Abstract
Social constructionists consider interviews as mutually co-constructing meaning. But what if the interlocutors do not seem to agree on what they construct? What if the interviewee has a particularly strong agenda, far from the intended research topic? Are these ‘failed’ interviews? We address this issue using a ‘deviant’ interview in a study of ‘being a neighbour’. First, we add to the discussion of interviewees’ category representativeness by acknowledging a situation when the interviewee insists on representing a category not intended by the researcher. Second, we address the notion of asymmetries of power, where it is often assumed that the interviewer has the upper hand. Through this case, we argue that the opposite may well be true. Finally, we argue that cases where the interviewee pursues a strong agenda may suggest new research areas. After all, strong efforts of resistance may indicate deeper cultural concerns.
Keywords
Introduction
Methodological literature on qualitative research interviews often warns against imposing the researcher’s definition and framing of a problem on research subjects. A textbook in research methods, thus, notes that to conduct a semi-structured interview, ‘Leave any assumptions you might have about the participant or the situation behind … Use neutral probes … to ensure that you do not lead your participant to make “acceptable” comments’. (Matthews and Ross, 2010: 231) By contrast, social constructionist researchers consider interviews as conversations, and thus view what conventionally may be seen as ‘leading questions’ as a natural part in the interview interaction. The interview is described as a site for co-construction, for shared knowledge and meaning-making (e.g. Koro-Ljungberg, 2008). Such interviewing implies including the interviewer’s participation in the analysis (e.g. Järvinen, 2001; Presser, 2004). But what if the interviewer fails to set ‘… the general pattern for responses, constraining as well as provoking answers that are germane to the researcher’s interest’, as Holstein and Gubrium (1995: 39) urge? What if the interlocutors do not seem to agree on what they construct? Our case deals with an interviewee who refrains from answering the interviewer’s questions in a reflective, thoughtful, or ‘pro-and-con-way’, as other interviewees did in the same project. Instead, this interviewee seemed to have an agenda of her own. Rather than simply explaining this ‘failed’ interview as a result of poor interviewing skills, we became curious. How was it different from the rest of the material, and why?
The interview in question stood out from a collection of 46 interviews investigating the meaning and experiences of being a neighbour. Like several other interviewees, ‘Emma’ states that she has harmonious and supportive relations with her neighbours. Unlike others though, Emma is the only one who speaks about her neighbours and neighbourhood uniformly and exclusively in positive terms, both in the past and present.
Nairn et al.’s (2005) recommendation to revisit an apparently failed interview inspired this article. Their analysis concerns a case where a group of students resisted conversation with the interviewer not only through silence but also through laughter. The researchers present their efforts to improve such interviews through changes of setting, getting to know the students in advance, and so on. Our case concerns another form of ‘failed interviews’: instances when an interviewee has an implicit yet obvious and strong agenda of her own. Needless to say, all interviewees have some sort of agenda, expressed in various kinds of identity work. They may highlight a preferred identity, underlining successes, sacrifices, competencies and so forth. Such presentations are often subtle and delicate, and achieved in collaboration with the interviewer (Åkerström, Burcar and Wästerfors, 2011). In this case, however, the interviewee’s agenda is especially strong, and could be described as ‘political’ in that she seems to act as an ambassador for her community.
The aim of the study ‘Neighbours and being a neighbour’ was to investigate reflections on and talk about being a neighbour, and had no interest in any particular geographical site. Emma ignored or distrusted this information. Instead, she continuously assumed she was selected as an inhabitant of Rosengård (rose garden) even though the interviewer initially tried to correct her. Rosengård is a council housing estate on the outskirts of one of Sweden’s major cities, which has become a symbol for the severely segregated, marginalized areas of Sweden. These areas, and especially Rosengård, are described in the media and research reports as having social problems, crime and a high density of immigrants or immigrants’ children (Hallin et al., 2010). This unbalanced, common image of Rosengård seemed to permeate the interview with Emma, causing her to appeal to another, equally unbalanced, description of her community: a praising picture.
Two well-known methodological topics of great significance to interviewing became apparent, although in different forms than we are used to from the methodological literature. First, we will add to the discussion of interviewees’ category representativeness by acknowledging a situation in which the interviewee insists on representing a category not intended by the researcher. Second, we address the notion of asymmetries of power that emerge between interviewer and interviewee where it is taken for granted that the interviewer has the upper hand. Our discussion, as in Nairn et al. (2005), concerns interviews marked by an inverted situation. Finally, we will emphasize that a focus on how interview participants’ talk may inform the interview topic (Gubrium and Holstein, 1997), and that such, perhaps, unexpected findings might suggest a new research topic.
