Feel-bad moments: Unpacking the complexity of class,gender and whiteness when studying ‘up’

Introduction: ‘This is boring. How long are you going to go on?’

The first encounter with my potential research participants within the Swedish Women’s Educational Association (SWEA), a network of Swedish women who live or have lived abroad, was a welcome dinner for members who had recently moved back to Sweden. The women, most of whom are privileged by class and colour, had been engaged in what Amit describes as a ‘voluntary movement’ made by ‘people who have the resources’ to be able to be on the move in a global landscape (2007: 2). At the dinner I was acting as a participant-observer but I also presented the research project, in order to make contact with women who were interested in being interviewed. After I had been describing the project for a few minutes, an elderly woman who sat at the same table as the members of the SWEA board and the president of the chapter raised her hand and asked: ‘This is boring. How long are you going to go on?’ I did go on, but I immediately became uncomfortable, because I felt that she not only wanted me to stop talking, but also tried to put me in my place.

As Skeggs argues, researchers ‘are located in the economic, social and cultural relations which we study’ (1997: 18). As a white feminist researcher from a working-class background, I invested much effort in thinking through questions of closeness/distance as well as similarities/differences. The present article aims to explore the feminist methodological challenges when doing fieldwork with privileged women. In order to do so the following questions are addressed: How do similarities and differences in multiple social positionings between the researcher and the research participant affect the fieldwork? In what ways can the uncomfortable moments during the research process be understood from the framework of feminist methodologies? What does the class-related discomfort experienced in the research interactions by the researcher and the participants reveal about how class ‘translates to feelings’ (Faber and Prieur, 2013: 8)?

Studying ‘up’ sheds a different light on feminist methodologies based on assumptions about similarity, shared experiences, intimacy and political alliance (see for instance Oakley, 1997 [1981]). This mirrors what Becker and Aiello highlight: ‘Calls to study up are easily found, but one major issue remains under-explored: the lack of fit between feminist, anti-racist, or social justice methodologies and studying up’ (2013: 63). Departing from previous feminist interventions in research about studying ‘up’, this article contributes by analysing class, gender and whiteness as a part of the shifting power relations between the researcher and the research participants. ¹

In addressing the uncomfortable aspects of doing research on priviliged groups, I adopt the work of Knowles (2006) as a starting point. Knowles emphasises that feminist methodological approaches have long been consensus-centred and focused on the importance of building ‘non-hierarchical relationships’ based on intimacy and familiarity with research participants, as well as advocating ‘allegiances as a legitimate part of the research process’ for researchers (2006: 394). In contrast, Knowles takes an encounter between herself and a British expatriate diving instructor living in Hong Kong as an example showing that an antagonistic relationship can also be productive. In adopting Knowles’s invitation to researchers to immerse themselves in the uncomfortable, the analysis in this article concentrates on how class, gender and whiteness are produced in three different fieldwork situations with women who hold privileges in a global postcolonial and capitalist landscape.

In the next section of the article, the feminist methodological challenges in studying women with privileges are discussed. Moreover, in this section the emotional dimensions of research relationships are highlighted as an important part of feminist methodologies. Thereafter, I describe this particular study on privileged Swedish return migrant women. In the following empirical sections, the first example illustrates how class dynamics as well as white privileges made the researcher feel uncomfortable in an interview setting. In the second example, I analyse how the researcher’s and the research participant’s different class backgrounds and gendered pasts affected the interview. In the third example, the focus is on the problem of discussing class even though the researcher’s and the research participant’s class backgrounds were similar. Finally, drawing from these examples, I suggest that uncomfortable research moments can offer us insights into the complex processes of the ways class, gender and whiteness are made.

