The Burden of Conviviality: British Bangladeshi Muslims Navigating Diversity in London, Luton and Birmingham

This article considers the convivial turn in migration and diversity studies, and some of its silences. Conviviality has been conceptualised by some as the ability to be at ease in the presence of diversity. However, insufficient attention has been paid to considering who is affectively at ease with whose differences or, more particularly, what the work of conviviality requires of those marked as other vis-a-vis European white normativity. Drawing on in-depth qualitative interviews with British Bangladeshi Muslims in London, Luton and Birmingham, we argue that a focus on ‘ease in the presence of diversity’ obscures the ‘burden of conviviality’ carried by some, but not others. We discuss three key types of burden that emerged from our data: the work of education and explanation, the work of understanding racism, and quite simply the work of ‘appearing unremarkable’.


Introduction
There is long-standing sociological interest in the interaction between ethnic groups, especially in urban spaces. In recent years, this has been framed as a 'convivial turn' in migration and diversity studies (Gidley, 2013). The concept of conviviality considers prominence within the British Bangladeshi imagination, as the heartland of the Bangladeshi community (Alexander, 2011), as well as its geographic proximity to the City of London, may in fact heighten British Bangladeshis' consciousness around how to navigate and 'manage' their Othering.

Multiculture and Conviviality
From the work of WEB Du Bois, or the Chicago School, sociological interest in the interaction between ethnic groups in urban space has a long history. More recent work suggests that ethnic and racial differences have become an increasingly normal part of city life and inter-ethnic interaction increasingly routine. Such work, dominated by ideas of super-diversity, multiculture and conviviality (Neal et al., 2013), affirms how everyday ethnic interaction across difference is commonplace. From prosaic multiculture (Amin, 2002), 'everyday cosmopolitanism' (Noble, 2009), 'everyday multiculturalism' (Wise and Velayutham, 2009) to 'commonplace diversity' (Wessendorf, 2014) several terms have been proposed to describe the intermingling of people in diverse settings (Lapina, 2015). This 'convivial turn' in migration and diversity studies (Gidley, 2013) has understood processes of cohabitation as 'empathetic' (Neal et al., 2019: 71) or 'unpanicked' (Noble, 2009: 46) ways of living with difference.
This approach has, however, prompted critique highlighting how: the notion of conviviality . . . slides away from its radical emphasis on uneasy and fragmented negotiations between connected others towards more familiar integrationist values in which difference is sanitized around contact and the hierarchies of cultural difference are flattened out or obscured. (Neal et al., 2019: 70) Celebratory discourses on encounters with difference mask underlying inequalities (Berg et al., 2019), and any politics of minority 'inclusion' is inadequate if it does not change the ideological terms of inferiority to white normativity (Gilroy, 2004;Valluvan, 2016). If ethnic differences are represented through the politics of 'recognition' and 'reconciliation' (Amin, 2012), requiring 'respect' or 'accommodation', then these identities of difference are presumed to be ontologically authentic and therefore reinforce hierarchically indexed ethno-racial positions (Valluvan, 2016). For Gilroy, conviviality occurs when ethnic differences do not require accommodation because they simply cease to require scrutiny; that is, there is indifference to difference (Amin, 2012). Despite the convivial turn, there is little understanding of what everyday inter-ethnic interaction entails for those racialised as requiring 'respect', 'recognition' or 'accommodation'. Within a context where 'indifference to difference' is a project in ambition rather than realisation -in which the presence of difference has not been 'habituated' (Noble, 2013) and where racialised bodies are suspect bodies -what does inter-ethnic interaction require of those marked as inferior vis-a-vis European white normativity (Valluvan, 2016)? This article addresses the work required of some in everyday mundane interactions that may otherwise be deemed convivial; it draws attention to the varied powerrelations active in different spaces, which have often been missing from the conviviality literature (Wise and Velayutham, 2014).
Moreover, we argue that the literature on anti-Muslim racism needs to be brought into conviviality debates, with a reflection of the multi-layered and intersectional nature of this racism. Our analysis shows how broader discourses of gendered anti-Muslim racism, building on earlier and parallel forms of anti-South Asian racism, framed through debates about Britishness, citizenship and the War on Terror, are internalised and resisted in everyday encounters. Underlying these encounters lie deep-rooted structural inequalities through which British Bangladeshis experience housing insecurity, one of the highest unemployment rates, static or worsening ethnic pay gaps and consequently have an estimated wealth of one-tenth of the wealth of white British households (Runnymede Trust, 2020). These structural inequalities and the systemic racism through which they are produced inform the everyday inter-ethnic interaction in ways that are not always seen so much as felt. We argue that the literature on conviviality and multiculture has sometimes obscured the background labour, required of some but not others, to ameliorate the interrogation of difference that in fact pre-empts the encounter. We suggest that this burden of conviviality is an expression of racist structures that are multi-layered the outcome of different forms of racialisation, as both South Asians and as Muslims.

