A social phenomenon is not necessarily sensitive a priori, as the German sociologist Jörg Bergmann (1998) has shown. It is all about framing, and topics can be sensitised or down-played based on who speaks and with whom in a specific situation (Silverman 1997). Nevertheless, migration can be seen as particularly inflammatory because it has been associated with racism, oppressive politics and ideology in the past and still is today. The way immigrants are characterised has become a matter of political and academic debate in the post-colonial world, as for instance discussion about the labour market or, as here, associations to crime. In this article, we briefly review some of the issues and concerns raised in these debates with regard to crime, focusing particularly on Sweden. But these concerns and issues arise in and pervade ongoing research processes as well. In the body of the article, we explore the details of these processes as a way to examine and discuss how qualitative researchers go about handling sensitivities related to research that involves ethnicities and crime. The article is based on our overall reflections on this area of research, having had experiences in this field that were characterised by a particular sensitivity that we have not encountered in other, similar fields where our work did not focus on ethnicity. The sensitivity we will discuss concerns choice of words and descriptions of behavior in relation to the research process: access, data collection, writing and discussions of findings.
This sensitivity, however, can be visible for researchers in various fields in migration research. According to an article in Curie, a Swedish science magazine, migration researchers often find themselves in the spotlight, and their results may be called into question (Östlund 2018). As Bosworth et al. (2018: 1–2) note ‘… the polarised nature of the political debates over migration and the competing objectives of the different stakeholders make this field of study particularly demanding’.
One especially contentious issue concerns crime and immigration. The tensions may trace to strident claims from radical right-wing political parties linking immigration to crime and other sorts of social unrest (Elgenius & Rydgren 2017; Rydgren 2017). Public political debates on crime as associated with immigrants have also spilled over into research in Sweden. An example is the heated debate among political party leaders about whether the Swedish National Council for Crime Prevention should perform a new study on crime that cites the country of birth of perpetrators (see, e.g., Barr et al. 2017, Brottsförebyggande rådet [The Swedish National Council for Crime Prevention] 2018). The Swedish media have brought Sweden’s Nordic neighbours into such debates, as when one of Sweden’s largest newspapers featured an interview with Grete Brochman, a Norwegian sociologist and migration expert. Brochman stated, ‘As a social scientist, I am a supporter of the best possible documentation of different social problems. If certain social categories systematically show worse results in relation to criminality, employment and conditions of living, we need data to manage this as a political problem’. The same article reported that both Denmark and Norway publish statistics on the country of origin of criminals and that Danish and Norwegian migration researchers are said to claim that ‘Sweden is more “careful” when it comes to analyses and debates that might give rise to conflict or confrontation between natives and immigrants’ (Barr et al. 2017).
Concerns about ‘care’ and ‘sensitivity’ affect not only statistical findings but also basic research processes, not least research based on ethnographies and other qualitative methods. These issues have emerged as critical in our own studies that focus on young people with an immigrant background – especially young men – as crime victims or targets of crime control.
The focus of this article is on the research process. Research about issues of immigration associated with crime may be seen as harbouring risks for those being researched and risks for those doing the research. Research identities may be questioned when writing about crime and immigrants. We discuss various forms of balancing societal ideologies in relation to ethnicities, finding ourselves in a situation similar to that described by the anthropologist Bonnie McElhinny (2001). She studied White police officers caught between conflicting demands, between being race-blind and race-conscious. In Sweden, where we conduct our research, we are urged to ignore ethnic categorisations or immigrant categories while being encouraged to be particularly observant of them to overcome structural discrimination. These conflicting imperatives create deep tensions in the research process: We need to show that our research does not presume or perpetuate claims of essential differences but at the same time does acknowledge differences in experiences between natives and persons with an immigrant background, or among various ethnicities.
Our case, as mentioned previously, concerns young people with an immigrant background and issues pertaining to crime can be seen as delicate. It concerns people who may be viewed with suspicion by some but also surrounded by a ‘sympathetic understanding’ because of a general knowledge of such oppression and discrimination (Åkerström & Burcar 2016: 256–258).
In contemporary Swedish society, the mass media show documentaries and reports of suffering refugees and the hardships of struggling new immigrants. At the same time, the mass media often describe young men of immigrant backgrounds in very negative ways. In Norway, we have seen examples of ‘sensational media stories about “problems” with youth gangs and forced marriages’ (Andersson 2003: 74). Lately in Sweden, ‘unaccompanied’ young men have been described in media as ‘dangerous’ and associated with discussions on terrorism and youth riots (Herz 2018: 2). Moreover, in the last few decades, the Swedish media have paid intense attention to crimes committed in socially disadvantaged areas, where many immigrants live; at times, almost daily reports have appeared with regard to ‘gang violence’ or young people attacking the police, ambulances or fire brigades (Elaies & Jaconelli 2020; Hallin et al. 2010; Hasni 2018; Rothoft 2017).
Such images also lead to a certain sensitivity, a sensitivity that has a long history, as noted. But the same phenomenon also can be discerned in fairly recent Swedish history in political and policy language and in recommendations about appropriate concepts. Because our interviewees and we as researchers relate to such changes, we will briefly describe the proscribed language changes. In Sweden, ‘immigrant’ (‘invandrare’) replaced the former term ‘foreigner’ (‘utlänning’) in the 1960s because the meaning of the latter term was considered to have shifted; it had come to mean something alien, derogatory and distinctive (Regeringskansliet Begreppet invandrare [the concept immigrant], Ds 2000:43, Kulturdepartementet). For a period of time, the suggestion was that instead of ‘immigrants’, the term ‘New Swedes’ (‘nysvenskar’) should be used, as ‘immigrants’ also was seen as derogatory as is related in this ironic quote from a the journalist Kalle Kniivilä, who is from Finland but who have moved to Sweden:1
“You are practically Swedish.” I have heard this many times, as a compliment. As if it were a shame not to be Swedish. Well-meaning Swedes have also come up with the fine term “New Swedes” which can be used instead of “immigrant”, “Arab”, “Finn” or other ugly words. (Kniivilä 2001, our translation)
However, New Swedes was not a successful concept and is not used very often. More and more, authorities seem to use ‘ethnicity’ instead, as in the call for research from the Swedish Crime Victim Compensation and Support Authority (Brottsoffermyndigheten 2014) and the Swedish National Board of Institutional Care (Statens institutionsstyrelse 2013). This might be a general Western trend; as U.S. sociologist Mitch Berbrier (2008) pointed out, whereas researchers earlier used words such as ‘race’, ‘nation’ and ‘tribe’, they increasingly use the word ‘ethnicity’.