Far too often, students and colleagues tend to neglect interviews they regard as ‘failed’ (for instance, when interviewees tend to talk about anything but the suggested topic). Rather, interviewers should pay particular attention to the ‘failed’, ‘deviant’ or ‘negative’ cases, and to put them through rigorous analysis (cf. Katz, 2001). Analyses of the often-neglected ‘odd’ cases might not only be analytically intriguing but also useful for teaching interviewing and qualitative methods.
The research interview as a site for various agendas
Sociologists interview individual people, but they do not interview them as individuals (cf. Kleinman et al., 1994: 43; Rapley, 2004: 29). Rather, the interviewee is chosen because he or she is a member of certain social categories (or a local culture) that are vital for the research project (e.g. ‘parents’, ‘deaf people’, ‘business men’).
With regard to researchers’ implicit use of membership categories when ‘collecting’ interviewees, Baker concludes, Interviewees are then made accountable members of that particular category in the sense that they then are meant to sound like members of that category and speak from within that particular, research-generated categorical incumbency and not some other. (Baker, 2002: 783–784)
Interviewees may, however, act as a representative for a category not intended by the researcher, and they may use the interview in creative and unforeseen ways not anticipated by the researcher. In Lois Presser’s (2004) study of violent male offenders, for example, the men who had murdered female partners and raped girls did not use the interview to explain their crimes, but instead accomplished moral selves by excluding themselves from the stigmatized social group of ‘violent offenders’. In a similar way, Juhila (2004) found her interviewees ‘talking back’ to the stigmatized identity of ‘people living in shelters’ in her study of shelter residents.
Such unintended presentations may well be expected as both interviewer and interviewee enter the situation with sometimes differing expectations of what will happen, why they meet, who the other is and so on. Methodological literature has pointed to the issue that arises when the interviewee does not agree on or know how to represent common ideas within a category. For instance, Baker (2002) tells about her studies of ‘adolescent-adult’ interviews where interviewees mostly were able to produce ‘a theory of adolescence’ in adult terms (p. 784). The interviewee who lacked such cultural knowledge, Baker says, obviously did not know how to talk as a member of that category. Another example is when interviewees do not even consider themselves as members of the research-intended category. In Järvinen’s (2001) interview study at institutions for heavily addicted alcoholics, it turned out that a few interviewees did not consider themselves alcoholics at all, and, consequently, they were not able to ‘represent’ the membership category the researchers had used to select interviewees for their study.
Järvinen (2001) discusses this phenomenon in terms of a too narrow preconception of alcoholics when designing the study, and she raises the important question of researchers’ ways of imposing definitions and reality descriptions on research participants. Similarly, the interview situation is often described as a site where the interviewer sets the agenda and more-or-less forcefully conducts an interview (e.g. Briggs, 2002). Certainly, the researcher is the more powerful of the two interview participants considering that recordings and transcripts rest in the researcher’s hands (e.g. Kvale, 2006; Rapley, 2004; Tanggaard, 2007; Van Enk, 2009). But during the interview, it might be the other way around, as is illustrated in the case we analyse here. The interviewee, Emma, seems to use the interview as a site to deliver a counter-narrative in dialogue with larger societal concerns, rather than in dialogue with an ‘active interviewer’ who sets the topics for conversation, or at least directs them (Holstein and Gubrium, 1995). In general, it is often elite interviewees – for instance, influential chief executive officers (CEOs), union officials or national politicians – who are assumed to have strong agendas in order to promote and protect their interests (e.g. Odendahl and Shaw, 2002; Thuesen, 2011). In the light of Emma’s interview, we emphasize that any interview may be understood as a site for doing public or political work (cf. Michael, 1996), and that such resources are not only restricted to the elite interviewees.
Methods and materials
The empirical example for this analysis is a set of interviews – and particularly one interview – that was gathered in the context of a graduate course as part of a qualitative methods exercise. Students in the master’s programme carried out 46 interviews on the topic ‘Neighbours and being a neighbour’. The interview guide was formulated on the assumption that ‘being a neighbour’ is an act of balance: there are not only codes of courtesy to honour but also limits to respect. The interview guide was put together accordingly, and included open questions regarding if, how and when interviewees socialized with their neighbours; the sharing of common areas of responsibility (such as a laundry room or the public areas); limits on what’s proper and not proper between neighbours and so forth. The qualitative approach was rather traditional in the sense that interviewees were asked about their experiences of ‘being a neighbour’, and the analytical ambition was restricted to organize the results into themes (cf. Roulston, 2010: 217). The material also includes students’ written ‘reflections’ on the interviews they conducted.