Feminist methodological challenges in studying ‘up’

Within feminist qualitative research, the relationships between researcher and research participants have been debated for a long time (Behar, 1996; Oakley, 1997 [1981]; Olesen, 2008; Reinharz, 1992; Riessman, 1987). In feminist research, mutuality, intimacy and empathy with the research participants have been emphasised: for instance, as Reinharz (1992) illustrated by the closeness/distance dilemma. In the distance approach, it is assumed that the researcher should not identify too much with the participants. Feminist researchers have advocated for closeness, intimacy and empathy in order to achieving a better understanding. A third approach is suggested by Reinharz, namely that the setting shall determine how the researcher works. Furthermore, she points to the fact that the researcher has different relationships with different research participants in the field and states that ‘as long as the researcher is self-aware, whatever happens is useful data’ (1992: 68). Dilemmas of feminist ethnographers have also been discussed by Behar (1996), who argues for an approach where researchers allow themselves to be vulnerable and make their emotions part of the ethnography.

The assumption that the research relationships should be non-hierarchical has been dismissed as ‘naive’ by Skeggs (2001 [1994]: 79). Phoenix throws into question the comfortable position for a feminist researcher interviewing women by arguing that ‘cosiness does not simply come from shared gender but is often partly the result of shared social class and/or shared colour’ (2001 [1994]: 51). Furthermore, when researching privileged women, the assumption that the researcher is almost always in a superior position within the research process becomes more complicated. For instance, as Ross (2001) argues in a study about politicians, researching women in elite settings implies different kinds of tensions for feminist researchers: the women participants are accustomed to controlling discussions and they are used to the exercise of power. Thus, doing fieldwork with privileged groups challenges some of the basic assumptions in feminist qualitative research. Becker and Aiello’s research on people with power and privileges also contests some of the methodological premises within feminist, anti-racist and social justice research approaches, such as giving ‘voice to the voiceless’ or trying to even out the power relations between researcher and participants: ‘When research participants have plentiful resources, enjoy privilege(s), and wield power regularly, such methods are inappropriate’ (2013: 71).

When Nader was seeking answers to why researchers prefer to study ‘down’ rather than ‘up’ in her influential article on ‘studying up’, one of her explanations related to the researcher’s own subjectivity: ‘Anthropologists value studying what they like and liking what they study and, in general, we prefer the underdog’ (1972: 303). In Priyadharshini’s (2003) analysis of the validity of Nader’s assumptions for contemporary research, she drew the conclusion that researching underprivileged groups contains ‘a feel-good factor’ for researchers. Doing research with privileged groups can place the researcher in a different position compared to doing fieldwork among subordinated groups. However, it is important to keep in mind that research interactions with privileged groups and individuals are not one-dimensional, but rather dynamic, since privileges and vulnerabilities are not always clear cut. An important insight from Knowles’s (2006) work is that difficult research relationships can produce vital information about the complexity of power and privilege. However, it is important to clarify that it is not the informant who is ‘difficult’, but rather the dynamics between the researcher and the research participants that create the difficulties (Knowles, 2006: 397).

The feminist turn to affect is useful in understanding the tensions that emerged during the fieldwork in this particular study. In the literature, an early call to investigate the importance of emotions in the research process, so as to understand the power dynamics involved, came from Ristock and Pennell (1996). They discuss how the complex and sometimes contradictory social histories (such as race, class, gender and age) shape the interaction between the researcher and the researched. Therefore, influenced by Foucault, Ristock and Pennell call for analysis of the ‘microphysics of power’; the argument they present is that ‘the interpersonal and the structural relations that affect the research process’ require examination (1996: 65). As Hoel suggests, there is a need for reflection on the emotional and affective dimensions in the research relationship that goes beyond the well-known feminist methodological discussions about ‘positionality and locatedness’ (2013: 41). Reflections on being vulnerable can, according to Hoel, be as important as reflections on the power dimensions involved. Expanding these arguments further, Ahlstedt (2015: 187) re-names the research process as ‘doing feelwork’. Furthermore, according to Ahlstedt, ethnographers do not sufficiently acknowledge how their personal histories and identities affect the feelings aroused in the field.