Recently, the term ‘racialised’ (rasifierad) has been used and debated in Sweden (Sandberg 2017: 50–51). This term was included in the Swedish Institute for Language and Folklore’s new list of words in 2014 and is defined as a ‘person who, due to appearance, is assigned a certain racial belonging and attributes associated with it’.2 Some anti-racist debaters argue that this term should be used to highlight the racism common in today’s society, whereas others argue that use of the term may risk instigating racist thinking instead of expelling it (Sandberg 2017: 50–51).
The people whom we interview or follow in field observations – who either identify as immigrants or live in marginalised neighborhoods – are often quite aware of the negative images in the media coverage described previously and in the language used and may set out to counter the negative media view of their neighbourhood and themselves (Andersson 2005; 2010; Jacobsson & Åkerström 2012; Uhnoo 2009). After all, they argue, the vast majority of young men with an immigrant background or people living in marginalised neighbourhoods are law-abiding citizens. Also, the very terms we used may, we thought, be seen as derogatory. Using the terminology of conversation researcher Harvey Sacks, social categories that refer to ethnicity, nationality, immigration and race in Anglo-American contexts can be ‘inference rich’ (Sacks 1995: 40–41). They can be potentially regarded as stereotyping, that is, ‘potentially treatable as an objectionable attribution of “stereotypical” characteristics to members of those categories’ (Whitehead 2009: 327).
However, issues of word choices and categorisations are not alone in having been described as sensitive. At times, reports of other types of behaviour (not talk) by stigmatised ethnic categories may be seen as increasing the stigmatisation and deemed as sensitive either by the researchers or by readers and publishers (Andersson 2003; Bäckman 2009; Jerzy Sarnecki and Jack Katz, personal communication).3
Concerning word choices and descriptions of various behaviours, we have needed to walk a fine line when seeking access, when we have collected our material and when we have written about and presented our findings. When people interact in different ordinary, everyday situations, one can assume that they are aware of the potentially delicate nature of some descriptive attributions, such as ways of speaking Swedish, visible religious signs or where they live, as neighbourhoods have potentials of signaling both stigmatisation and prestige. The same holds true for researchers doing fieldwork or qualitative interviews. Studying issues touching on ethnicity or immigrants entails not only categorising others but also being categorised as a researcher by the members of the local culture being studied.
Below, we will see how ethnographers may go about handling, reflecting on, or interacting within the almost built-in intricacies when researching issues of deviance, crime or crime victims as associated with immigration or ethnicities.
In this article, through examples from ethnographic studies in Sweden, we elaborate on the intricacies and delicate formulations by interviewers and interviewees and the sometimes more unguarded expressions by the young interviewees.
The illustrations from our own research include the following:
Our common theoretical inspiration for these studies has been drawn from an interactionist, ethnomethodological perspective. In line with this approach, we have used so-called ‘active interviewing’ (Holstein & Gubrium 1995), where the interviewer and the interviewee are seen as co-constructing the phenomena and social categories in question.
In the current article, we have analysed the different studies, as Vaughan (2014) proposes in arguing for the advantages of exploring a particular phenomenon with qualitative case studies. Such a comparison can illuminate more general traits of events, activities or phenomena than collecting more of the same or larger samples of similar formats in the same context.
In this article, we thus return to previously collected material in the form of interviews and field notes, along with more general research notes and email conversations. Our empirical material is very rich, so we have chosen to use illustrations that we believe are representative of and persuasive for the arguments we wish to make. In some cases, we have stayed with some examples to expand them with more detail and nuance.
Interviewing potentially vulnerable individuals and groups, such as victims of crime (which we have done in several studies), requires care and ethical sensitivity. In addition, talking about ethnic backgrounds in studies about young victims and offenders can be delicate and ethically problematic. We have therefore very carefully and consistently followed the Swedish Research Council’s ethical guidelines (Vetenskapsrådet, n.d.).4
In these varied research projects concerning youth and crime, we have encountered the delicate nature of this specific field. These intricacies and considerations become visible during different parts of the research process. We have come across problems concerning sensitivity with accessing young people who have been placed or have placed themselves in categories that relate to ethnicity or to ‘youth with an immigrant background’. When collecting data, we encountered difficulties in asking questions (the researchers’ issues) and in retold stories (the interviewees’ issues). Furthermore, sensitivity concerns arise when we, or other researchers, write up reports or articles, and we as researchers have to deal with responses from colleagues to what has been presented at conferences and seminars.
To gain access to the young people we were interested in, we needed to contact different categories of professionals for assistance or permission. Access may always be a problem in ethnographies, especially so perhaps when it comes to vulnerable categories protected by various gatekeepers, as was the case for Cochrane (2018), who studied asylum-seeking mothers from Iran and Afghanistan in Australia. She accounts for the hesitation on the part of gatekeepers, such as representatives from different non-governmental organisations, in terms of both preventing exploitation of their beneficiaries during research and the sense that this subject was over-researched.
In our case, however, we had conducted studies in the same areas but not explicitly focusing on ethnicity (focusing instead on victims, detention homes and on voluntary policing) and not met with any difficulty even though many whom we interviewed had an immigrant background – both staff and clients – which was relevant and integrated in initial presentation of research questions and in suggestion for data collection. In the cases discussed here, though, requests for access explicitly mentioning ethnicity triggered prevention or delaying manoeuvres when heads of institutions, teachers or superintendents understood that the studies involved young people with immigrant backgrounds being victims of crimes. Furthermore, the professionals who acted as gatekeepers explicitly noted the categories when accounting for their reluctance. They referred to that it could be sensitive to talk to immigrant youths about experiences of crime when explaining why they were hesitant to help us get in touch with interviewees. This could include what we could label a ‘vicarious sensitivity’, that is that gatekeepers deem our candidate interviewees (the young people) being especially sensitive, or it would seem to imply a risk for themselves to get their own organisation involved in a topic deemed sensitive by others, thus stigmitising the organisation.