The interviews were tape-recorded and lasted for at least half an hour but more often for an hour. The interviewers transcribed all interviews verbatim; we have access only to the interview transcripts, since the students were told to erase the recording when the transcription was done. Excerpts (the written interview reflection as well as the transcribed interviews) were translated from Swedish to English by the authors with the help of professional translators. The longer excerpts are enclosed, in Swedish, following Nikander’s (2008) appeal for increased transparency and continuous efforts to produce ‘good enough transcripts and translations’ in qualitative research. 1
When we decided to look into the interview with Emma more thoroughly, one of us carried out another interview with Emma over the telephone, this time to talk to her about how she had perceived the previous interview. In addition, for the purpose of sharing the cultural knowledge of this area – for sketching the rather stereotypical common image or public discourse of Rosengård – we used last year’s media reports in the Swedish daily paper Sydsvenskan. Both Emma and her interviewer, Bella, have agreed to have their interview analysed for this article.
Analysis and findings
The analysis begins by contextualizing the interview with regard to Rosengård as a symbol for socially disorganized immigrant housing areas, and the interviewee’s disagreement on the purpose of the interview. In the following section, we will address how Emma’s talk differs from the other interviewees’ accounts, concurrently pointing to what Emma accomplishes with her descriptions. Emma’s narrative is not only a contrast to the other interviewees but also to the common image of Rosengård. In this sense, it could be said to be a counter-narrative (e.g. Bamberg and Andrews, 2004; Fisher and Goodley, 2007). Finally, we conclude by discussing what might be learned from this interview that seemed to go wrong.
Contextualizing the interview
From a much larger body of material, the one interview that draws our attention is the interview with Emma, a young woman, aged 23 years, who was brought up in Sweden, but was born in the Eastern European country that her family emigrated from. Emma speaks Swedish fluently, has a college degree, and is about to move to Stockholm where she has gotten a job as a ‘job coach’. For 12 years, Emma has lived with her parents and younger brothers in Rosengård, a segregated immigrant area, which is situated just outside central Malmö, a large city in the south of Sweden. Such building complexes are sometimes referred to as ‘concrete suburbs’ (Andersson, 2003), as they were mass-produced (in concrete) during the 1960s. Rosengård, as are many of the ‘concrete suburbs’, is a neighbourhood that is heavily marked by social problems such as unemployment, crime and dependency on social welfare.
To nearly all citizens of not only Malmö but also Sweden in general, Rosengård most likely triggers associations of social disorder. It is not a very physically deteriorated area, not a ‘slum’; it consists of fairly modern apartment buildings. Above all, the problems associated with Rosengård are social: high unemployment, poor school achievement often accounted for by the high level of ethnic diversity (around 50 languages spoken by immigrants from 111 countries), welfare dependence with almost a third of the residents financially supported by social welfare and a rate of migration to and from other neighbourhoods that is higher than average. 2
The area has a long history of attracting media and police attention, not to mention attention from social researchers (e.g. Carlbom, 2003; Flemström and Ronnby, 1972; Hallin et al., 2010; Popoola, 1998; Ristilammi, 1994). The media cover Rosengård almost daily: fires in garages, cellars, and cars; attacks on fire brigades and police cars with stones and eggs; muggings and thefts; bad housing conditions; poor schools and so on (Hallin et al., 2010). Recent articles on Rosengård in the daily paper Sydsvenskan consist of headlines such as ‘Police aim for dialogue in Rosengård’ (11 December 2009), ‘More fire alarms in Rosengård’ (23 November 2009), ‘Alarm on gunshots in Rosengård’ (2 November 2010) and ‘Rosengård is singled out as the worst ghetto in Scandinavia’ (7 November 2010).
Perhaps this intense exposure explains why Emma instantly draws the conclusion that the fact that she belongs to the category ‘people living in Rosengård’ was the reason for her being interviewed. In accordance with these assumptions, it appears that Emma sets out to counter the negative media view of her neighbourhood by presenting a problem-free picture. The interviewer has difficulties in being ‘the active interviewer’ as Emma continues to ‘make a commercial’ for her neighbourhood, rather than reflecting on being a neighbour, in a more neutral, open, or thoughtful way, as other interviewees had done.