In recent feminist research on emotions, Ahmed’s framework is particularly useful. Ahmed focuses on what emotions do (Ahmed, 2004a). Hence, the ways in which we live with and within different power structures are important so as to understand the ways in which ‘the past is carried in the body’ (Spanger, 2015: 114). Ahmed (2004a) understands comfort and discomfort by relating to which bodies can feel at home in the world, or not feel at home. For Ahmed, to be comfortable is to feel safe and at home, but also to feel that you belong in certain spaces. The possibility to feel at home is central to Ahmed’s understanding of how different power structures make different bodies become oriented in different directions. Ahmed’s phenomenological understanding of emotions open up for an analysis of the unequal as well as structural dimensions of our lived experience of ‘fitting in’ and ‘passing by’ as well as feelings of discomfort and disorientation (Ahmed, 2004a).

Research methods and research context

This project investigates the lived experiences of return migration of a group of privileged migrants. The empirical material consists of 46 interviews with returning Swedish migrant women; plus participant observation during eight months of fieldwork within two networks for Swedes who have returned back ‘home’ after years abroad. ² Additionally, the women were asked to draw a social network map during the interviews. The women participating in the study were between 33 and 80 years old. A majority of them were in their sixties or seventies. Their lives abroad were mainly structured around taking care of children, supporting their spouses by arranging social events and dinners, becoming employers and being in charge of domestic workers and involvement in voluntary work, charity work and social clubs. Their lives after returning to Sweden include extensive travelling, participating in different networks and doing voluntary and/or charity work. Some of the women in the study have gone back into the labour market after returning to Sweden.

The interviews were based on a semi-structured questionnaire and lasted between one and two and a half hours. Both the interviews and fieldnotes were manually coded, following thematic fields on belonging, migration, gender vulnerabilities as well as whiteness and class privileges. The analytic steps can be described as going through the stages of ‘rereading, strategic reading and careful contextual rereading’ (Tollin, 2011: 32, my translation). The analysis was focused on what the women told me and how I responded, how they narrated their stories and what they or I did in the field.

This article is based on three uncomfortable moments from the fieldwork, but its arguments are informed by the whole sample of 46 women. The three selected cases are chosen, first, because they illustrate a pattern in the material that emerged during the analysis, namely how the emotional aspects of the interaction during fieldwork could provide important sociological knowledge. Second, the uncomfortable moments experienced during these three interviews have crystallised my theoretical insight on how emotions are linked to the lived experiences of gender, whiteness and class as well as the structural dimensions of those positions.

The Swedish Women’s Educational Association facilitated my possibility to conduct participant observations. The SWEA network includes over 7000 members worldwide, even though some of the largest SWEA chapters are actually in Sweden. The internal class structures within SWEA are clear: upper-class women hold the more important positions and middle-class women the less prestigious ones, mirroring the class hierarchies in society at large (Lundström, 2010a). By means of contacts within SWEA, I was also able to follow another network, ‘Swedes Back Home’. This is not a network for women only, but a group of women from the network formed a subgroup to which I was invited.

The possibility of gaining access to the different SWEA chapters varied. Two chapters were welcoming and allowed me to attend their meetings, while in the chapter described at the beginning of the article, the president turned my requests down time after time. Later on, I was told by research participants that a small group of upper-class women in this chapter were well-known within SWEA for their way of exercising power over other members. Quite a few women in the network raised this as a problem and described chapter leadership as ‘posh ladies’. A number of research participants also expressed empathy, recognising that the episode during the dinner mentioned above was an uncomfortable moment for me. Several of the women who contacted me in order to be interviewed told me that one of the reasons they wanted to participate was precisely because they reacted to the dismissive comment made by the woman at the dinner. This volunteering for the interview may reflect one of the advantages of studying ‘up’. When research participants possess privilege, they may not feel threatened by the fact that a researcher wants to gain access to them since they do not feel any pressure to agree to participate. Underprivileged groups might feel that pressure because they occupy a different power position.

