Abstract
The last decade has seen a proliferation of sociological research on minorities’ responses to ethnoracial stigmatisation. Compared to social and cultural structures, face-to-face interaction has thus far received insufficient theoretical attention, even though social situations are a prime site for stigmatisation. This has resulted in three shortcomings in the research: (1) a tacit individual–structure dualism; (2) neglect of situational variation; and (3) neglect of dialogical concerns behind people’s responses. Building on 20 in-depth qualitative interviews with Norwegians of ethnic and/or racial minority backgrounds, and homing in on non-confrontations to ambiguous stigmatisation, we investigate the role of face-to-face interaction as such. We rely on Erving Goffman’s theories of interaction order and interaction ritual. The interaction order highlights social interaction’s semi-autonomous existence vis-a-vis both human personalities and social structures, and is sustained by interaction ritual, that is, the mutual protection of selves in interaction. Among our key findings is that non-confrontations are concurrently self- and other-oriented, mirroring Goffman’s concept of facework. Regarding situational variation, we distinguish, for instance, situations of hierarchy from those of assumed equality; interactions with strangers from established relations; situations framed by professional ethics from those not; work from private settings and public spaces.
Introduction
I’m at the point where I have to specify my role before we start any consultation, by saying ‘I’m the doctor, the examining doctor, my name’s Asha.’ I say that very specifically at the outset. Because there’ve been so many times, despite me wearing a doctor’s uniform and coming in and asking all the typical ‘doctor’s questions’, that very often patients will look at me with suspicion and ask: ‘But when will the doctor come?’ ‘No, I’m the doctor.’ ‘Huh?’ they’ll say, and then, ‘oh, okay’, they stop themselves, and ‘okay, please continue’.
Asha is a physician in her early thirties. She was born in Norway to Somali parents. She wears the hijab. What Asha describes with patients also happens with colleagues, when nurses from other wards sometimes mistake her for a health assistant. Although these experiences with patients and colleagues are humiliating, Asha never confronts them. She politely corrects them by stating her role, so she can continue with hospital procedures. These are ambiguous incidents, as shown by Asha’s reluctance to blame them on her racial/religious appearance or on her youthful appearance. Because these incidents have become routine, Asha always begins a consultation by introducing herself as the examining doctor. This protects her and spares patients or colleagues embarrassment. This article analyses how racialised people negotiate ambiguous stigmatisation. We introduce Erving Goffman’s interaction ritual (1967/1982) and interaction order (1983) to the research field.
The last decade has seen a proliferation of studies of how people respond to overt and ambiguous forms of ethnoracial stigmatisation (Ellefsen et al., 2022; Fleming et al., 2012; Lamont et al., 2016). Numerous responses are described, from confronting to ignoring and withdrawing. Asha resorts to a common option: non-confrontation (Dazy, 2023; Imoagene, 2019).
This is a topical enquiry. Stigmatisation threatens fundamental values of liberal democracy: individuals’ right to live their lives, participate in democratic exchange, work and civil society, without being reduced to discredited social categories, and without suffering reprisals from those categories. Stigmatisation can limit people’s life-chances and lead to segregation and reduced societal trust. The study of responses may offer inspiration for democratic discourse, daily actions and anti-racist policies (Piwoni, 2023, p. 98).
For that to happen, enquiries must be accurate. Although social situations are a prime site for stigmatisation, and we are hence dealing with interactional phenomena, face-to-face interaction has received scarce theoretical attention from researchers. This neglect takes three interconnected forms. First is a tacit dualism of individuals and structures. In critiques of psychological interpretations, sociologists highlight how people’s strategies vis-a-vis stigmatisation – whether it is implicit or blatantly racial – are constrained or enabled by social structures and cultural repertoires (Lamont et al., 2016). In contrast, the situational domain is left largely in the shadows. Second, variations across situations are insufficiently incorporated in existing interpretations (but see Bickerstaff, 2012; Piwoni, 2023). Different situations carry different frames (Goffman, 1974), which influence the range of responses available. Third, there is little attention to dialogical concerns in interaction. Researchers depict non-confrontations as self-oriented acts (Ellefsen et al., 2022; Rawls & Duck, 2017). As we will show, those who experience stigmatisation also carefully consider the perspectives of their interactants, including those who overstep personal boundaries. Moreover, they are acutely attentive to how their response may affect the situation.
Goffman’s interaction ritual (1967/1982) and interaction order (1983) respond to these observations. When connected with Stigma (1963/1990), these theories help us understand why people often do not confront transgressors in situations of stigmatisation. They provide heuristic tools to examine the concerns behind non-responses and non-confrontational responses and hone our attention to the observation that situations have force of their own. Goffman’s claim that the interaction order has ‘sui generis’ existence means that, instead of tacitly treating interactional behaviour as ‘dependent variable’ or ‘outcome’ – of personalities and/or structural/cultural forces – we identify the logics of interaction in interaction itself.
We rely on interviews with Norwegians who are ethnic and/or racial minorities (N = 20), most of whom are professionals like Asha. The Norwegian context offers a vantage point for investigation of the topic. Since the late 1960s, Norway has transformed from a predominantly white population to a multiracial one. In 2023, immigrants and their children constituted one-fifth of Norway’s 5.4 million inhabitants. Changes pertained to labour migration from Pakistan and Turkey, and later Central/Eastern Europe, family migration, and refugees from countries like Sri Lanka, Somalia, Bosnia and Afghanistan. Norway is an intriguing case also because the comprehensive welfare state creates opportunities for inter-ethnic interaction, as citizens and employees. Kindergartens are state-financed and affordable, public schools enrol 96% of pupils, and higher education is mostly state-run and tuition-free.