For the assignment, the interviewers were asked to briefly discuss how they experienced the particular interview they had carried out. Thus, apart from the transcript they also handed in reflections on the interview situation. The interviewer, Bella, explains that she contacted Emma with the help of her mother’s acquaintances. According to Emma’s wishes, they decided to meet on her campus for an interview. Bella writes, The only problem I encountered was that Emma [the interviewee] thought she was singled out for an interview on neighbourhoods because of her living in Rosengård. She thinks that people who do not live there often have the wrong idea about Rosengård. I explained to her that the purpose of the assignment was not to analyse different neighbourhoods but rather [to talk to people about] relations between neighbours. Besides, I did not have the faintest idea she lived in Rosengård. (Excerpt 1)
Reading the interview transcript, it soon becomes clear that Bella’s initial information about why Emma was selected and the purpose of the interview are in vain. Emma insists on talking about her neighbourhood in more general terms, telling about ‘how people are’. She acts as a spokesperson for the residents, and restricts herself to sunny, positive and lively terms. In a later interview with Emma, conducted by one of the co-authors, she explains that she thought she was selected as an inhabitant of Rosengård, and that she was used to people always asking her about how it is to live there (‘Do you dare to go out at 2 o’clock in the morning?’) because they tend to believe the media reports on its danger: ‘They just see media reports, that “Rosengård is on fire”’. In this later interview, she concludes the conversation with the statement: ‘I’m very proud of who I am and that I’m from Rosengård’. It is important to stress that the aim of this analysis is to distinguish how Emma’s interview differs from the rest of the material and to discuss why this may be the case. We have no intention of questioning her experience of Rosengård as a pleasant neighbourhood characterized by solidarity between its members.
Speaking as an ambassador
In most of the interviews, the interviewees acted as representatives for the research-intended category ‘neighbour’; they related concrete experiences as they told stories of both good and bad neighbours. Emma, on the other hand, seems to take on the role of an ambassador speaking for an immigrant collective living in Rosengård. Rather than the descriptions of riots and violence that are frequently reported in the media (which she never mentions in the original interview), Emma invokes a picture of an old-fashioned joviality among immigrants, which may be formulated as a contrast to a common cultural discourse of ethnic Swedes as being reserved and socially inhibited (Daun, 1998):
Describe how you have contact with your neighbours!
I have good contact with my neighbours, um … we always greet each other when we meet outside. Perhaps it’s a little different since I live in a place where there are many immigrants. I think we can call each other friends, all of us. Chit-chatting about this and that every time we meet, we always stop and talk a little, more than the ordinary ‘hi’. My mom is friends with many women who live in my apartment complex [my building or close by] so I usually say hi to them too, just because my mom knows them. And then they ask how my mom is doing, and the family, and you ask them the same. It’s just how we talk to one another. But if there is something we are short of at home and if we need it right away, we have good contact with people who live in our building, we just go downstairs and ask if we can borrow something. It’s no problem. (Excerpt 2)
There are several elements that make Emma’s account seem more like a deliberate counter-narrative rather than a fairly straightforward description of ‘contact with your neighbours’. First, while answering the interviewer’s question about how she has contact with her neighbours, she also delivers an assessment that is unequivocally formulated as a positive statement: ‘I have good contact with my neighbours’. Furthermore, this is made instantly, with no pause for reflection or hesitation, which often was the case in other interviews (see the following). Second, she uses several extreme case formulations (‘always’, ‘all of us’, ‘every’): a classic rhetorical technique to ‘convince others to buy a product, to believe in an idea, or to support a project’ (Pomerantz, 1986: 219). This suggests that the interviewee is eager to convince the interviewer and that she suspects the listener of scepticism. Third, with the phrase ‘… I live in a place where there are many immigrants’, the interviewee invokes a collective. In the first five lines, Emma refers to a community of neighbours when she talks about ‘we’ and ‘us’ (in contrast to the last lines when ‘we’ refers specifically to her family). Fourth, in portraying her community, she seems to suggest that it differs from others in being more communitarian (implying better), when adding ‘more than the ordinary “hi”’ after ‘we always stop and talk a little’.
Neither extreme formulations nor images of a community of neighbours were common in the other interviews. For comparison, below is an example from another interview conducted by another interviewer:
The first question concerns how you have contact with your neighbours.