The women’s class backgrounds were not a clear-cut category: they ranged from women who went to the same school as members of the Swedish Royal Family to women who were upwardly mobile from the working class. Class, here, is understood as a relational concept, which describes the relations of exploration and conflict that exist between different groups and their interests in a capitalist structure based on unequal power relations (Skeggs, 1997, 2004). I consider class to be dynamic, and reproduced through symbolic, as well as material, factors (Lawler, 2005; Skeggs, 2004). The Swedish upper class can be seen, as Holmberg suggests, as ‘a particularly privileged layer, which owns assets and exercises power in society through exclusive social networks and organizations’ (forthcoming, my translation). What constitutes the middle class historically is, according to Lawler (2008), the distinction from both the ‘lower’ classes and the aristocracy. Following Lawler, middle class here is used in order to ‘denote a normatively desirable and indeed “normal” and “natural” status’ (2008: 247).

In the following, I use my position, as a white upwardly mobile woman who is younger than a majority of the women participating in the project, in order to investigate the feelings from the fieldwork (see Ahlstedt, 2015). For instance, how did it happen that I could not establish the same closeness with the women who were – similarly to me – upwardly class mobile in this project, as I had managed to do in my previous research project on women’s upward mobility in the Swedish context (Sohl, 2014)?

‘It wasn’t dangerous at all being interviewed’

Sabine lived abroad for six years with her husband. She began the interview by telling me that they lived in South Africa during the apartheid years, since her husband had an executive position with a Swedish company that ran a business there. Sabine’s family lived in a large house with a big garden with a pool and a high wall around it, rented by the company her husband worked for; the house was owned by one of the leaders of the apartheid regime. Sabine explains further: ‘So we lived in the stronghold of apartheid.’ The ties between the Swedish company Sabine’s husband worked for and the apartheid regime were close, even though the sanctions debate and the solidarity movement in Sweden clearly wanted Swedish companies out of South Africa. Knowing that, I asked Sabine what it was like, as a Swede, to live in South Africa during these years and Sabine responded: ‘It was safe, because no blacks except the maids were walking the street.’

At this point in the interview, I was not able to ask more than the most basic follow-up questions. It was clear that I lost the thread when Sabine so openly talked about the racial and economic privileges she lived with during apartheid. Class and race privileges, in the interview with Sabine, are intimately bound up with and manifested in the lifestyle she and her husband were able to maintain within the system of global capitalism because of the company he worked for. Drawing on Fechter’s discussion on how expatriate women’s ‘ideological labour’ within capitalism – as well as women’s participation in the colonial enterprise – tends to be downplayed, I suggest that Sabine’s story highlights both a structural dimension, about how a Swedish company took economic advantage of the racist apartheid regime, and the lived experience of privileges of class, gender and whiteness (2010: 1294). For me, the interview with Sabine was one of the most challenging in the project, and during the interview I had a strong sense of being uncomfortable as a result of Sabine’s description of her life as a white woman with class privileges living in ‘the stronghold of apartheid’. In the interview situation, I realised that I was provoked by the fact that Sabine and her husband had been benefiting economically from, as well as being a part of upholding that racist apartheid system.

Moreover, the discomfort I felt in the interview with Sabine can be understood drawing on Ahlstedt’s (2015) analysis when she was doing research on queer partner migration: how awareness of her own privileges as white, Western and middle class made her frustrated and angry by the lack of awareness demonstrated by her research participants with similar backgrounds. In the interview with Sabine, my feelings of discomfort can be understood both as a consequence of my own anti-racist awareness and as a reaction linked to the Swedish political context. The Swedish self-image rests on the notion that Sweden is one of the most egalitarian and gender-equal countries in the world (Berggren, 2014; Lundström and Twine, 2011). The national self-perception is based on Sweden being associated with ‘good’ values and the country is (falsely) assumed to lack a racialised history, but such history is actually part of the national project of tolerance and solidarity (Pred, 2000). Against this background, my reaction can be understood as a reflection of this skewed national self-image. Sabine’s sincerity and openness about her own privileges and benefits from living in a country with policies of racial segregation enforced by law presents a sharp contrast to the Swedish – contested and debated – national self-image. Therefore, as I listened to Sabine, I was both surprised and antagonised by the sense of entitlement that she expressed when talking about her privileged lifestyle in South Africa during the apartheid years. My own reaction can, however, also be understood in relation to Ahmed’s framework on the cultural politics of emotions and racism. She suggests that the ‘very claim to feel bad (about this or that) also involves a self-perception of “being good” ’ (2004b: 28). In other words, my feeling of being uncomfortable can be understood as a way of distancing myself from Sabine’s way of expressing her white privileges.