The next section reviews previous research, emphasising conceptions of social situations. After discussing Goffman’s Stigma (1963/1990) and connecting it to interaction order (1983) and ritual (1967/1982), we describe our methods and data. In the findings section, we highlight how those experiencing ambiguous ethnoracial stigmatisation protect themselves, their interactants and situations. We compare non-confrontations with direct and subtle confrontations, identify pre-emptive techniques, and connect the themes through examinations of situational variation.
The study of responses to ethnoracial stigmatisation
To position our contribution, we note that the research reviewed here covers the ‘whole spectrum’ of stigmatisation, from hate speech and racial slurs, via microaggressions and everyday racism, to indeterminable incidents. The topic of our article is ambiguous stigmatisation, which is however compared to instances of overt stigmatisation, which also abound in our data.
Regarding the individual–structural dualism, sociologists criticise psychological research for not considering cultural/structural contexts (Lamont & Mizrachi, 2012). Sociologists argue for putting meaning-making front and centre of analyses (Fleming et al., 2012), and for seeing responses to stigmatisation in relation to ‘broader social factors’ (Lamont & Mizrachi, 2012, p. 366).
The examination of cultural repertoires that shape responses to stigmatisation gained traction with Lamont and co-authors’ research and subsequent book, Getting Respect: Responding to Stigma and Discrimination in the United States, Brazil and Israel (2016). They compare the strategies of African Americans, Afro-Brazilians, Arab Palestinians and Ethiopian and Mizrahim Jews. Among many differences, confrontations are most prevalent among African Americans. Afro-Brazilians, in comparison, emphasise politeness. The researchers’ explanatory frameworks are (1) historical, socioeconomic and institutional elements; (2) cultural repertoires; and (3) levels of groupness (pp. 20–33).
Macro-level correctives to psychological/individualist accounts are important but sidestep the situational domain. This observation holds for many studies. Dazy (2023) shows how French Muslim elites cultivate non-confrontational strategies to stigmatisation for political reasons, strategies which they attempt to inculcate in their co-religionists. Dazy argues that such ‘[m]itigation of anti-Muslim hostility must be understood at two levels: personal and collective’ (p. 690). This mirrors Lamont et al.’s emphasis on broader social factors. In Imoagene’s (2019) study of second-generation Nigerians in Britain, non-response is a dominant strategy. Pointing to the importance of context, Imoagene identifies British colonial history, national identity and repertoires of social class (p. 266), all which shape individuals’ responses to ethnoracial exclusion (p. 264). Although such contexts are crucial, they point to the macro-world, bypassing questions of situational meaning.
Existing research investigates individuals’ responses to stigmatisation and interprets these responses as constrained or enabled by the cultural repertoires available. This is the implicit individual–structure dualism.
The next form of situational neglect is situational variation. This has been addressed in some contributions, which we build on and extend. In a study of German Turks, Witte (2018) identifies four responses to stigmatisation: confronting, deemphasising, avoiding/ignoring and boundary work. Witte constructively distinguishes situational reactions from discursive strategies, which researchers often conflate. Whereas confronting, deemphasising and avoiding/ignoring are situational phenomena, boundary work is not (p. 1429). Witte does not, however, distinguish between types of interactional contexts. Nor is the situational as such theorised.
Some authors attend to situational variation. In a study of first-generation French Black people, Bickerstaff (2012) usefully distinguishes between responses according to relationships (impersonal vs personal) and social contexts (school, work, public spaces). Piwoni (2023) criticises the literature for not differentiating ‘between responses to incidents of racism in a concrete situation as it happened and responses to incidents outside the concrete situation’ (p. 108). We concur. Importantly, both Bickerstaff and Piwoni stress the contrasts between strategies available when people experience stigmatisation in their professional life and when they do so in personal life. Piwoni warns that if we do not address such situational variation properly, we risk comparing apples to oranges (p. 115).
For all the value of Piwoni’s and Bickerstaff’s contributions, they attend less to why and how face-to-face interaction matters. A theoretical understanding of interactional concerns, like balancing social situations, and how this differs from retrospective backstage settings, is missing. This is the third neglect of situations/interactions: dialogical concerns. Existing research interprets non-responses/non-confrontations as self-oriented acts: people ignore stigmatisation to maintain their ‘personal dignity’ (Rawls & Duck, 2017) and ‘self-worth’ (Lamont et al., 2016, p. 248). Non-response is described as an ‘individual and informal way of coping’ (Ellefsen et al., 2022, p. 443; Imoagene, 2019, p. 275). In contrast, we show how non-confrontations like deemphasising and smoothing over, as well as subtle confrontations like humorous rejoinders, are concurrently self- and other-oriented.
Stigma, interaction order, ritual
Whenever stigmatisation is under scrutiny, Goffman’s Stigma (1963/1990) is a standard reference (Lamont et al., 2016). Goffman’s interest was the encounters between ‘the stigmatised’ and ‘the normal’, with special attention to the stigmatised’s interactional navigation. Defining stigma as ‘an attribute that is deeply discrediting’ (1963/1990, p. 13), Goffman emphasised a relational understanding; that we direct our attention towards relationships and interactions. Stigma is not a thing in itself, but rather determined by the relationship between a label linking a person to a set of characteristics that is devalued in particular situations (1963/1990).