London, Luton and Birmingham
The research focused on the experience of British Bangladeshi Muslims in three urban settings -Tower Hamlets in London, Luton in Bedfordshire and Aston/Smethwick in Birmingham. In working across these different field sites, our aim was to consider how space and geography (and consequently different dynamics and demographics) inform people's experiences of citizenship. In doing so, the article demonstrates significant heterogeneity not only in terms of participants' experiences of anti-South Asian and anti-Muslim racism, but also in terms of their responses.
The Bangladeshi community is well established in the UK. The earliest settlers arrived in the 19th century and by the time of the last census there were 447,200 Bangladeshiorigin men, women and children (Office of National Statistics [ONS], 2011). The vast majority originated from the region of Sylhet, and much of the migration flow has resulted from family sponsorship (Kibria, 2011). These two factors have produced significant geographic concentration in 'Inner London' and the borough of Tower Hamlets in particular, where almost half of the Bangladeshi population is located (222,127). Luton was chosen as a satellite town to London because it represents an interesting comparator to Tower Hamlets in terms of ethnic concentration. It has a smaller Bangladeshi population than Tower Hamlets (13,744 in 2011), living alongside a sizeable Pakistani population of 24,279 (ONS, 2011). Luton is also the 'home' of the English Defence League (EDL) a far-right anti-Muslim organisation and has received significant media attention as a result.
The region with the second-highest concentration of British Bangladeshi Muslims in the UK is the West Midlands, with a population of 52,477 (ONS, 2011). The majority live in the city of Birmingham (32,532) where one of the largest concentrations is in the ward of Aston. The city has played a vital role in shaping the politics of race in British society (Solomos and Back, 1995). Furthermore, as religion has increasingly become a key theme in contemporary political debates, the city has continued to feature in public commentary as a site of concern (Miah, 2017), notably in relation to education; that is, the Trojan Horse affair (Holmwood and O'Toole, 2017) and the 'No Outsiders' controversy. Birmingham is, therefore, also a key site of contestation for some of the issues around identity and discrimination that this article explores.

Methodology
In total, 120 British Bangladeshi Muslims were interviewed in 75 interviews in the form of semi-structured dyadic interviews with parents and children (45 interviews), narrative interviews with participants over 60 years of age (14 interviews) and semi-structured interviews with members of civil society (16 interviews). The research was conducted as part of two overlapping projects that sought to examine the relationship between experiences of citizenship and transnational practice. The parent-child dyads took the form of semi-structured interviews and reflected a mix of mothers, fathers, daughters and sons, to capture the way in which citizenship is mediated not only by 'race' and ethnicity but also by gender and generation. A generational approach has been less usual in Britain than in the USA but the impact of generational change has been of growing interest (Heath, 2014), partly in response to a history of homogenising ethnic groups in research and policy. The interviews sought to examine socio-political in/exclusion concerning experiences of citizenship; and to consider these local experiences of citizenship alongside transnational social, political and religious engagement (Redclift and Rajina, 2019).
The theme of the 'burden of conviviality' emerged from discussions regarding experiences of citizenship, and anti-South Asian and anti-Muslim racism. The issues were raised in all three interview types. They were more common in the dyad interviews, which may be because of the conversational nature of interviews with more than one participant, and because there were more young women recruited for these interviews than for any other interview type. It also may be because different generational responses to the 'burden of conviviality' provoked greater discussion during multi-generational interviews. These discussions were least common in the narrative interviews, in part because conversations here focused more on biographical themes of migration and settlement and historical experiences of racism.
In Tower Hamlets and Luton initial access to interviewees was gained with the help of the Swadhinata Trust, and in Aston/Smethwick access to interviewees was gained with the help of the Community Connect Foundation. In all field sites a sample was drawn using purposive sampling methods and the criteria for the inclusion of cases was a spread of socio-economic backgrounds and an even spread of mothers, fathers, daughters and sons. After the intial interviews were conducted, snowball sampling began. Alongside the snowball sample the researchers contacted a broader range of relevant organisations and drew on personal networks to try to ensure sample selectivity was minimised, but it was not completely eradicated. Interviews were conducted by one of the co-authors, a British Bangladeshi female academic. They were conducted in English or Sylheti depending on the preferences of interviewees. The majority of the interviews took place in interviewees' homes and in local community centres to ensure participants were as relaxed and comfortable as possible and minimise any power asymmetry between the researcher and participants. It was clear from the data generated that the researcher's 'insider' status helped to build rapport and the interviews were rich. The majority of interviews lasted approximately 90 minutes, and some of the narrative and civil society interviews were as long as three hours. All interviews were audio-recorded and transcribed, and names have been replaced with pseudonyms in all cases. The data were coded in Nvivo and thematic analysis applied. The examples of data chosen to illustrate our argument are those coded under a theme labelled 'the burden of conviviality', separated into the sub-themes we discuss below.
The production of community involves labour 'not just because it is hard . . . but because it is productive, transactional and cumulative' (Noble, 2009: 53). However, recognition of this labour, like much work on conviviality, has largely focused on the labour, habits or practices of those embodying normative whiteness. In relation to the 'production of community' there is a notable absence of work that examines the labour, habits or practices of those minoritised as 'different' in contexts in which differences remain remarkable. We found that this labour principally entailed appearing 'not so different' as to be considered threatening. This arose principally from two sources, the association of Muslims with terrorism and negative stereotypes about Muslim women. We discuss three key types of labour that emerged from our data: the work of education and explanation; the work of understanding racism; and quite simply the work of 'appearing unremarkable'.