Well (pause) as a matter of fact we don’t have that much contact with our neighbours. We’ve said hello. (pause) [ … ] There is an old lady who is – what’s it called – the apartment house’s little police officer, who knows and checks up on everything. We don’t have much contact with her. There are also a few couples our own age that we say hi to and chat with and so on. But, none that we visit or anything. (Excerpt 3)
This excerpt is typical of how the rest of the interviewees answered the opening question. The interviewee starts to comment on the frequency of neighbour contacts (‘not that much’) and explains interactions with neighbours (‘we’ve said hello’). Also, those interviewees who told about ‘good contact’ did so in a matter-of-fact way, describing what neighbours they personally talked to.
The laundry room as a telling example
The interview with Emma differs in yet another way. Whereas the other interviewees always told a story or two of a particularly disturbing neighbour or a particularly picky or complaining neighbour, Emma delivers no such story. Most frequently, these stories concern the laundry room: more than half of the interviewees who shared a laundry room with other tenants (20 out of 36) had something to say on the matter. The main variation of the laundry room story was complaining about other people’s behaviour or about being the subject of complaints of improper laundry room behaviour. In general, in Sweden (with its many apartment houses), conflicts related to the laundry room seem to be important. In fact, they are so important that a magazine for apartment renters (Hem och hyra) conducted a survey where a thousand people answered questions regarding disputes in the laundry room. 3 In the light of this fairly common idea, that the laundry room is a place for potential conflicts, Emma seems to use the laundry room as a particularly telling example (Wästerfors and Holsanova, 2005) to illustrate her laudatory account of her neighbourhood. She has nothing but praise for her neighbours:
Are there any problems when it comes to the common areas, spaces shared by all the tenants? Say, for instance, the laundry room or the yard?
We have a laundry room, no problems there. Everyone cleans up and no one takes laundry from another. It’s very seldom we lock the door. We feel that it’s not necessary since we trust our neighbours. And there are three staircases [separate entrances], which make sixteen households in the building that share the laundry, so we trust each other very much. We have never had any troubles with the courtyard either. The solidarity between us is pretty good – we simply get on well. I cannot think of any incident regarding the laundry room or the outside common areas that has caused a problem. We simply get on well. (Excerpt 4)
Once again, Emma uses extreme formulations (Pomerantz, 1986) when she declares that there are no troubles whatsoever between the neighbours (‘everyone’, ‘no one’ and ‘never’). She also makes use of numbers to support the claim that they all trust each other: as many as 16 households share the same laundry room. Still, nothing bad has ever happened. Furthermore, in this excerpt, the collective story is particularly conspicuous. She refrains from talking about ‘me’ or ‘I’, but instead presents herself as the voice of the community: ‘we’ do not have any problems, summing up her description on this section of her talk with ‘We simply get on well’. In other parts of the interview (see Excerpts 2, 5 and 7), the ‘we’ reference is not as obvious as in Excerpt 4. Still, references to ‘I’ – the personally experienced – seem to reinforce the construction of a community ‘we’ by the persuasiveness of a first-hand account (cf. Potter, 1996). In Excerpt 2, when Emma is asked about how she has contact with her neighbours, she takes the opportunity to construct an immigrant collective where ‘we can call each other friends, all of us’. At the same time, she personally testifies to the fact that in her neighbourhood they (a) always greet one another, (b) chat every time they meet, (c) have inter-generational conversations and (d) borrow items from one another when necessary.
Emma’s description also differs from the 16 interviewees who, just like her, did not have any complaints about the laundry room. Whereas Emma uses rhetorical devices (extreme formulations and the counting of households) to convince the listener (cf. Potter, 1996), the other interviewees just stated that ‘it works just fine’ or said that they simply had a high level of tolerance since they valued good relations with their neighbours. The latter implies that there might be problems but that they have decided to ignore them, which is quite different from Emma’s declarations.
Again, it is important to make clear that the point is neither to verify nor doubt whether Emma has had laundry room problems and that, more generally, she actually finds her neighbourhood attractive. The point is rather how she presents her home quarters; she is constructing a flawless version of a neighbourhood in sharp contrast to the current (although not mentioned in the original interview) common cultural knowledge of Rosengård as a community characterized by social problems: crime, everyday disturbances, neglected houses and so on.
In sum, Emma’s talk is obviously different from the rest of the 45 interviewees’ in various ways. First, in the original interview, not only does she talk about her contact with her neighbours but she also evaluates them – she delivers her version without hesitation or signs of hesitant thoughtfulness. Second, she makes use of extreme formulations in an effort to construct an unambiguously positive description of her neighbourhood. Third, by invoking a collective ‘we’, she makes herself a spokesperson for the category ‘people-living-in-Rosengård’. Fourth, Emma’s laundry room version is a forceful telling example rather than less dramatic comments that ‘it works just fine’.