Goffman’s treatise on stigma, despite its enormous influence, has also been criticised. Goffman has been accused of being silent on issues of ethnic/racial justice, like the racial political upheavals that erupted when he was writing Stigma (Tyler, 2018). Through his silence, Tyler argues, Goffman contributed to what Bhambra calls ‘epistemological injustice’ (2021) by formulating a sociology ‘for normals’ (Tyler, 2018, pp. 754–755). An adjacent critique is that Goffman ignored where stigma originates, and how stigma is produced through political and structural forces (Link & Phelan, 2001; Tyler & Slater, 2018). In aiming to reconceptualise (and decolonise) our understanding of stigma, Tyler (2018, 2020) sees stigma as perpetuated through power structures of colonialism, welfare retreatment, neoliberalism, etc. Stigma can be technology in the hands of governments (Tyler, 2020).
These are important correctives. Notwithstanding, the sociology of stigma should caution against throwing the ‘interactionist baby’ out with the bath water. For where stigma is produced –and where power relations like ‘white ignorance’ (Mills, 2007) stem from to influence interactions – we must attend carefully to structural and historical processes of inequality. Here, microsociology is insufficient. In contrast, for how stigma is negotiated in everyday life, through interactional processes, structural/political/historical accounts only provide so much analytical assistance. These are sociological questions that require different data and research. We may hence draw an analytical distinction between structural/political production of stigma and interactional face-to-face management of it. Our project is the latter. Here, Goffman’s theories hold in valuable insights of continued relevance. For influential studies of racism that make productive use of Goffman’s theories and concepts (without losing sight of historical/structural processes) see e.g., Rawls and Duck’s Tacit Racism (2020).
Whereas Tyler and others might be read as calling for a ‘vertical’ (and historical) extension of Goffman (to include power, structures and politics), we suggest a complementary ‘horizontal’ extension, by aligning Stigma with Goffman’s other work, most notably interaction order and ritual. This theoretical intervention improves our understanding of why, when and how people – in face-to-face encounters – respond (or not) to ethnoracial stigmatisation.
Goffman relied on two concepts from Durkheim: sui generis and ritual. He transplanted Durkheim’s sui generis idea of society’s existence in itself to face-to-face interaction, which he argued has semi-autonomous existence vis-a-vis both individuals and social/cultural structures (1964, 1983). In Rawls’ interpretation (1987), Goffman’s interaction order, as order, connects intimately to the second Durkheimian idea, ritual. Ritual emerges from people’s longing for acceptance and their constant search for respect (a recurrent theme in Goffman’s works, Stigma not the least; Smith, 2006). Seeing human selves as fundamentally fragile, Goffman theorised the human personality as sacred, a totem that is the object of worship in modernity (as collective gods retreat). The individual is a frail little ‘god’ striving for ‘worship’ (respect/self-affirmation). Its predicament is that it must itself acquire worship, which it can only obtain from other people, who are equally fragile little gods. Enter interaction ritual and facework (1967/1982), the collaborative communication that upholds our projected self-definitions (‘face’) upon which we rely. Interaction is fragile (and sacred) too: since we rely on our interactants to ratify our self-definitions, and others rely on us, we go to lengths to protect interaction. We are here governed by two interactional ground rules: ‘the rule of self-respect’ and ‘the rule of considerateness’ (1967/1982, p. 11). Interaction ritual around totemic selfhoods that need perpetual bolstering adds up to the interaction order sui generis.
Whilst Goffman made no reference to ritual in Stigma (1963/1990), the ideas can be connected (Smith, 2006; Vassenden & Rusnes, 2022). An ethnoracial stigmatisation – like Asha being mistaken for an aide – can be conceptualised as a face-threat, a violation of her ‘sanctity’ as a person. It also disturbs the interaction order. From the rules of self-respect and considerateness arise considerations like ‘I must respond here, for my own self-respect’ vs ‘I may appear inconsiderate if I confront this person, if she is unaware of her intrusion’. The humiliated little god, whose ‘worship’ is undermined, finds itself in an extra predicament: this person, like any other, is reliant on the interaction order. Because ‘[o]ne’s own face and the face of others are constructs of the same order’ (Goffman, 1967/1982, p. 6), we are inclined to avoid threatening topics, those that may violate the face of the other. Even when our own face has been threatened, and we are humiliated, the threshold for making a ‘return face-threat’ is high.
Far be it from us to suggest ‘universal laws’ of (non)responses to ethnoracial stigmatisation. Our point is more straightforward: the theoretical observations of mutual dependence for upholding of face/selfhood, and people’s attempts to protect the interaction order advance our understanding of the issue. The stigmatisation itself – the ‘input’ – also matters greatly in creating conditions for collaborative ritual communication to uphold selves in interaction. Most empirical cases we present are about ambiguous ethnoracial stigmatisation. Often, it is precisely ambiguity that makes people avoid confrontation. Saving face and protecting the interaction order may be less relevant in situations of overt racism and clear-cut stigmatisation. Conceptual clarifications of ‘ambiguous ethnoracial stigmatisation’ and how it connects with ‘everyday racism’ are appropriate. In pointing to subtle exclusion, these are clearly overlapping concepts, and we will use them interchangeably. Essed (1991) defines everyday racism as manifestations of broader racism occurring through repeated interactions. Through accumulated experiences, minorities acquire racial awareness but not necessarily certainty about incidents (Bourabain & Verhaeghe, 2021; Essed, 1991). We conceive of ambiguous stigmatisation as a continuum from incidents of everyday racism that are recognisably ethnoracial, to incidents where the recipient is hesitant to interpret the incident as ethnoracial, despite experiencing devaluation. Ambiguity can also pertain to whether interactants or audiences will recognise the incident as such.