One such study concerned the actualisation of ethnicity in young victims’ retold experiences. We were interested in whether ethnicities would be emphasised and a ‘discrimination discourse’ or ‘racist discourse’ would be used, or whether issues of ethnicity would be activated at all. In short, we investigated the performance of identities and self-presentations in the interviews referring to multiple selves, such as ‘young man’, ‘immigrant’ and ‘victim’. We decided to search for ‘ethnic Swedes’ who had been mugged or assaulted by young people with an immigrant background, as well as the reverse, that is, young people with an immigrant background who had been victimised by ethnic Swedes. Through the District Court, Burcar Alm succeeded in finding what she assumed were ethnic Swedish youths who had been subjected to crimes by persons she considered to have an ethnic background other than Swedish by the simple method of searching for surnames. This method is certainly not very reliable, but we did find many interviewees this way. People in the reverse category – youths with an immigrant background who had been assaulted or robbed by ethnic Swedes – were difficult to find using this method, and the search process took time. We invited a researcher, Anna Rypi, who previously wrote about victims of crime (Ryding5, 2005), to help us find young people with immigrant backgrounds who had been subjected to crimes by Swedish youth.
The following is taken from Rypi’s description of how she searched for interviewees through other people’s contacts and through different organisations. After a few failed attempts, a support centre for crime victims was contacted:
After a week, the crime victim assistant emailed me and wrote that they had decided that they could not ask the victims if they wanted to be interviewed because they would not ask about their ethnicity, as this could be perceived as offensive. (Logbook from field diary, our translation)
Accordingly, the representative of Crime Victims Sweden did not want to participate based on the rationale that associations with ethnicity can be perceived as offensive. She took on a position of guardian gatekeeper, as did various school representatives:
Of those who responded, someone referred to a person responsible for student health. Other heads of the institutions were less cooperative and seemed to think that it was a very sensitive subject. One head of the institution said, for example, “You cannot just come here and talk about this [project] during a lesson, you do not know what you stir up in the classes.” The same answer is given by a teacher (Rypi, email).
The professionals who were asked were unwilling to help us with contacts. Someone referred us to student health, although it is not clear why, and others explained their lack of helpfulness within therapeutic discourse, ‘You do not know what you stir up’. Rypi, however, continued her efforts and eventually succeeded in getting permission to visit a school and present the project. This situation was one of the few in which the young people themselves were hesitant to be included in the study. A few pupils asked rather provocatively, ‘Do you mean there’s a difference?’ Perhaps they thought that ethnicity or their ‘immigrant-background status’ would be used as an explanation for being involved in violence or crime, even if as victims. In general, though, the attributed sensitivity among the young people appeared to be established local knowledge among teachers, heads of departments and employees at the crime victims support centre, even before the young people were asked if they wanted to be interviewed.
A reflection of the difficulty contacting young people with immigrant backgrounds who have been victims of crime is that another, more positive attitude could be seen as attractive. ‘Immigrant boys’ are said to be portrayed as more problematic, for example, as more criminal than young people with Swedish backgrounds (see Lindgren 2009, on mass media representations and Peterson 2010, on police), and the interviews could be perceived as an opportunity to provide a different picture. The interviews could recount experiences from young people with immigrant backgrounds who would be presented as worthy of sympathy (Clark 1987) as victims of crime rather than as perpetrators. School leaders, crime victim support and teachers could have willingly offered their help and encouraged young people to participate as interviewees.6
Returning to the pupils’ comments, when Rypi presented her study of victims with an immigrant background in a school class, questions about ethnicity or being an immigrant appeared to be provocative. The pupils used the sensitivity in a rhetorical manner when they asked the interviewer, ‘Do you mean there is a difference? (between ethnic Swedes and those with an immigrant background)’ This question implies that there are no differences, and the students can be said to have been drawing on an ideology of equality. Their assumption is similar to that of a head of a youth care institution to whom a doctoral student spoke to gain access for a study of everyday life at these institutions. We wanted to investigate whether ethnicity was important in everyday life at these institutions, given their over-representation. It was not easy for the doctoral student to convince the head of the institution to take part in the ethnographic study, even though the Swedish National Board of Institutional Care, which governs the institution, had funded the project and prioritised the question. Before the doctoral student met with the head of the institution, she had emailed and spoken to him several times, and he had expressed strong scepticism:
After some persuasion, today’s meeting was booked. The head of the institution begins by establishing that at this institution, ethnicity is not of any real or major importance, and he thinks it is less important than socioeconomic conditions, e.g., throughout the whole conversation, the head of the institution is very keen to indicate that ethnicity is less relevant than what the media and “others” want to emphasise (from field-note in Åkerström 2016: 21).
The institution head’s initial hesitation in granting access, as indicated by the multiple phone calls and initially unanswered emails, can be seen as attempts at ‘cooling out’. Such a strategy may pacify those who are frustrated and lessen their involvement and commitment, which is not uncommon in bureaucratic organisations (Sellerberg 2008). We cannot know for certain if this strategy was only about the study’s focus on ethnicity, but it did seem to be part of the problem.
Even when the doctoral student succeeded in organising a first meeting, there appeared to be an attempt to ward off the study. The head of the institution explained that the study was uninteresting and also tried to counter descriptions published by the media. It can be seen as noteworthy that the head of the institution and other employees at the youth care homes in other interviews pointed out several differences in family backgrounds, upbringing, experiences and ways of acting as they relate to national or ethnic backgrounds, as did the pupils in Rypi’s interviews on experiences of being a victim, although they had been indignant about her wanting to interview them.