Sticking to the agenda
Järvinen (2001) draws attention to an interview situation from her own study in which researchers interviewed people at institutions for heavily addicted alcoholics: ‘Interviewers approached the interviewees as “heavy alcoholics” willing to relate their “drinking careers” and discuss their experiences (and possibly unmet needs) of treatment and care’ (p. 269). This approach worked fine in most cases: many interviewees identified themselves as alcoholics and they organized their life stories around how they developed drinking problems. However, interviewees who did not fit into this therapeutic model of ‘I acknowledge my problems and want to do something about them’ attempted to pursue an alternative line of self-presentation (cf. Baker, 2002: 784). According to Järvinen (2001), the narrow preconception of the alcoholic study forced ‘deviant’ interviewees to struggle ‘… to construct another identity for themselves than the one the interviewer, faithful to the therapeutic model of the study, suggested’ (p. 281).
In the case of Bella’s interview with Emma, however, it seems to be the interviewer who has to manage a respondent who is determined to defend her neighbourhood, although no accusations against it have been raised. In fact, the interviewer has made it clear that she is not at all interested in the specific neighbourhood. Still, Emma takes every opportunity to present a positive view of Rosengård. Bella continues the interview in accordance with the interview guide, now asking about ‘social relations in the neighbourhood’:
What about social relations? Do you spend time with your neighbours?
Nah, it’s not often I hang out with the girls who live in my complex. Most of them are younger than me. And there are no girls my age in my entrance [ … ] But the guy who lives downstairs – I used to see him sometimes. My brothers hang out with all the guys who live in our entrance and some other guys who live in our apartment complex. I have a friend, but she lives in the next complex. And my parents see others here when they have the time. I don’t think there are many people who know their neighbours as well as we do in this complex. Could be, eh … because we all have lived here for quite some time, and since most of us have a foreign background we share a feeling of solidarity. Do you see what I mean? (Excerpt 5)
Emma first accounts for why she does not spend time with some (the other girls are too young); this statement is then followed with descriptions of her own and her family’s neighbour relations. Then she returns to her ‘image-building’ of the collective, summing it up in: ‘… we share a feeling of solidarity’. In a sense, Emma treats the discourse on Rosengård as more important and urgent than Bella’s questions. Although both Bella (by continuing the questions on her interview guide) and Emma stick to their respective agendas, it seems reasonable to say that Emma, the interviewee, is the one most in control of the interview. Basically, Emma co-operates with the interviewer, in that she offers personal accounts of who she says ‘hi’ to and chats with. These personal accounts, nevertheless, seem to serve the overall purpose of telling a counter-narrative of Rosengård: this is a cheerful and pleasant place to live.
The same tendency is evident when potentially problematic relations or phenomena are explicitly brought forward, such as neighbours who refrain from helping out when ‘something bad is happening’. Emma stresses that it is an imaginable situation rather than a factual one: ‘I haven’t encountered that, but that’s something that can happen’. Another question explicitly dealt with problems in the neighbourhood concerning crime and safety. This is also answered with a negative, which makes an interesting comparison with other interviewees and with the common image of Rosengård.
Many studies about neighbours deal with safety and fear of crime – is this something you have thought about?
I really don’t know what to say about that. I feel safe in my courtyard and with my neighbours so I don’t know what to say about that. We have never had a break-in, or anyone at our courtyard, as I recall at least, so I don’t know what one should say about that [the interview ends here]. (Excerpt 6)
In the rest of the material, 26 interviewees answered the same question. 4 About one half stated that they were quite worried about their possessions and/or personal safety, mentioning break-ins in apartments or cellars, or being afraid to walk home alone at night. The other half mentioned, just like Emma, that they felt safe in their neighbourhood but also added comments and stories about stolen bicycles, cars set on fire, break-ins and the importance of locking the door and having neighbours watching out.