Before moving to methodology, we have another clarification regarding Goffman’s interaction order in the study of racism. In the US, Rawls and Duck (2020) have extended the theory by proposing the idea of separate Black and White interaction orders, where the former connects to oppression and ‘double consciousness’ (Du Bois, 1903/2008). Others apply the interaction order more as a micro-expression of larger racial structures (Meghij, 2019; Rosino, 2017). We find the latter application unsatisfactory. Whilst it may provide insight into how larger structures shape our daily interactions, conceiving of interaction orders as reflections of structures, be they of social class, ethnicity/race, capitalism or the welfare state, hinders us from applying Goffman’s ideas. We do not argue, and neither did Goffman, that the interaction order exists in isolation from structures, ideologies or history. Goffman stated that between the interaction order and social structure, there is ‘loose coupling’ (1983), which, however loose, is still a coupling. Goffman’s sui generis point was never that the extra-interactional is irrelevant, but that the interaction order’s analytical autonomy helps us understand the force of interaction in its own right.
Methodology
Between May 2021 and January 2023, we conducted 20 in-depth interviews with 13 women and 7 men. Most are from immigrant families, with Somali (8), other African, Asian or Latin American backgrounds. Three interviewees were transnational adoptees to white Norwegian parents. Sixteen participants had all or most of their upbringing in Norway. The rest came as young adults. All have Norwegian citizenship. All but two are ‘visible minorities’ through phenotypes and/or wearing hijabs. The two are perhaps light-skinned enough to ‘pass’ as white Norwegians, but have ‘non-Norwegian’ names, and experience vicarious stigmatisation through their significant others (e.g. a sister who wears the hijab).
Interviewees were recruited mostly through our extended networks. We alternated the interviewing; Handulle interviewed those recruited by Vassenden, and vice versa.
Respondents ranged in age from 25 to 60 with an average age of 38 years (median 37). Half of the participants have children. All but one had higher education, with health and welfare occupations being the biggest groups (n = 10): doctors, psychologists, nurses, social workers. Other occupations included university academics, engineers and entrepreneurs. They lived in different parts of southern Norway: Oslo, Stavanger and smaller cities.
Due to the Covid-19 pandemic, we started interviewing over Zoom (13 interviews), before transitioning to face-to-face interviews. Other researchers have noted that remote online interviewing need not compromise data quality (Zadkowska et al., 2022), which was our experience too.
Interviews lasted 90 minutes on average, and were conducted by either Handulle, who is an ethnic minority woman (7 interviews), or Vassenden, an ethnic majority man (13 interviews). As interviews are interaction rituals in their own right (Vassenden & Mangset, 2024), our positionalities merit reflection. Although neither appeared to dampen narratives, we often asked towards the end of interviews if participants thought it would matter had the interviewer been of majority or minority background, respectively. Interviewees typically answered in the negative, but some also told Vassenden that, in retrospect, they had perhaps ‘sugarcoated’ their stories a bit. We also note here, that among those interviewed by Handulle, proportionately more reported having often confronted stigmatising incidents. That might be entirely coincidental. Still, with situational interviewing (Vassenden & Mangset, 2024), perhaps a minority–minority interview more closely resembles the backstage reworking of incidents. Here, one might have in mind how one wants to respond to stigmatisation (Piwoni, 2023, p. 111). A majority–minority interview may more closely resemble the actual cross-racial encounters.
Social desirability bias can work in contradictory ways. In majority–minority interviews, respondents might take care not to alienate the interviewer, and to exaggerate the frequency of their non-confrontations. A minority–minority interview might pressure some respondents to display knowledge of identity issues and exaggerate their confrontations. This, of course, we cannot decide. We restrict ourselves to believing that the combination of our positionalities represented a strength. This extends beyond who did the interviews. Orupabo, who did not participate in the interviewing, is also a visibly minority woman.
Readings and analyses of transcripts were conducted by the entire team, with careful attention to signs of social desirability bias and face-saving. We tried to play to the strengths of being an ethnoracially mixed group of researchers when planning, recruiting to and conducting interviews, and in the analyses. We cross-validated all readings of transcripts, and strived to understand the actors’ points of view of all interactants in the incidents described – interviewees and their interactants (minority and majority alike) – by continuously asking ourselves ‘what is going on here?’
As per regulations from the Norwegian Agency for Shared Services in Education and Research, who approved the study, participants received written information in advance, for informed consent. This stated that the interview’s topics were situations of ambiguous stigmatisation and ‘everyday racism’. Beyond that, we left it to respondents to narrate, with the interviews set up to be ‘less than semi-structured’. Participants generally needed little prompting, with interviewers keeping track with questions like ‘what did you do afterwards?’ or ‘do I understand you correctly, that. . .?’ In the latter half of each interview, we introduced two to three vignettes and asked the interviewees to put themselves in the situation and think how they would have acted, or how the people in the vignette would react. One vignette used in all interviews was: ‘A professor from an Indian background, working at a Norwegian university, is mistaken for a cleaner by a new colleague.’ Overall, the vignettes successfully elicited similar experiences to the interviewees’ – many had already mentioned similar experiences – and allowed us to make finer distinctions of situations, like by toying with audiences.
Interviews were transcribed verbatim, and the data analyses combined abductive (Tavory & Timmermans, 2014; Vassenden, 2018) and thematic analysis (Braun & Clarke, 2006). Working independently, the three of us familiarised ourselves with all data through close readings of transcripts, writing memos and notes. Next, we sat down face-to-face to develop categories/themes for more hands-on coding. Categories included social arenas/contexts; types of stigmatisation; typical response; reasons for response; relational consequences; and backstage reworking.
For how findings emerged from data, we mention situational variation. This showed in two main ways. First was in reported responses, where confrontations are more common in some situations than others. The second, and equally important, was in the explanations that interviewees gave for their responses, like why one remained silent. These explanations contrasted distinctly across situations.