When researchers collect empirical material on ethnicity, immigrants and crime, difficulties may be observed in both field observations and conversational interviews. These difficulties may relate to the expressive caution observed by Silverman (1997) with regard to delicate subjects. In a study of young girls from immigrant backgrounds, ethnologist Åsa Andersson (2003) describes a related delicate moment. During a conversation with a few teenage girls, one of them mentioned with ‘excited eagerness’ that one of the girls, Haideh, has been engaged since she was 12 years old, followed by ‘the laughter of the girls and my strained cheerfulness’ (Andersson 2003: 125). After Haideh explained that it is common among her compatriots, Andersson braced to keep asking about the subject, and it is clear that it was not easy; she restarted with a rather loaded question and then faltered, ‘But do you think it’s good to get engaged so young?’ (Andersson 2003: 126, our translation). Andersson relates how she left her neutral research role and end up with a value-laden question revealing her own sets of cultural values.
Sensitivity is not confined only to particular subjects (i.e., early engagements). The sensitivity may be about the designations used to refer to ethnic or national categories. In interview studies, in which one does not know the participants’ word choices through long-term observation, the problem is probably more apparent. As noted, Rypi had difficulty finding interviewees, and when she finally found them, choosing words was not simple for her. Below is an excerpt in which she is talking to Darin, who has a Middle Eastern background, about school. He has said that with regard to bullying and mugging, there is a divide: ‘Usually, it is foreigners and Swedes. Foreigners feel more at home, are more comfortable with those who have a different background’. Rypi continues to try to develop Darin’s differentiation between the categories:
Anna: Hmm, mmm. But it is thus divided like between Swedes and immigrants. Maybe that’s an incorrect word? It is thus divided between Swedes and non-Swedes?
Darin: It depends. It depends mostly on where you are, if I am to be honest. Different schools, my school, there the majority are foreigners. (Åkerström 2013: 52, translated by authors)
Darin has no problem using the distinctions ‘foreigners and Swedes’. Nevertheless, Rypi signals her openness to the designation, which provides an opportunity for the co-actor to correct it. The little addition, ‘Maybe that’s an incorrect word’, shows that the interviewer is not an expert in terminology and is willing to change. In other words, it is a form of ‘hedging’, an ‘intentional signal of minimal commitment, an indication of willingness to receive discrepant information, change opinion, be persuaded otherwise or be better informed’ (Hewitt & Stokes 1975: 4). Darin does not signal notice of Anna’s small insertion but continues to talk about the subject, maintaining his designation ‘foreigners’ and smoothly rejecting Rypi’s suggestion. In this short conversational exchange in the previous excerpt, three different designations were used to label the same category: immigrants, non-Swedes and foreigners.
During the interviews, the interviewees could at times ask the interviewer or fieldworker about their ethnicity, as when Burcar Alm was asked if she was ‘en svenne eller blatte’, that is a Swede or a ‘wog’ (blatte is a derogatory word for those immigrants who are visibly not ethnic Swedes). Such questions were asked if the interviewees had an immigrant background, and the researcher by name or appearance signalled that she also had such a background. This was true for the doctoral students who did fieldwork on detention homes: Goran Basic and Maria Seger, as well as for Burcar Alm. They all encountered at some point these kinds of questions from the young people they met.
We thought that such questions often suggested efforts to create something of a social bond. Burcar Alm possibly gained trust when she answered ‘both’ to the question about whether she was Swede or a ‘wog’, because she then could be assumed to at least partly identify with the young people and she implicitly accepted their way of categorising.
The interviewees could also turn the issue of a researcher’s ethnicity into research advice, however. Darin, quoted previously, complained that the police and media produce an image of immigrants as being criminals, even though Swedish youth have provoked troubled situations, too. Relating to this, Rypi asks if Darin can explain why it is so hard to find interviewees with immigrant backgrounds who can give their version of events. This young man offers several explanations. Foremost, ‘interviews’ are associated with police interrogation or media reporting, and taking part in an interview may be seen as ‘snitching’, a statement that was also found in Walklate’s (1998) study of a deprived neighbourhood in England. Darin offers a solution:
Darin: One of the things could be that they would talk to someone who is more like them. Then an immigrant who might know how it is and how it feels to be like that … They would more easily open up to such a person.
Anna: hmm, mmm. Even if it’s an older one?
Darin: Older may also be OK. It is easier for, for example, Ahmed to express himself to Mohammed than to Karin or Ann-Kristin.
Darin’s ideas emphasise one of the ‘markers’ of someone belonging to an ‘immigrant category’, one’s name (Carling et al. 2014). His advice to let immigrants do the interviews, as it is easier for ‘Ahmed to express himself to Mohammed’, is based on a general conclusion that emerges from the idea of a spirit of community among immigrants, who supposedly share experiences. It is also, however, about an epistemological hope that being an immigrant makes it easier to understand and talk to other immigrants (for similar findings, see Damsa & Ugelvik 2018: 205). As mentioned previously, in a study of safety work in Landskrona (Wästerfors and Burcar 2014), Burcar Alm was asked by potential interviewees with ethnic minority backgrounds if she was ‘a svenne (Swede) or a wog’, as if the answer to this question would decide if they would talk to her. The answer ‘both’ seemed satisfactory. She and her research colleague were asked the same question, although in less explicit terms, by adult safety workers with immigrant backgrounds.
Regardless of whether Darin’s advice is helpful, we can see how he is given an expert role as an interviewee. The issue touches on researchers’ discussions of positionality or how researchers’ positions affect interaction with informants touching on issues of identities as ‘insiders or outsiders’ (Aliverti 2018; Carling et al. 2014; Tewksbury & Gagné 1997). In the research project on safety work in Landskrona, Burcar Alm’s affiliation was determined based on the answer ‘both’. She experienced it as being assigned an insider identity.
When researchers write up their findings, they have many choices to make in terms of expressive caution. One aspect concerns what type of words to use when referring to the categories being studied. An obvious difference is the use of “race” in research, surveys and census data in the United States, whereas this concept is not used in Europe.