We end this section on ‘sticking to an agenda’ with a discussion of power relationships. One basic aspect of such a relation is who determines such an agenda. When discussing power relations between interviewer and interviewee, researchers often regard the interviewer as the more powerful of the two (e.g. Kvale, 2006; Rapley, 2004; Tanggaard, 2007; Van Enk, 2009). Indeed, it is the researcher who plans the interview and has the interview recording or transcript at his or her disposal for future use, often beyond the control of the interviewee (see, for instance, Watson, 2006). Even during the interview, the power relationships are visible in that they tend to work in favour of the interviewer’s version of reality: Rapley and Antaki (1998) have shown how interviewers can generate views in line with their own position, rather than with that of the interviewees (see also Baker, 1997). In the case of Bella’s interview with Emma, there seems to be something different going on. The interviewee seems to be the one who takes the lead in the way of directing the general tone and the message that she wants to convey. This is a more subtle power relationship than the one Nairn et al. (2005) described. In our case, the interviewer reported that it was difficult to ‘discuss’ the various themes, and that Emma delivered ready-made answers, which were difficult to respond to with follow-up questions.
Discussion
‘Failed’ interviews often raise interesting methodological questions, and they may also be useful for teaching interviewing and qualitative methods (Roulston, 2010). In the discussion of what may be learned from a ‘failed’ interview, we will summarise two methodological points: first, regarding category representativeness, and second, asymmetries of power in the interview situation. Then, the discussion will focus on a third point: analysing how people talk may inform what they say and thereby indicate a new research topic.
First, respondents who belong to ‘high-interest’ social categories, as for instance, ‘handicapped’, ‘criminals’, ‘immigrants’, or, in this case, ‘someone living in Rosengård’, may make unwarranted assumptions about the researchers’ interest in them: that they are asked for an interview because they belong to a particular social category. Such assumptions are probably well founded, as social researchers tend to turn to individuals and expect them to represent a particular category (Baker, 2002; Kleinman et al., 1994: 43). Furthermore, interviewees who strongly identify themselves with a certain group may be reluctant or astonished at the idea of speaking about topics from other subject positions in their lives. In those cases, the interviewer faces a hard task in asking questions that do not necessarily relate to ‘being disabled’ or ‘being a criminal’ or ‘being someone living in Rosengård’. Above all, it may be difficult to convince the respondent of the researcher’s interest, ‘beyond’ this over-shadowing social category. In the case discussed in this article, the interviewee took the researchers’ interest in her for granted, guided by assumptions of what people and the media generally know about the subject (or category). Hence, interviewees may use researchers’ interests in them to correct or refute perceived misunderstandings, thus avoiding or refusing to share personal experiences or stories with the interviewer. Instead, the interviewee with an agenda may turn the interview into an arena for producing – and insisting on – a preferred version of a problem.
Second, the interview presented here is an example of the asymmetry of power in the interview situation, being more in favour of the interviewee (Emma), as opposed to what is often highlighted in the methodological literature. Mostly, the assumption is that the interviewer is the more powerful party (e.g. Briggs, 2002; Kvale, 2006; Rapley, 2004; Tanggaard, 2007; Van Enk, 2009) – an idea that is not always recognizable to researchers in concrete interview situations. After all, we cannot force people to agree to an interview nor to talk about things they do not want to talk about.
Many researchers, and not only graduate students, have experienced interview situations of indeed feeling inferior to the interviewee, and unable to get to the point of ‘mutual co-construction of meaning’ around themes that are central to one’s research. There are several ideas that may explain why this happens, for instance, when the researcher is considerably younger than the interviewee or when one is interviewing powerful members of the community (cf. Odendahl and Shaw, 2002). In addition, as this analysis shows, interviewees with an agenda might take control of the interview for the purpose of correcting perceptions they deem unfair, no matter how peripheral this ‘debate’ is to the researcher. ‘Losing control’ of the interview may well be due to an inexperienced interviewer (here, for instance, Bella fails to pose follow-up questions), still the analysis of how the conversation proceeds is a clue to what the interviewee considers important.
Our third and final point relates more directly to the fact that an analysis of how interview participants talk may lead to unforeseen findings and suggest new areas for research. After all, strong efforts of resistance may indicate some deeper cultural concerns. In the case of Bella’s interview with Emma, far from the quite feeble inquiries on ‘being a neighbour’, the analysis points out an alternative and (for these researchers) new research question: ‘How do people living in areas marked by low social prestige respond to the negative and stigmatizing associations they encounter in everyday life?’ Kusenbach (2009) points out that research on how the location of residence is associated with identity is sparse. She investigates the stigma of ‘trailer living’. The main way her respondents managed the stigma implied by their community (‘white trash’ implications) was by distancing themselves from the label ‘trailer trash’ and identifying themselves as ‘mobile home residents’, while identifying some other mobile home residents as belonging to the category ‘trailer trash’. There was no embrace of a collective identity, ‘we in trailer homes’, such as the ones found in Emma’s accounts, ‘we from Rosengård’. Our case, similar to findings in ethnographies of contemporary Swedish ‘concrete suburbs’ (Andersson, 2003; Bäckman, 2009), seems to suggest that the embrace of a collective identity might exist among residents of a marginalized community, and may warrant a broader study on the topic. In Andersson’s (2003) ethnographic study on teenage girls living in another multi-ethnic and socially stigmatized neighbourhood, it is evident that many of them are annoyed at what they describe as biased media reports, demonizing their neighbourhood: ‘The suburb is also discussed in terms of the girls’ emotional assessment of it as their home district and as a safe and tolerant place where people are seen as equal regardless of origin’ (Andersson, 2003: 246, translated).