Later on, when I met Sabine at a SWEA event and she was talking to another SWEA member about being interviewed by me, she said: ‘It was not dangerous at all being interviewed. We just talked for a bit.’ At that point I did remember my own sense of being uncomfortable during the interview, and understood that our experience of the interview situation was completely different. As Knowles (2006) describes, she found it hard to handle when the research participants used stereotypical orientalist views or prejudices about women, even though that was a pattern in her material. As Knowles explains her own reaction in the research encounter with the British expatriate diving instructor: ‘[The] Diver’s story grated against my politics, my conceptions of women and myself; and it was difficult to repress my feelings of antagonism towards him’ (2006: 397). In the case of Sabine, she seemed to regard the interview as an occasion where we just ‘talked for a bit’, while I was provoked by the class and racial privileges Sabine described. Being provoked by a research participant when studying privileged groups may not be the best way for a researcher to act, but instead of pretending that these tensions do not exist, it is better to acknowledge them and try to understand the sociological implications of these tensions (see Knowles, 2006: 402). In the next example I elaborate on this argument by analysing an interview encounter where the different class backgrounds of the research participant and myself created the tension.

‘One of my nannies came from there!’

Elsa moved away from Sweden with her husband and describes her life abroad as ‘very nice and privileged’. Her upbringing in Sweden was privileged as well, since she went to the same school as the members of the Swedish Royal Family. Both abroad and in Sweden Elsa has hired domestic assistance, such as cleaners and nannies. Because of her upbringing and her way of living, she has gained the feeling that other people see her as what she describes as ‘posh’. Towards the end of the interview Elsa suddenly asked me if I had lived all my life in the university town where I reside now. I told Elsa that I moved there 18 years ago, and then she asked:

The place where I grew up is, if people have heard of it, associated with the white working class. Until Rimbo was mentioned the interview had been quite relaxed, but I immediately became uncomfortable after having to ‘reveal’ my (class) background. Lundström’s experience from fieldwork with the SWEA network in the US was that the anonymity provided by the migratory context made her ‘comfortable with the fact that the women could hardly locate my social or geographical background’ (2010b: 81). By contrast, I was afraid that the women would ask me about my background or where I grew up, since that would clearly indicate that I was raised in a working-class environment. As suggested by Ristock and Pennell, both as researchers and in our lives outside academia, we sometimes experience ‘uncomfortable gaps between our own feelings and intentions, on the one hand, and how we are perceived, on the other’ (1996: 70). I was not comfortable talking about my own class background with Elsa. This contrasts with my previous research project (Sohl, 2014) on upwardly mobile women of my own age possessing similar feminist and political values. Ahmed describes discomfort as ‘a feeling of disorientation: one’s body feels out of place, awkward, unsettled’ (2004a: 148). With Elsa, I was afraid to ‘reveal’ my class background because it reminded me of my family’s position in the Swedish class society. To me, her comment was a way of distinguishing herself from the working-class positions that used to be mine – and still are my parents’. The encounter with Elsa made me aware of the different class trajectories Elsa and I had experienced. In fact, my mother’s first job was working as a maid and nanny for a wealthy family. My expected life route – staying in the working class as most people do where I grew up – could have potentially positioned me as one of Elsa’s employees. The tensions that I experienced in the interview with Elsa can be understood using Knowles’s discussion on the ‘researchers’ personal baggage’, namely the ‘emotional, political and intellectual positioning aspects of a researcher’s make up and modus operandi’ (2006: 394, 403). Rather than avoiding the personal aspects of research encounters, Wilkins highlights the benefits of analysing one’s own personal responses, because researchers do sometimes ‘take it personally’ (1993: 93). Wilkins argues that even though feminist methodologies explore the personal aspects, such as emotions and autobiographical dimensions, in research relationships, there is still a tendency to overlook ‘difficulty and difference’ (1993: 94).