In Sweden, as we mentioned previously, ‘ethnicity’ has become a common concept replacing the term ‘immigrant’. Furthermore, ‘ethnicity’ may be used to denote all people who do not have a Swedish background. The expression ‘other ethnic backgrounds’ has become accepted shorthand for ‘other than a Swedish background’. This shorthand was actualised and discussed during seminars on several manuscripts in a 6-year large research program on Policing Ethnicity (Forte 2009–2011), reported in (Peterson and Åkerström 2013). These researchers were not alone, as this terminology – describing someone of ‘other ethnic background’ – can also be found in reports on how authorities choose epithets (see, e.g., The Swedish Integration Board, Integrationsverket [2006: 5]).7
Another wording issue concerns the inclusion of derogatory words used by the participants in a study. In a report on the meaning of ethnicity in everyday interactions in youth care institutions, the following scene was used to illustrate how ethnicity could bridge the divisions between staff and the young people in care through playing around with slang that might otherwise be derogatory:
The sun shines and one of the girls, Simone, and the treatment assistant, Halim, have gone out to the terrace. Simone paints her nails, and Halim reads a newspaper. There is music that Simone has chosen from loudspeakers. Soon, I (the researcher) and Petter, a person from the staff, join them. Madine, a young girl, comes out on the terrace. She is on her way to buy snacks.
“What’s up gangsters?” she says.
“What’s up my nigger” answers the treatment assistant Halim (field note) (translated from Åkerström 2016: 40).
Åkerström hesitated to include this scene in a report because she feared that the field note would reflect negatively on the staff, given that the ‘n-word’ is a very negative epithet: often used as a racial slur. In the end, it was included because it was said with a smile and expressed some of the playfulness that occurs at times between staff and the young people at the youth care institutions, similar to the ‘play fights’ Wästerfors (2014) described. His analysis showed that such observations also served as a contrast to using the same or similar derogatory epithets in conflicts at the youth care institution.
Hesitancy with regard to what to write and include may not always concern negative epithets but could involve descriptions of behaviour not seen as proper for a category that could be further stigmatised. This is illustrated in an ethnography written by ethnologist Maria Bäckman (2009) about ethnic Swedes living in council housing estates on the outskirts of one of Sweden’s major cities. These areas are described in the media and research reports with a strong focus on social problems, criminality, and the high rate of immigrants or immigrant descendants.8
Bäckman’s Swedish informants spoke about being uncomfortable with their Swedish neighbours who claim that immigrants are responsible for littering. ‘Bodil was at this occasion worried about how to treat the subject (littering) without appearing racist. I had, shame on me, to confess that I did not feel very sympathetic towards this dilemma. Silently, I had dismissed most of her stories as stereotypical prejudice’. Afterwards, Bäckman is walking around alone in the area when she sees an elderly immigrant woman with a baby carrier: ‘Suddenly she [the immigrant woman] stops and bends down. She pulls out a trash bag with some effort from the wire platform under the carrier. With unexpected force, she swings the trash bag several meters into the bushes beside the gravel path’. Bäckström writes of her embarrassment about her ‘initial self-righteousness as an enlightened and unprejudiced academic’ and how Bodil’s dilemma has become her own (Bäckman 2009: 72). If she includes it in her book, the concern is that she may contribute to a ‘generalised image of “immigrants” who are not capable of taking care of their own litter’ (Bäckman 2009: 74). If she decides to exclude it and write about it in connection with her informants’ narratives, she will ‘diminish my informants … and make them more prejudiced than I believe them to be’ (Bäckman 2009: 75). By including her reflections on whether to publish her observations, Bäckman illustrates the balancing act of both wanting to protect a stigmatised category, while also wanting to faithfully report about her informers’ retold experiences and her own observations.
The various difficulties do not end when the material has been analysed and a paper, report, or article is written. When presenting our findings we met audiences and readers, which involved different knowledge-power systems than the one’s we encountered when in contact with gatekeepers or informants. This of course requires reflexivity from the researchers; writing this article is, indeed, an instance of such reflexivity.
As noted previously, qualitative material collected through fieldwork or qualitative interviews involves interactions, but the researcher is also embedded in interactions through relations with colleagues. At seminars and workshops where we have presented and discussed our findings, issues such as what designations we used often came to the fore. Issues of power were also often raised when lack of power is a priori attributed to ethnic minorities, even though situations may at times illustrate the opposite. Such situations may be the sociological analytical point of interest that the authors want to discuss but where the discussion returns to a general statement, often along a simplified psychological interpretation: ‘because of the lack of power, this specific group excersise domination or violence…’. Through the resistance to talk among our colleagues or other audiences about situations that actually show power from those generally described as subordinated, the nuances of the empiric material are disregarded and powerlessness is instead constantly repeated.
One of the points of ethnographic studies is to be open and faithful to the complex organisation of the local context, but questions by participants in seminars or conferences have been raised as to whether the interviewer corrected the interviewees perceived to be giving morally upsetting answers. This occurred mostly in relation to those who seemed to have power over others, such as during seminars on ‘ethnified police practices’ that included reports on what policemen express about various ethnic categories (Görtz 2015). Furthermore, at times, we have gotten stuck on the words we use. In an email from Burcar Alm, from a conference where she had presented a paper on the rhetorical use of ethnicity during interviews with young crime victims, a certain fatigue (because of getting the same comment over and over again) is clear:
It is interesting to see that once again the discussion did not go beyond the problems of the term ethnicity and our sample (Anna, however, avoided it this time since I presented first). It was a good mood and no problem in that way, but it seems to be a problem to highlight ethnicity in this way. You get stuck here … (Email from Burcar Alm to Åkerström, October 2010)
This theme continues in a subsequent project several years later on how ethnic minorities perceive the police. Burcar Alm sent the following email to a project colleague after presenting their findings at a research conference:
What I received as comments was really just one thing – the use of the term “ethnicity” and the importance of talking about racialisation … Strange indeed – suddenly I’m back in the same context as when I, Malin, and Anna studied young crime victims with another ethnic background. As soon as you talk about something that concerns ethnicity, you get stuck in this term and can’t proceed … (Email from Burcar Alm to colleague, May 2017)
The interactions in which the researcher is embedded also include relations with another public – the unknown recipients. We do not control our texts. Words and quotes can be taken out of context and used in ways that we do not wish. Knowing the potential risks of unwanted attention, misunderstandings and uses, we may be particularly careful in what we report if we want to protect our identity as serious researchers.
