies in bullying talk. Education and Society, 15, 13–31. Hepburn, A. (1997b). Teachers and secondary school bullying: A postmodern discourse analysis. Discourse and Society, 8, 27–48.Hepburn, A. (2000). Power lines: Derrida, discursive psychology and the management of accusations of teacher bullying. British Journal of Social Psychology, 39, 605–628. Hepburn, A., & Bolden, G. (in press). Transcription for conversation analysis. In J. Sidnell & T. Stivers (Eds.), Blackwell handbook of conversation analysis. Oxford, England: Blackwell. Hepburn, A., & Brown, S. J. (2001). Teacher stress and the management of accountability. Human Relations, 54(6), 531–555. Hepburn, A., & Potter, J. (2010). Recipients designed: Tag questions and gender. In S. Speer & E. Stokoe (Eds.), Conversation analysis and gender (pp. 137–154). Cambridge, England: Cambridge University Press. Heritage, J. (2002). The limits of questioning: Negative interrogatives and hostile question content. Journal of Pragmatics, 34, 1427–1446. Heritage, J. C., & Greatbatch, D. (1991). On the institutional character of institutional talk: The case of news interviews. In D. Boden & D. H. Zimmerman (Eds.), Talk and social structure: Studies in ethnomethodology and conversation analysis (pp. 93–137). Cambridge, MA: Polity Press. Heritage, J. C., & Raymond, G. (in press). Navigating epistemic landscapes: Acquiescence, agency and resistance in 570 ◆ PART VII. CRITICAL REFLECTIONS responses to polar questions. In J.-P. de Ruiter (Ed.), Questions. Cambridge, England: Cambridge University Press. Houtkoop-Steenstra, H. (2000). Interaction and the standardized survey interview: The living questionnaire. Cambridge, England: Cambridge University Press. Jefferson, G. (1985). Notes on a systematic deployment of the acknowledgement tokens “Yeah” and “Mmhm.” Papers in Linguistics, 17, 197–216. Jefferson, G. (2004). Glossary of transcript symbols with an introduction. In G. H. Lerner (Ed.), Conversation analysis: Studies from the first generation (pp. 13–31). Amsterdam, Netherlands: John Benjamins. Koole, T. (2003). Affiliation and detachment in interviewer answer receipts. In H. van den Berg, M. Wetherell, & H. Houtkoop-Steenstra (Eds.), Analyzing race talk: Multidisciplinary approaches to the interview (pp. 178–199). Cambridge, England: Cambridge University Press. Lee, Y-J., & Roth, W-M. (2004, January). Making a scientist: Discursive “doing” of identity and self-presentation during research interviews. Forum: Qualitative Social Research, 5(1), Art. 12. Retrieved from http://www.qualitativeresearch.net/fqs-texte/1-04/1-04leeroth-e.htm Leudar, I., & Antaki, C. (1996). Discourse participation, reported speech and research practices in social psychology. Theory & Psychology, 6, 5–29. Maynard, D. W., Houtkoop-Steenstra, H., Schaeffer, N. C., & van der Zouwen, J. (Eds.). (2002). Standardization and tacit knowledge: Interaction and practice in the survey interview. New York, NY: Wiley. Morgan, D. L. (2010). Reconsidering the role of interaction in analysing and reporting focus groups. Qualitative Health Research, 20, 718–722. Myers, G. (2004). Matters of opinion: Talking about public ideas. Cambridge, England: Cambridge University Press. Poland, B. D. (2001). Transcription quality. In J. F. Gubrium & J. A. Holstein (Eds.), Handbook of interview research: Context and method (pp. 629–649). London, England: Sage. Pomerantz, A. M. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J. M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 57–101). Cambridge, England: Cambridge University Press. Potter, J. (1996). Right and wrong footing. Theory and Psychology, 6, 31–39. Potter, J., & Hepburn, A. (2005a). Action, interaction and interviews: Some responses to Hollway, Mischler and Smith. Qualitative Research in Psychology, 2, 319–325. Potter, J., & Hepburn, A. (2005b). Qualitative interviews in psychology: Problems and possibilities. Qualitative Research in Psychology, 2, 281–307. Potter, J., & Hepburn, A. (2007). Life is out there: A comment on Griffin. Discourse Studies, 9, 277–283. Potter, J., & Mulkay, M. (1985). Scientists’ interview talk: Interviews as a technique for revealing participants’ interpretative practices. In M. Brenner, J. Brown, & D. Canter (Eds.), The research interview: Uses and approaches (pp. 247–271). London, England: Academic Press. Potter, J., & Puchta, C. (2007). Mind, mousse and moderation. In A. Hepburn & S. Wiggins (Eds.), Discursive research in practice (pp. 104–123). Cambridge, England: Cambridge University Press. Potter, J., & Wetherell, M. (1987). Discourse and social psychology: Beyond attitudes and behaviour. London, England: Sage. Potter, J., & Wetherell, M. (1995). Discourse analysis. In J. Smith, R. Harré, & L. van Langenhove (Eds.), Rethinking methods in psychology (pp. 80–92). London, England: Sage. Puchta, C., & Potter, J. (1999). Asking elaborate questions: Focus groups and the management of spontaneity. Journal of Sociolinguistics, 3, 314–335. Puchta, C., & Potter, J. (2002). Manufacturing individual opinions: Market research focus groups and the discursive psychology of attitudes. British Journal of Social Psychology, 41, 345–363. Puchta, C., & Potter, J. (2004). Focus group practice. London, England: Sage. Raymond, G. (2003). Grammar and social organisation: Yes/no interrogatives and the structure of responding. American Sociological Review, 68, 939–967. Sacks, H. (1992). Lectures on conversation (Vols. 1 & 2). In G. Jefferson (Ed). Oxford, England: Basil Blackwell. Schegloff, E. A. (2007). Sequence organization in interaction: A primer in conversation analysis (Vol. 1). Cambridge, England: Cambridge University Press. Stivers, T., & Hayashi, M. (2010). Transformative answers: One way to resist a question’s constraints. Language in Society, 39, 1–25. Stokoe, E., & Edwards, D. (2008). Did you have permission to smash your neighbour’s door? Silly questions and their answers in police–suspect interrogations. Discourse Studies, 10(1), 89–111. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self categorization theory. Oxford, England: Blackwell. Wetherell, M., & Potter, J. (1992). Mapping the language of racism: Discourse and the legitimation of exploitation. London, England: Columbia University Press. Willig, C., & Stainton-Rogers, W. (Eds.). (2008). The SAGE handbook of qualitative research in psychology. London, England: Sage.