Findings
Here, we first describe the main patterns and the broader concerns that participants emphasise in their non-confrontations. For contrast, we next examine confrontations, as they are also present in our data. Subtle confrontations bridge to how facework is creatively ‘stretched’, most notably through humour. Thereafter, we present pre-emptive measures that participants take in anticipation of stigmatisation. While situational variation is topical throughout, the final part connects the analyses regarding contrasting situational pressures.
The main pattern: Dialogical non-confrontation
Alfred, 45, moved to Norway from Rwanda in his late teens. He is a professor of chemistry in Oslo. When shown our professor vignette, he shared his own experience: The wife of a new colleague visited the department. She met Alfred in the hallway and asked if he knew where her husband’s office was. Alfred showed her the way, and shortly after, overheard the woman telling her husband that ‘the cleaner’ had helped her. Alfred never mentioned this episode to the colleague or anyone else until the interview. His reasons for not confronting in this and similar incidents pertained to both himself and interactants: he did not want to make a person feel bad for making a mistake. Next is solidarity between colleagues and the risk of ruining social relations. In addition, one might appear as a victim, or get a reputation as a troublemaker.
When interviewees report about such incidents, the main reported response is non-confrontational facework like Alfred’s: ignore, smooth over, politely correct mistakes about identities, let humiliating comments pass, or simply remain silent. Participants generally say that for them to call someone out, there cannot be doubts about the incident: the other person’s intentions must be malicious, or the message unmistakably racist.
Solveig is a social worker in her mid-thirties, who was adopted from an African country to ethnic Norwegian parents. She has told older relatives not to use the word ‘neger’ (negro). Although it did not quite sink in, Solveig dispassionately explained to them why this is a hurtful racial epithet. (In Norway, it is now often referred to as ‘the N word’, 1 a phrase used by several interviewees.) At work, she generally lets insults pass if she considers them unintentional; but she has also engaged in conversations over lunch if colleagues have used ‘the N word’ – explaining to them why that word is hurtful. Regarding mistaken identities, Solveig shared an incident from a hiking trip. When getting the key for her rented cabin, the desk officer tried to compliment Solveig, saying ‘Oh, that’s such a nice Norwegian name!’, implying that she did not look Norwegian.
I [showed] clearly, with facial expressions and everything, and being short . . . I saw that she was dismayed, and I felt such remorse, that I was so. . . I didn’t say anything, but everything was like. . . I just took the key, and I left. I said nothing and then I saw that she felt bad too. [. . .] Because I was so short. . . normally, I appear quite happy and outgoing, and I remember so well that, in that setting, I didn’t have the energy to, sort of ‘Yes! I’m Solveig, it’s like this and that’ [. . .] when I see that the intention hasn’t been bad, more like uninformed, clumsy, then I think it’s difficult to be. . . Well, should I have been angry? She didn’t mean it. But I became a bit sad, and a bit pissed [off]. I find that difficult. And she acted nice, you know? ‘Such a wonderful Norwegian name!’ Or ‘your Norwegian is so good!’
Solveig told us about numerous other experiences, like the comment on being fluent in her mother tongue. She does not confront interlocutors when this happens. When shown our vignette with the professor mistaken for a cleaner, she referred to a previous job, where she had worked with refugees in Oslo. A colleague whom she did not know once tried to stop her from entering the office building, confusing Solveig with a refugee client. Solveig said nothing and stepped inside. She could tell that the colleague ‘wanted to disappear’. Solveig mentioned how people can make these blunders in a split second. ‘From that perspective, it’s almost like I feel sorry for the person who messed up.’ Still, these cumulative incidents can be humiliating and emotionally draining.
Solveig said nothing in the lodge and office-building incidents. In the former, she could not help expressing her displeasure through facial expressions, and by leaving without a word. Both are certainly responses. Apparently instinctively, Solveig communicates her emotional reaction and that she indeed noticed what happened, but she stops her response from becoming antagonistic. In other words, facework can be a fine balancing act.
Feeling sorry for the other person, as Solveig did, is only one way of playing by ‘the rule of considerateness’ in facework. Still, even if one does not feel sorry for the interactant, and even if one wants to react with fury, one is mindful of how one appears. Non-confrontation is also about protecting one’s own face, as considerate. Interviewees were more likely to mention this than emotional connection with transgressors. Asha explained her own non-confrontations vis-a-vis colleagues who expressed surprise that she is the physician, by asking: ‘will others see me as a difficult person? As someone who is easily offended?’ Here, non-confrontation, and being considerate, is about keeping others from seeing one as inconsiderate, a troublemaker, or as unsafe to interact with. This mirrors Goffman’s observation that ‘[o]ne’s own face and the face of others are constructs of the same order’ (1967/1982, p. 6).
The double nature of face also showed in Fadumo’s story. Fadumo is of Somali background and works for a business consultancy in Oslo. Her accounts of stigmatisation included racial jokes at work; for instance, when playing chess during a coffee break, her opponent said: ‘No, Fadumo, you’re black, you must be black in chess too!’ Or when a power failure made the lights go out, one colleague asked, ‘where did Fadumo go?’ Although she has on occasion put some of these colleagues straight, she (‘unfortunately’) lets most of the comments slide. ‘Of course, you can see from my facial expressions that this isn’t funny at all’, Fadumo said, but like Solveig, she mostly stops herself from raising her voice. She gave several explanations: she sometimes thinks that she does not have to make people look even more stupid than they already do. Another point is how reliant she is on others at work, especially her superiors:
Quarrelling with colleagues is not easy, because. . . You can’t escape them, you’re eight hours in the workplace with them. And many of them are my superiors, so if I need help. . . say, I actively seek them out for help, and then [if I had called them out]. . . It’s difficult to bite the hand that feeds you.