Contemporary concerns may affect how we choose our study participants and finance our research. Public concerns, however, also can influence the framing of a subject, such as migration being seen as a ‘problem’ by funders who may expect researchers to deliver solutions (Skilbrei 2018). Such framing may also mean that migration researchers find themselves in the spotlight, and their results may be called into question when they, for example, deliver unexpected findings that risk them being cast as debaters rather than social scientists. Such a report may include unemployment among immigrants as decreasing in Sweden, or results showing that immigration bring in revenue, according to researchers interviewed in the Swedish science magazine Curie (Östlund 2018).
Social scientists engage in many research topics that may be sensitive. Some issues may involve issues beforehand that researchers understand are sensitive in the sense that they deal with behaviour that is commonly regarded as discreditable or incriminating, ranging from drug use to paedophilia to corruption (Ranzetti & Lee 1993). However, at times, the delicacy may not be understood by researchers beforehand because the situation is not immediately understood as sensitive; for instance, breastfeeding or not (Murphy 2011) or Swedish employers’ subsidies (Sellerberg 2020). In other words, what is delicate or not has to be interpreted by members in various social categories.
In this article, the focus is on sensibilities during the research process in qualitative studies concerning issues of ethnicity and crime, involving young people who are either crime victims or offenders. Using mainly our own research projects, we have exemplified difficulties in gaining access, collecting material and writing and presenting one’s findings. All qualitative research requires that the researchers gain trust, continually negotiate access and participation, and reflect on how to represent their informants.
We argue, however, that ethnicity and crime-associated issues may be especially delicate because they not only concern those directly involved but also draw critical surveillance by those who read these publications or attend lectures or seminars amidst a societal and political gaze that is so morally and politically charged.
In attempts to gain access, like other researchers, we have had to balance between a rhetoric of equality and acknowledging that young people with an immigrant background may have special experiences that are different from those of ‘insiders’ (c.f. Ogbu & Simons 1998), in our case, young ethnic Swedes. Thus, researchers have to present their requests and engage in attempts to co-construct the issue at hand to avoid appearing naively unaware of differences in experiences.
When collecting data, ethno-national categories or dichotomies come into play that involve an us–them boundary between ‘Swedes’ and ‘immigrants’, regardless of the researcher’s perspective. Carling et al. (2014) have criticised the simplifying ‘outsider–-insider’ divide in migration research, based on their experiences from different studies where in which some of the researchers had an immigrant background and others had an ethnic Norwegian background. Moreover, age, gender and contemporary media attention to special issues may influence the way we position each other. That ethnic categorisations may be actualised was visible in our material when the interviewees asked Burcar Alm and some of our colleagues working on our projects whether they were ‘Swedes or wogs’ during interviews and field observations. Similar comments have been reported in many studies. An example is Aliverti (2018: 89), in her court study in Britain, in which she describes staff making inquiries about ‘where she came from’ and interviewees asking about her origins based on her name and accent, making her visible and different. Tewksbury and Gagné (1997: 138), who have done research on many different stigmatised groups, also noted that ‘it was not uncommon for community members to assume or ask if we shared the statuses that stigmatised members’.
During the process of writing up one’s findings, we may find ourselves torn between different sets of ethics (Eldén 2020). Researchers have to navigate between a professional ethic that requires (more or less) writing about field notes and interviews without any biases, antipathy or sympathy and a ‘care ethic’ requiring that researchers protect their informants. There may not be an easy solution, as described about Bäckman (2009) previously, wrestling with her narratives from Swedes living in marginalised areas who were torn about mentioning littering among new immigrants because they did not want to appear prejudiced. If Bäckman were to have decided to exclude the information, she felt that she would have diminished her informants, would have made ‘them more prejudiced than I believe them to be’ (Bäckman 2009: 75).9
We can relate to how we think the research community and unknown readers will respond to what is written in advance. Of course, we want to do our empirical material justice, but when what is expressed is sensitive, that affects us as researchers, too. However, presenting our material is not simply about reporting what was said but about commenting and analysing. If those we interview express words we do not appreciate, we must still let them speak. As Gubrium and Holstein (1997: 123) write, the interviewees’ explanations are not ours: ‘… interacting persons artfully construct reality by doing things with words, talking reality into being …’. Thus, we do not try to understand ‘the real’ – their worlds, their stories as such – but rather make their practices visible: ‘… how practitioners of everyday life constitute, reproduce, redesign, or specify locally, what the institutional and cultural contexts of their actions make available to them’ (Ibid.,115).
Although research on migration or ethnicities may be sensitive and fraught with difficulties, these sensitivities can harbour analytic opportunities for researchers writing about their findings. Particularly in conversational analytic and ethnomethodological studies, discursive strategies have shown how to avoid potentially negative categorisations based on assumptions about prejudices, stereotypes and sensitive distinctions, among others. Consider, for instance, Whitehead’s (2009) study of a ‘race-training workshop’, where the participants did not know each other beforehand but found it easy to recognise each other in terms of being, for example, Afro-American, Latino or Asian. The empirical material provided illustrations of varied strategies, such as carefully generalising race; localising race by, for instance, neighbourhood; and alluding to race only to deal with ‘the systematic potential for difficulties resulting from referring to race’ (Whitehead 2009: 328).
Ordinary qualitative interviews may also provide rich and fertile ground for analysing delicate expressivity. Such interviews are always sites of self-presentation and impression management (Järvinen & Mik-Meyer 2005), and interviews may be understood as ‘the work of accounting by a member of a category for activities attached to that category’ (Baker 1998: 131). For example, we analysed in detail an interview with a young ethnic Swedish man in our crime victim study, who refrained from mentioning that the people who had attacked him were young immigrants (Åkerström and Burcar Alm 2016). When the interviewer finally brought this up, the crime victim explained that there is no correlation between ethnicity and crime and that he did not even want to tell the police at first, although he eventually did. However, once the subject was initiated, he proceeded to discuss the over-representation of crime among young men with an immigrant background and the danger of a lack of discussion about this in Swedish society. In his long and complicated discussion, he carefully balanced his expressions to avoid coming across as xenophobic and to appear as someone who reflects on and is concerned about societal developments.