We conclude by discussing the analytic approach that led to our new research interest. Many qualitative researchers have strongly emphasized the significance of the interview context on how interviews unfold (e.g. Rapley, 2001; Van Enk, 2009). Apparently, the interviewees’ expectations of interviewers and their research purposes are important clues to the ethnomethodological invitation to analyse interview conversations with regard to what the interview participants aim to accomplish (e.g. Baker, 1983, 2002; Roulston, 2006; Ryen, 2004). The interview analysed here also points to the necessity of cultural or ethnographic knowledge that goes beyond the immediate interview context. Without knowing anything about the wider cultural context, the social status of Emma’s neighbourhood, it would be hard to claim or even notice an interviewee with an agenda to ‘prove them wrong’ (the media, people in general, etc.).
We have argued that in this case, cultural understandings and expectations are made discernable (even if not openly stated). The interviewee, Emma, rather than talking to Bella, the interviewer, is responding to or arguing against a ‘public view’ or a ‘public discourse’. In this case, the interviewer turns into a passive audience, a feature of interviews that is quite common, although not always treated as important guidance for analysis. Where such ‘non-interaction’ answers or statements are given during interviews (Jacobsson, 2000; Thelander, 2006), we may hear and read them as clues to efforts at change or resistance; respondents seem to engage in a Bakhtinian dialogue with a meta-narrative that they perceive as strong or hegemonic during a certain period of time, or particularly intense in a special situation or context, or among a category of people. A starting point for a dialogic process, Bakhtin (1981) says, is when hegemonic discourses are questioned, when alternative stories start to emerge. In this case, Emma actively makes use of the interview for the purpose of telling that alternative story.
Concurrently, counter-narratives of Rosengård crop up in various contexts and in different forms. For instance, newspaper stories of particularly noteworthy people who are said to contribute to the area contradict the many negative reports. The success story is another form; the most spectacular is about Zlatan Ibrahimovic, a world-famous soccer player who grew up in Rosengård. Social researchers are likewise keen to present a nuanced picture of the area. For instance, a thematic issue of the journal Praktik och Teori [Practice and Theory] (2009) was devoted to Rosengård with an explicit aim to counter prejudiced ideas of the vicinity. Why this resistance? In her study of mobile homes and stigma, Kusenbach (2009: 401) stresses, ‘… the intricate bonds that exist between the places of our daily living and ourselves’. The belittlement and vilification of one’s neighbourhood are serious threats to identity, and the inhabitants of such areas often resist or challenge negative stereotypes in various ways (Palmer et al., 2004).
In contrast to the rest of the material, the interview with Emma suggests a very different research question than the one we started out with: a research question identified and prompted by the analysis of an apparently failed interview where the interviewee’s obvious agenda was forced through a dialogue, not with the interviewer but with an existing and deprecating discourse on her neighbourhood.
An interview may seem to have failed, straying far from the intended topic, or in other ways irrelevant to the research questions. But by paying attention to these ‘failed’, odd or deviant interviews, we might find them surprisingly lucrative analytically. Here, the deviant interview has triggered more analysis than all the other interviews together. As such, an initially failed interview in terms of the here-and-now interaction (Rapley, 2004) can get a career of its own by triggering analytic curiosity. Ultimately, the failed interview might turn into a rather successful one.
Footnotes
Acknowledgements
The article has benefited from valuable discussions with Jaber Gubrium on the excerpts, and from Ann-Mari Sellerberg’s and Maggie Kusenbach’s thorough reading of previous versions, as well as helpful comments from the anonymous reviewers. We also thank ‘Bella’ and ‘Emma’ for allowing us to examine their interview in detail.
Funding
This research was funded by FAS - Swedish Council for Working Life and Social Research (Dnr 2009-0011).