As highlighted by Skeggs, ‘Class is a relationship between people who inherit not just different categories, but also who inherit the values of those categories, the inequalities and injustices’ (2010: 355). The reasons for my unwillingness to reveal (or one could even say my fear of revealing) my class background with these privileged women was grounded in my understanding of class as an ‘inequitable system of difference and differentiation’ (Lawler, 2008: 256), which manifested itself in the interview situation with Elsa. As argued by Skeggs, class struggle cannot be understood only in terms of collective action: ‘it is also about the positioning, judgements and relations that are entered into on a daily and personal basis’ (2004: 173). Moreover, as Bourdieu describes it, a class is a way in which ‘social distances are inscribed in the body’ (1987: 5). To ‘know one’s place’ is for Bourdieu an unconscious and embodied process. Here it is important to point out that the knowledge of one’s own location also includes knowing others. Based on how class is lived and how classed subjectivity is created, I understand the encounter between me and Elsa as a moment when we both became aware of our places in the class hierarchy.

Since a majority of researchers have middle-class origins, the similarity in class background with research participants belonging to the middle class has been highlighted as a methodological asset (Benson, 2016; Riessman, 1987). However, when differences in class background between researchers and research participants are present, this brings into focus the ways in which class is lived. As suggested by Faber and Prieur (2013), distinctions and the drawing of boundaries are central to how class is made. The emotional dimensions of class can be understood using Kuhn’s (2002 [1995]: 117) often cited quote: ‘Class is something beneath your clothes, under your skin, in your reflexes, in your psyche, at the very core of your being.’ Thus, my emotional reaction in the research encounter with Elsa can be understood as a classed moment where the feelings moving around reflect relations not only in the field but also outside the field.

In the following discussion, I will extend the analysis to cover how class is played out in a more outspoken way than in the interview with Elsa. However, in contrast to the two previous examples, the woman being interviewed was left feeling uncomfortable, not the researcher.

‘I don’t really like that question’

Ellen has lived abroad for a long time with her husband. They returned to Sweden nine years ago and are now living in a house in the wealthiest area in a larger town. I ask Ellen about the importance of class in the process of moving from Sweden. She does not understand the question and asks me: ‘What did you say, about …?’ After I have repeated the question, Ellen declares that she ‘grew up in an ordinary working-class home’ and then she explains:

Ellen: No, I am not thinking about class and I definitely don’t belong. … If I ever think about class, I think about the upper class and the aristocracy because in this part of Sweden there are a lot of aristocracy and workers. … But today I can’t say I am a worker. … I am not thinking in those terms. I don’t really like that question, because it’s not me.

Up until I asked Ellen about class, the interview situation had been quite relaxed. We were sitting in a room in their house, looking out over their beautiful garden. Immediately after bringing up class, the interview situation became tense. The interview with Ellen was one of the last in the project, and by this stage in the research process I had begun to see ‘the feel-bad factor’ as important data. Ellen describes herself as an upwardly mobile woman, moving from a working-class background to an upper-middle-class position. She discussed her upbringing in the working class with ease, but it was clear that my questions about her current class position made her uncomfortable. As Skeggs (1997) shows, the reason that working-class women in her study did not want to use the concept of class is because it captures precisely why they do not want to use it. Ellen, who has a working-class background but now finds herself in an upper-middle-class position because of her upward mobility, expresses an awareness of difference in values with which class positions are connected; and I argue that this might be why she tried to avoid the concept of class.