Fadumo is also afraid to come across as thin-skinned or to get a reputation in the workplace as a complainer. Moreover, confrontations, no matter how minor, may generate countermoves. Fadumo knows that if she criticises someone for stereotyping Somalis, she might not be taken seriously: ‘“I wasn’t talking about you. You’re okay!”’ She also suspects, that if she does speak up, the company might stop hiring minorities. So, Fadumo does not confront her prejudiced colleagues. That is not say that their jokes do not offend her. They do, and Fadumo is applying for positions elsewhere.
Confrontations: Direct or subtle
Non-confrontations are widespread. For comparison, we will now cite confrontations.
Given how incidents of disrespect are cumulative, some interviewees reported having ‘snapped’ after, say the 100th incident. Thomas, an academic with a Brazilian background, ‘always call[s] out the other person, and that is for political reasons’. Ayanle, an economist with Somali parents, says that he tends to ‘answer first, and think later’. He adds that his answers are so pointed that the other person cannot respond in kind. Ayanle tries to dominate these situations, which is facework in its own right, but one when it becomes ‘less a scene of considerateness than an arena in which a contest or match is held’ (Goffman, 1967/1982, p. 24). Thomas and Ayanle, who claim to ‘always’ confront, are outliers, and seemed to have more overt ethnoracial disrespect in mind than Asha, Alfred or Solveig did.
Another reported confrontation involves significant others. Illuminating accounts described interviewees accompanying their mothers who are not proficient in the Norwegian language, on shopping trips or to cafes. Several who say they typically do not speak up for themselves, at work or elsewhere, do speak up for their mothers when they receive poor service or are treated disrespectfully. Here, facework is less about protecting one’s own face, as that of one’s mother, who may be less able to do so. These confrontations are played as subtle but unmistakable exposure of prejudice. Esra is a social worker with Turkish parents. Being light-skinned and not wearing a hijab, she might pass as ethnic Norwegian, or perhaps Italian. Her mother wears the hijab and long dresses. She also does not speak any Norwegian. When Esra believes that a waiter or shop assistant has treated her mother poorly, she will talk to her mother in Norwegian about the bad service, as a way of letting the offending party know that she is aware of their behaviour.
I say [to mum] that the other person was rude not to greet us. . . Even though mum doesn’t understand any of it! But it’s kind of. . . to demonstrate in a way, that ‘this is not OK!’ to the person behind the desk, or the waiter. [. . .] This has happened repeatedly. Sometimes, of course, I don’t [bother]. . . But it’s like when we enter a shop, and there is no one else there, and we come in, and don’t even get a ‘hello’. Then, someone else comes behind us and suddenly ‘hi, do you need any help?’ That’s when I sort of respond like this, telling my mum in Norwegian: ‘My God, how rude! We didn’t get a hello, no one offered us help, but they ask the ones who come behind us, and we were the only ones in the shop to begin with.’
Esra’s story speaks to several of our findings. First, it illustrates situational variation – Esra is among those who were most insistent that they never confront anyone at work, but in situations with her mother she does give a corrective response. Next, her confrontation/demonstration exemplifies creative stretching of facework. Importantly, however, Esra navigates carefully within the frame of the interactional context. She stays within the language of service, without explicitly mentioning prejudice. Within that frame, and via the rules and resources of that interactional context, she communicates clearly to shop assistants that she is aware of their prejudice.
Return digs within confines of ‘rule of considerateness’
In the tenor of interactionism, people creatively construct meaning in their lives. In protecting situations and personal integrity (one’s own and that of others), people toy with expectations and challenge rules. Here, jokes are subtle confrontations carefully crafted not to violate ‘the rule of considerateness’. Humour can sustain facework and soften stigmatisation. In a study of Hungarian Roma, Dobai and Hopkins show that one function of humour is to smooth interaction and avoid or ease embarrassment for both the stigmatised and potential stigmatiser. Roma persons can tell ‘Gypsy jokes’ to help majority members feel more comfortable in interaction (2020, pp. 453–454).
Hassan intriguingly illustrates the use of humour. He grew up in Norway with Somali parents and runs a small company. When he experiences unintentional stigmatisation, like being praised for speaking Norwegian so well, Hassan resorts to a polite but cheeky ‘Thanks! Your Norwegian is also very good!’ When asked if he enjoys living in Norway, he replies with a smile, ‘Yes, I do! Do you?’
Hassan’s rejoinders serve several functions. First, in accordance with ‘the rule of self-respect’, Hassan presents himself as self-ironic, witty and a native speaker, whilst exposing the other person’s ignorance. However, when Hassan does so through friendly teasing, his jokes importantly invite the other person to laugh at their mistake, and perhaps even results in some intimacy. It is a dig and an invitation in the same go.
Pre-emptive techniques
Participants often try to pre-empt stigmatisation through providing information at the start of encounters with people they do not know. Early provision of information was evident in several interviews. Whether one would confront or not if a stigmatisation were to happen, this allows everyone to move forward without disruption and embarrassment. To oneself, it prevents being humiliated and/or becoming upset.
Although feeling stigmatised, Kaja (social worker; adopted from South Korea) responds with neither confrontation nor humorous corrections when people ask her where she is ‘really from’. Moreover, she often volunteers information about her origins. Kaja may tell new acquaintances very early that she was adopted. Her reasons for doing so show intriguing facework. It prevents people she likes from showing their true colours by asking the dreaded question ‘where are you [really] from?’ By saying so herself, she avoids having to change her opinion of them. At the same time, she does not have to ‘go around wondering what other people really think of me’. This is about protecting her image of others as likeable. At work, it is a measure of self-protection, to keep her colleagues from placing her in the same category as their Asian clients.