Another analytical opportunity is to study how the same words and utterances may have a completely different meaning depending on who is saying them and to whom. The rhetoric of academics and debaters may be different from those they intend to represent. As Swedish researchers, we cannot just designate somebody with an immigrant background who is born in Sweden as a ‘foreigner’ or an ‘immigrant’. Certainly, we cannot use the word ‘blatte’ (wog), which would be met with objections. As a contrast, we find that the young people with immigrant backgrounds whom we studied use these designations not provocatively but in a factual way or even as a symbol of a social bond (see also Andersson 2003; Bäckman 2009; Jonsson 2007).
An oversensitivity can lead to a fear and eventually avoidance of doing studies at all if they involve issues of ethnicity. This outcome would be regrettable because this area is both important and scientifically engaging. Through our discussion in this article, we hope to contribute to raising awareness of the sensitivity associated with studying, for example, ethnicity and crime, while not diminishing interest in doing these types of studies. Sensitivity should primarily be about how we treat those we meet and how we treat our material and not indicate that the research area in itself is sensitive and should be avoided. As Ogbu and Simons (1998) note in connection with education and difficulties for teachers of minority students, the point is not to discuss explicit strategies but to reflect on the nature of the problem.
In sum, the importance of sensitivity recurs throughout the research process when ethnicity and issues related to crime are involved and is related to the identities of the researchers and other relevant parties. Sensitivity not only references using the correct words in the correct context but also encompasses understanding when and why something becomes sensitive. Researchers in this field should be aware, understand and be able to handle the often subtle markers and signals of sensitivity, as well as the caution required with interacting parties, which is an ample subject of study itself.
1https://www.glasnost.se/2001/finns-det-nya-svenskar-har/. Downloaded 2021-01-31.
2https://www.isof.se/sprak/nyord/nyord/aktuellt-nyord-2002-2015/2014-05-05-rasifierad.html. Downloaded 2021-02-01.
3In the case of the Swedish criminologist Jerzy Sarnecki and the American sociologist Katz, such reports involved publishing data. In Sarnecki’s case, it involved discussions about whether to report on criminality among people of certain ethnicities, and in Jack Katz’s case, it involved issues concerning the manuscript of the book Seductions of Crime.
4The Swedish Research Council has published guidelines for many years, and the latest is “Good research practice” (2017). Our commissioned evaluation of the youth project that started in 2006 was not required at that time to pass the Swedish Ethics Board, but it was ethically approved by the authority that commissioned the research, the National Board for Institutional Care. The others were reviewed by the Swedish Ethics Board, dnr. 2011/454, dnr. 2016/1077 and dnr. 2017/1077.
6This was actually the case when we contacted a youth activity center for our project on perception of police among ethnic minorities. The leaders seemed to think that it was important that the young people express their views.
8In a report from the Council of Crime Prevention (2018), Relations to the judicial system, the reader is presented with words describing these areas as having ‘social problems’, ‘criminality’, ‘fear of crime’ and ‘gangs’.
9Representation of the people researched may be may be contested by them. Well-known examples are studies of small societies in which identities, in spite of efforts to anonymise them, have been discovered (Scheper-Hughes 2000; Vidich & Bensman 1968). Authors may themselves note that the reception of their writings is not altogether approved, as in the minor classic Street Corner Society (Whyte 1943/1993: 351 and 352).
The authors have no competing interests to declare.
Åkerström, M. 2013. Känsliggörandet av etnicitet i forskning. [Doing sensitivity in research on ethnicity] In: Peterson, A. et al. (eds.) Den sorterande ordningsmakten [The sorting order in plural policing]. Malmö: Bokbox. pp. 37–68.
Åkerström, M and Burcar, V. 2016. Etnicitet: ett engagerande men delikat samtalsstoff–brottsutsatta unga mäns återberättade erfarenheter [Ethnicity: An involving but delicate subject-young male victims’ retold experiences]. Socialvetenskaplig tidskrift, 22(3–4): 245–263. DOI: https://doi.org/10.3384/SVT.2015.22.3-4.2340
Aliverti, A. 2018. Spotting foreigners inside the court room. In: Fili, A, et al. (eds.) Criminal Justice Research in an Era of Mass Mobility. Abingdon and New York: Routledge. pp. 85–99. DOI: https://doi.org/10.4324/9781315269771-7
Andersson, Å. 2003. Inte samma lika. Identifikationer hos tonårsflickor i en multietnisk stadsdel. [Not the same similar. Identifications among teenage girls in a multiethnic neighborhood]. Stehag: Symposium.
Andersson, M. 2003. Immigrant youth and the dynamics of marginalization. Young, 11(1): 74–89. DOI: https://doi.org/10.1177/1103308803011001077
Barr, B, Lapidus, A and Sörbring, K. 25 February 2017. Sprickan om invandrare och brottsligheten. [The rift on immigrants and crime]. Expressen. [online access at https://www.expressen.se/nyheter/sprickan-om-invandrare-och-brottsligheten/ accessed 6 February 2021].
Bosworth, M, Franko, K and Pickering, S. 2018. Criminal justice research in an era of mass mobility- a brief introduction. In: Fili, A, et al. (eds.) Criminal Justice Research in an Era of Mass Mobility. Abingdon and New York: Routledge. pp. 1–11. DOI: https://doi.org/10.4324/9781315269771-1
Burcar, V. 2012. “Unga mäns brottsofferberättelser–‘svensk’ och ‘invandrare’ som retoriska resurser”. [Young men’s victim narratives-‘swede’ and ‘immigrant’ as rhetorical resources In: Gottzén, L, et al. (eds.) Andra män. Maskulinitet, normskapande och jämställdhet. [Other men. Masculinity, norm-setting and gender equality.] Malmö: Gleerups. pp. 81–97.