The interview with Ellen illustrates a frequent situation during the research, namely that the women tried to avoid my questions about class. Since their migration experience as part of a white transnational elite was marked by class, I was interested in exploring their own views about the importance of class in their lives. I argue that Ellen’s disapproval of the question about class can be seen both as a way of downplaying her own privileged class position and – at the same time – as a way of performing an upper-middle-class position by telling me that the question is not a proper one to raise. As Lawler argues, it is assumed that ‘while class does matter to some people, it ought not to; and, furthermore, that to notice “class” (in the sense of social distinctions) is to display a lack of “class” in the sense of a set of personal characteristics’ (2008: 254). What is interesting here is that Ellen in the interview describes her life as ‘extremely privileged’ without using class as the frame for understanding her own position. One way to understand this is that the ideals of social justice and equality are concepts connected to the self-image in Sweden (Pred, 2000). Therefore, Swedish society sometimes seems to try to ‘cover up’ class divisions (which are of course extremely real), just as it seeks to cover up other sorts of social divisions, such as racialised ones.

However, it is also possible to analyse Ellen’s unwillingness to talk about her current class position in relation to the internal class hierarchies within the SWEA organisation, as previously mentioned. From that perspective, Ellen’s problem with the question about class can be seen as reflecting her position as an upwardly mobile woman, participating in a network run by women with upper-middle-class and upper-class origins. Class hierarchies were an important part of my struggle in getting access to the field and, as mentioned earlier, class hierarchies also informed the relationships between the members of SWEA, which may be reflected in the interview with Ellen. From that perspective, one dimension of class here might be a shared feeling between me and Ellen of being uncomfortable with ‘the upper class and the aristocracy’.

Unlike a majority of the women in the study, Ellen has a background in the working class. In the interview with Ellen, this classed moment carried the possibility of us sharing experience: we were both upwardly mobile women with working-class origins taking part in the activities of the SWEA network. Despite these similarities, I did not succeed in making her comfortable enough to talk about her class belonging, rather she seemed to be insecure during that part of the interview. The limitations of woman to woman rapport are described by Duncombe and Jessop, who argue that rapport sometimes does not occur ‘because the social and emotional distance between researcher and interviewee proves too great’ (2012: 113). Here, Ellen’s uncertainty about talking about her class belonging could be understood in the light of the emotional difficulties surrounding upward mobility (Sohl, 2014). Ellen’s struggle with the concept of class can also be understood in relation to the Swedish context. This can be analysed as a way of inscribing herself in the Swedish equality discourse; the Swedish society seems to be ‘classless’ in comparison with other countries. However, the women’s ways of speaking (or not) about class can be understood by drawing a parallel with the way Knowles (2003) argues that whiteness is rarely taken up explicitly in white people’s narratives. Similarly, talking about class from a privileged class position – even if in the past Ellen belonged to the working class – can be challenging because this might potentially imply that you occupy that position at someone else’s expense.

Conclusions

In this article, I have drawn on my own class baggage (see Knowles, 2006) when analysing feelings of being uncomfortable as a researcher, or causing uncomfortable feelings to my informants. As researchers, we sometimes ask people about their experience of things that they might not otherwise talk about. As highlighted by Richardson (1990), most people do not talk about how sociological categories such as race, class and gender shape their lives. However, as shown by Skeggs (1997), working-class women can be very aware of their class position without talking about it. It can be argued that my strategy of mentioning class in itself might have evoked various uncomfortable feelings among some of the participants. Therefore, it may be possible to understand the women’s emotional reactions to my questions about class in the interview situation as an expression of their class belonging and position within the class structures. In addition, as Lawler argues, class cannot be understood solely based on objective external criteria: ‘class is not simply an “objective” position which one occupies, but becomes configured into subjectivity’ (1999: 6). Thus, class can be understood as part of an unequal system of injustice with both political and personal dimensions that affect relationships between the researcher and the research participants.

When I struggled to understand what was happening in the interviews using my own class-related feelings in the encounters analysed here, I came to realise that these uncomfortable moments may provide important insights into the complex processes by which class, gender and whiteness are constituted. However, there is no doubt that, as a researcher, one can also seek to avoid class as a question in the same way my respondent Ellen did: because class can entail difficult judgements and struggles for us all. Despite this temptation to avoid class, holding on to such classed ‘feel-bad moments’ from the field can provide a truly invaluable – though painful – research strategy.