Tone, who was adopted from Vietnam to ethnic Norwegian parents, is a senior psychologist in Oslo. Like Asha, she looks younger than her years. At work, she asserts her senior position by wearing business attire.
When I enter a room, it’s very seldom that they just assume I’m the senior psychologist. [. . .] I go there [work] wearing, for example, a blazer or high heels. I wear clothes in a somewhat formal way to [chuckles]. . .
. . . demonstrate?
[concurrently:]. . . demonstrate who I am. And what I’ve always heard, is that I look very young, and that I look like a student. So yeah. . . I become more formal in my way of dressing.
Hodan (Somali parents) has often been trailed by security guards when shopping, especially when she was younger. To avoid being followed and to put the security guards and store staff at ease, Hodan approaches them first.
I talk to them. So that they leave me alone. . . I say ‘hello, hey, how much is this sweater?’ Then I make eye contact and we have a conversation going and maybe she can relax, and I can be allowed to look around in the shop.
By initiating contact with salespersons or security guards, Hodan allays their suspicions. If she did nothing, and the surveillance became more obvious, she might leave the shop without making a purchase or react angrily, thus violating her own face by appearing aggressive, and that of the employee, who appears prejudiced. By initiating contact, and de-escalating the situation, Hodan protects the interaction order.
Situational variation
The previous cases indicate situational variation, which we now examine more closely. In our data, important interactional contexts are: (1) work-life; (2) private life – friendships and family; (3) civil society; (4) schools/kindergartens; (5) public spaces; (6) shops and cafes. Across contexts, there is an important contrast between situations where interactants are situational equals, as colleagues, or as parents in children’s sports; and situations of hierarchy, like with superiors at work. This dimension crosses with whether interactants have established relations and are thus mutually dependent, like colleagues and relatives, or will likely never cross paths again, like strangers in public spaces. These situations pressure for different forms of interaction ritual, from which follows our point (with Goffman) that rules, frames, and hence availability of responses also belong to the situations and to the interaction order.
Let us revisit Solveig’s stories and consider her alternatives. In the hiking incident (‘nice Norwegian name’), Solveig would likely never see the receptionist again. In that sense, she is ‘free’ to respond critically. Yet, as this is a fleeting encounter, she is also free to leave, which she did. The stigmatising ‘input’ may not compare to when her older relatives or colleagues use derogatory terms (‘neger’). Nevertheless, with relatives but also colleagues, her responses are shaped by the need to sustain their relationship. Potentially, that could have led to continued silence on her part. But what she opts for is non-confrontational but corrective dialogue, which (hopefully) stops her relatives/colleagues from using the term again. Hence, Solveig probably tries to prevent future stigmatisation from people she must interact with again and/or who are close to her. As she said, ‘the closer the relation, the worse it is, more painful in a way’. This will motivate her to repair the problem in a way that she does not with strangers. Moreover, simply leaving the situation, like Solveig did in the cabin lodge incident, is not as easy in an office lunch or a family dinner. That, we surmise, might alienate interactants.
Recall next the jokes directed at Fadumo at work (power failure; chess). The backdrop to her non-confrontations differs from Solveig’s. As Fadumo explained, the incidents took place within local structures of power, as she mentioned in the excerpt about needing help from senior colleagues. Fadumo’s workplace is male-dominated, Fadumo is the only person of colour, and among the youngest.
As determinants of responses, power and hierarchy were evident in several interviews. Hodan has Somali parents, is a graduate in human geography, and works for a private company in Stavanger, on temporary engagements. Whereas Fadumo talked about colleagues and more informal hierarchies (of ethnicity/race, gender, age), Hodan referred to her boss, i.e. formal power and hierarchy. Her boss, she said, frequently commits microaggressions (similar jokes as with Fadumo). Hodan tends to ignore them, mostly to secure future engagements. She blamed herself for ‘making excuses for him’, instead of calling him out.
Especially illuminating for contrasting situational demands are work settings where one is in the superior position and bound by professional ethics. This is the domain of doctors, nurses, psychologists and social workers, who engage with patients and clients/users. Here, confrontation is not an option. The doctor/nurse/social worker is the one in power, as a gatekeeper of treatment or welfare. Professional ethics give doctors, for example, explicit responsibility for the protection of patients who are vulnerable, and have legal rights. When engaging with patients, Asha’s response to humiliation is determined by these requirements. Connecting the different situations in which she bites her tongue, Asha may do so because of her authority over patients and as an equal to her colleagues. Yet, her non-confrontations arise from very different (situational) demands. Some of Asha’s stories about patients and colleagues were markedly dissimilar. When a patient refused outright to be treated by a black doctor wearing a hijab, Asha did not respond with anger (despite feeling angered). Instead, she politely but firmly informed him that she was the only doctor available to him, and the alternative was no treatment at all (for which she has management support). If patients are visibly unhappy that Asha is the doctor but do not say anything, she carries out the examination like with any other patient, despite her discomfort. Compare this to a story from when she was a resident: Asha had some difficulty inserting a syringe and a nurse questioned Asha’s capabilities in front of the patient. Asha attributed the disrespect to her ethnoracial appearance. Confronting the nurse in the moment was out of the question. After the appointment, Asha asked the nurse to accompany her to another room afterwards and gave her a full dressing down. With the patient, in other Goffmanian terms (1959/1987), this was impression management on the hospital’s frontstage, where all staff members play their roles in support of collective performance of good treatment. Backstage, Asha set the record straight with the nurse for the sake of her own self-respect, clearly overruling considerateness.