Carling, J, Erdal Bivand, M and Rojan Ezzati, E. 2014. Beyond the insider-outsider divide in migration research. Migration Studies, 2(1): 36–54. DOI: https://doi.org/10.1093/migration/mnt022
Clark, C. 1987. Sympathy biography and sympathy margin. American Journal of Sociology, 93(2): 290–321. DOI: https://doi.org/10.1086/228746
Cochrane, B. 2018. Expectations and realities of fieldwork. In: Fili, A, et al. (eds.) Criminal Justice Research in an Era of Mass Mobility. Abingdon and New York: Routledge. pp. 69–84. DOI: https://doi.org/10.4324/9781315269771-6
Council for Crime Prevention. 8 May 2018. Brå överväger studier om brott och migration. [“Brå considers studies on crime and migration.”] https://www.bra.se/om-bra/nytt-fran-bra/arkiv/nyheter/2018-05-08-bra-overvager-studier-om-brott-och-migration.html. (Verification date 2020-11-16).
Elaies, N and Jaconelli, J. 2020. Stigmatiserade på två skilda sätt- en diskursanalys av den mediala framställningen av brott och förorter. [Stigmatized in two different ways- the representation of marginalized areas in the mass media]. Unpublished thesis (BA). Department of criminology, Stockholm University.
Herz, M. 2018. ‘Becoming’ a possible threat: masculinity, culture and questioning among unaccompanied young men in Sweden. Identities. Global Studies in Culture and Power, 26(4): 431–449. DOI: https://doi.org/10.1080/1070289X.2018.1441692
Hewitt, J and Stokes, R. 1975. “Disclaimers”. American Sociological Review, 40(1): 1–11. DOI: https://doi.org/10.2307/2094442
Holstein, J and Jaber Gubrium, J. 1995. The Active Interview. New York and London: Sage. DOI: https://doi.org/10.4135/9781412986120
Integrationsverket. 2006. Upplevelser av etnisk diskriminering hos utrikes födda personer i Sverige 2005. Delrapport 1. [Experiences of ethnic discrimination in Sweden among people born in other countries]. Integrationsverket: Norrköping.
Jacobsson, K and Åkerström, M. 2012. Interviewees with an agenda: Learning from a ‘failed’ interview. Qualitative Research, 13(6): 717–734. DOI: https://doi.org/10.1177/1468794112465631
Kniivilä, K. Finns det nya svenskar här? [Are there any new Swedes here?] Available at https://www.glasnost.se/2001/finns-det-nya-svenskar-har/ [Last accessed at 07-02-2021].
Lindgren, S. 2009. Representing otherness in youth crime discourse. Youth robberies and racism in the Swedish press 1998–2002. Critical Discourse Studies, 6(1): 65–77. DOI: https://doi.org/10.1080/17405900802560116
McElhinny, B. 2001. See no evil, Speak no evil: White police officers’ talk about race and affirmative action. Journal of Linguistic Anthropology, 11(1): 65–78. DOI: https://doi.org/10.1525/jlin.2001.11.1.65
Murphy, E. 2011. Anticipatory accounts. Symbolic Interaction, 27(2): 129–159. DOI: https://doi.org/10.1525/si.2004.27.2.129
Ogbu, J and Simons, H. 1998. Voluntary and involuntary minorities: A cultural-ecological theory of school performance with some implications for education. Anthropology & Education Quarterly, 29(2): 155–188. DOI: https://doi.org/10.1525/aeq.1918.104.22.168
Östlund, H. 22 August 2008. Positiva rön om invandring väcker infekterad debatt. [“Positive findings about immigration generate poisoned debates”]. Curie. [online access at https://www.tidningencurie.se/nyheter/2018/08/22/positiva-ron-om-invandring-vacker-infekterad-debatt/ accessed 6 February 2021].
Peterson, A. 2010. From great Britain to Sweden-The import of reassurance policing, local police offices in metropolitan Stockholm. Journal of Scandinavian Studies in Criminology and Crime Prevention, 11(1): 25–45. DOI: https://doi.org/10.1080/14043850903488367
Ryding, A. 2005. Välviljans variationer. Moraliska gränsdragningar inom brottsofferjourer [The variety of benevolence- moral boundaries in crime victim help centers.]. Lund: Lund Dissertations in Sociology.
Rydgren, J. 2017. Radical right-wing parties in Europe. Journal of Language and Politics. 16(4): 485–496. DOI: https://doi.org/10.1075/jlp.17024.ryd
Sacks, H. 1995. Lectures on Conversation, (Lecture 6). Malden, MA: Blackwell. DOI: https://doi.org/10.1002/9781444328301
Scheper-Hughes, N. 2000. Ire in Ireland. Ethnography, 1(1): 117–140. DOI: https://doi.org/10.1177/14661380022230660
Sellerberg, A-M. 2008. Waiting and rejection: An organizational perspective-‘Cooling out’ rejected applicants. Time & Society, 17(2–3): 349–362. DOI: https://doi.org/10.1177/0961463X08093428
Skilbrei, M-L. 2018. Manoevering in tricky waters. In: Fili, A. et al. Criminal Justice Research in an Era of Mass Mobility. Abingdon and New York: Routledge. pp. 28–40. DOI: https://doi.org/10.4324/9781315269771-3
Tewksbury, R and Gangé, P. 1997. Asumed and presumed identities. Sociological Spectrum. 17(2): 127–155. DOI: https://doi.org/10.1080/02732173.1997.9982157
Walklate, S. 1998. Crime and community, fear and trust. British Journal of Sociology, 49(4): 550–569. DOI: https://doi.org/10.2307/591288
Wästerfors, D. 2014. Playfights as trouble and respite. Journal of Contemporary Ethnography, 45(2): 168–197. DOI: https://doi.org/10.1177/0891241614554087
Wästerfors, D and Burcar Alm, V. 2020. ‘They are harsher to me than to my friend who is blond’–police critique among ethnic minority youth in Sweden. Journal of Youth Studies, 23(2): 170–188. DOI: https://doi.org/10.1080/13676261.2019.1592129
Wästerfors, D and Burcar, V. 2014. Safety work with an ethnic slant. Social Inclusion, 2(3): 88–101. DOI: https://doi.org/10.17645/si.v2i3.27
Whitehead, K. 2009. Categorizing the categorizer. Social Psychology Quarterly, 72(4): 325–342. DOI: https://doi.org/10.1177/019027250907200406
Whyte, WF. 1943/1993. Street Corner Society. Chicago: University of Chicago Press. DOI: https://doi.org/10.7208/chicago/9780226922669.001.0001