Across contexts outside of work too, non-confrontation dominates, yet with more individual variation, and with interviewees’ explanations being starkly different from what they are in the workplace. If teachers or staff make an inappropriate comment to a parent who is dropping off children, that parent will likely not respond critically, but mostly out of consideration for the children – as mentioned by Alfred (in contrast to his reasons for not confronting departmental colleagues). The status equality of friendships or parenthood in children’s sports also makes responses deviate from those at work, by becoming more critical of or more sympathetic towards the majority person. Some who do not respond at work, for the sake of collegiality, ethics or power, report doing so in private contexts. Suad, a social worker with Eritrean parents, regularly bites her tongue at work – albeit (like Asha) more with clients than colleagues. In contrast, when another parent at her son’s handball team commented on her fluency in Norwegian, she retorted with, ‘Well, what did you expect!?’ At work, she never responds to clients’ comments on her background.
Situational variation is key to understanding where, why and how minorities respond to stigmatisation. Beneath this lies Goffman’s basic structures of interaction order and ritual.
Discussion and conclusion
Regarding interactional dimensions, we suggest distinguishing situations of hierarchy from those of equality; interactions with strangers from established relations; situations where one is the victim of stigmatisation from those when significant others are; workplaces from private settings and public spaces; and the importance of maintaining professional ethics at work. Another dimension that we have not explored much in this article is the social audience. Confrontations might be more expected from someone who experiences a stigmatising incident if bystanders are also minorities. ‘The rule of self-respect’ now applies to a different game from a dyadic majority–minority incident. One concern may be that friends do not see one as allowing oneself to be insulted. Audiences also open interactional possibilities, like in talking indirectly to transgressors via a third person, like with Esra and her mother.
Understanding such variation is crucial, and our study is only a beginning. It continues Piwoni’s (2023) point about comparing apples to oranges lest we fail to distinguish properly between situations. The difference between, for instance, medical examinations, where physicians are regulated by professional ethics (Asha), and interactions on the sideline of children’s sports (Suad) are illuminating. With a stronger contrast, we must separate certain strictly hierarchical situations of power, even potential violence, like with the police or in border security-checks, from disrespect in situations where people are interactional equals, like with strangers in public spaces. Confrontation may feel necessary for self-respect in the latter situation but can cause bodily harm or get a person imprisoned in the former. In Lamont et al. (2016), datapoints include Arab Palestinians being silent in security checkpoints (p. 2), and African Americans confronting whites who cut in line in shops (p. 87). Yet, in explaining why confrontations are generally more prevalent among the latter than the former group, the authors cite the macro-contexts of culture and history rather than situational requirements, which in these two examples are polar opposites. Whilst Lamont et al.’s analyses are nuanced and productive, this fundamental analytical level is largely missing.
In this regard, we highlight the promise of engaging Goffman’s theories ‘in total’, beyond single pieces like Stigma (1963/1990). This pertains to seeing face-to-face interaction as a partly separate domain of the social world, induced by ritual protection of totemic selfhood. Society’s almost infinite situational variations rest on top of the basic ritual interactional structure of self-respect and considerateness.
With Goffman’s interaction order sui generis (1983), we treat accessibility of responses as belonging also to – even being a feature of – the interaction order. This opens new avenues for theory and research. It can refine our understanding of how larger repertoires influence responses, with repertoires sifted through situations with semi-autonomous existence and their own force. This addresses the individual–structure dichotomy in previous research, and the need for more accurate analyses. But it also speaks to the necessity of including social/cultural structures (Tyler, 2020). If any readers have taken our arguments to suggest that macro-contexts do not matter, we must correct those impressions. Goffman’s ‘sui generis’ was arguably intended as a heuristic device as much as an ontological argument. This is how we have intended it. The broader repertoires documented by Lamont et al. (2016) and others are real, and consequential. (So are personalities.) Nevertheless, we must carefully consider how the interaction order in its manifold manifestations mediates repertoires.
We emphasise interactions but have examined them only from one part. We should also study majorities; and in ways that live up to the classical interactionist premise of taking the actor’s point of view. Which ritual concerns arise when someone realises a blunder? Asha’s patients clearly try to backtrack. They too protect the interaction order, by correcting themselves and quickly deferring to Asha’s status (‘please, continue’). We need also to understand how majorities experience confrontations, as when Suad replies ‘well, what did you expect?’ to people complimenting her fluency in Norwegian. One hypothesis about failed interactions is that someone like Suad ‘snaps’ in, say, the 100th instance of arduous face-protecting interactional labour. To the interactant, this can be a first-time experience. Suad’s reaction may seem unjustified. Whereas Suad likely sees the other person as representing the 99 preceding interactants, the majority person may object to being placed in that category. Likewise, we need to know the ‘effect’ of Hassan’s humorous rejoinders (‘Your Norwegian is good too!’). Do they, as we surmise, build sympathy?
When face-protecting interactional labour is hidden, it remains invisible for the white majority and conceals humiliation from transgressors. Thus, interaction rituals also highlight the asymmetries in incidents of ambiguous stigmatisation. Whilst one party may barely notice and remain unaffected, the other party is restricted. For many informants, enacting non-confrontation to stigmatisation is emotionally draining. Protecting the interaction order even when humiliated, can leave people feeling exhaustion, anger and shame. Future research should examine how such experiences feed into group formation, and segregate life-worlds.
Footnotes
Acknowledgements
We are very grateful to the editors of The Sociological Review and the three anonymous reviewers for providing detailed and insightful comments and suggestions. They have all significantly contributed to raising the quality of our paper. We also thank Turid Rødne, Ugo Corte, Merete Jonvik, Cato Wittusen and Gurli Olsen (all at the University of Stavanger), for commenting on earlier versions. The paper was also presented at the EU Society for the Study of Symbolic Interaction Annual Conference 2023 in Cardiff, where participants provided helpful comments.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.