The Work of Education and Explanation
Whether this was about being asked to justify personal decisions around dress and behaviour or to take responsibility for acts committed by other Muslims, many described feeling they had to explain and defend their religion although attitudes to this varied. Some almost felt they had a 'duty to educate'; to better inform the non-Muslims they came into contact with. Much of this kind of discussion focused on the labour of disassociating from terrorism. One such participant was Kobir in Birmingham who told us how he was treated differently after 7/7, by people he had previously considered friends: the few people doing some bad things over there. I'm not responsible for it. They don't represent me at all. (Birmingham, Civil Society, Rothon (43, male, second generation)) Amina believed that if Muslims did not engage in this work, then animosity towards them would increase: So I think it's our job, as Muslims, to show them . . . with our manner, speech, the way we walk, talk . . . Even if our neighbours are non-Muslim, we should still take care of them, keep an eye out for them, say hello to them, you know; see if they need any help with anything . . . So through that, they will see that: 'Hold on a minute, these guys are Muslim, but they're not attacking me, they're not threatening me, they're not going about attacking people with bombs.' (Luton, Dyad, Amina (39, female, second generation) and Omar (18, male, third generation)) Amina has to demonstrate, through her behaviour, that she is not 'going about attacking people with bombs'. Educating people in this way is simply her 'job', as a Muslim. Comments like this highlight the way in which conviviality, in its more radical formulation, is only possible upon freeing racialised bodies of suspicion (Valluvan, 2016). For some participants, this work of explanation and education was specifically gendered. Women like Jennifer and Farina were additionally required to explain or justify their choice of dress both outside and inside the home, particularly because of the perceived association between the hijab and extremism: Jennifer: If you were an Asian woman at (the school where she worked), you were always justifying yourself. After a while, it got a bit exhausting and I think as times have changed as well, especially 9/11 and things, I think who and what you like has become a big part of how people perceive you, unfortunately.
Farina: . . . and then with family, I had to keep justifying the fact that it made them inferior . . . it made them feel guilty and their guilt made them say things . . . (like) 'why do you have to wear the hijab?', 'why can't you wear a blouse that's short?' . . . 'why do you have to do this and why do you have to do (that)?' (Tower Hamlets, Dyad, Farina (48, female, second generation) and Jennifer (29, female, third generation)) Through the repeated association between Islam and terrorism the word terrorist 'sticks to some bodies' and 'by generating the other as an object of fear' that fear 'is then taken on by the other, as its own' (Ahmed, 2004:76). Farina's accounts exemplify how this fear has been internalised by her family, which leads to the questioning of her sartorial choices. Explanation or justification is, therefore, required by Muslims and non-Muslims alike. Ethnic and religious dress can be seen as an expression of pride and love for one's heritage, or an individual expression of identity, but group in/exclusion is also made apparent through modifying the body (Rajina, 2016). For Farina's daughter Jennifer, in the post 9/11 landscape, the choices you make as a South Asian Muslim woman become markers of an identity that is up for dispute. For British Bangladeshi Muslims this builds on longer characterisations of South Asian women (Brah, 1992;Parmar, 1982;Rashid, 2016); anti-South Asian racism and anti-Muslim racism intersect.
As Bilge (2010) argues, the veil has become an over-determined cultural signifier, and female participants frequently mentioned the 'extra effort' that was required by those who wore the hijab to show that they were sufficiently British. As Karina explains: You're having to answer all the time . . . you have to justify, let people know . . . these . . . British Bangladeshi women who are living in this country . . . they can speak English, they're doing good jobs and they're providing for their families . . . I have to make an extra effort to tell people that even though I'm wearing a hijab, I can speak English, I can do stuff. (Luton, Civil Society, Karina (44, female, second generation)) Karina attempts to disrupt enduring stereotypes about South Asian women's (lack of) English. Moreover, her reference to her ability to 'do stuff' is telling. Women regularly discussed their frustration at the range of orientalist stereotypes the hijab inspired in non-Muslims around passivity, powerlessness, exoticism and danger. The hijab both disqualifies the wearer from agency and symbolises a certain kind of militancy and empowerment through religion (Bilge, 2010;Rashid, 2016). The data above speak to this idea of educating people to understand that these women were not defined by the veil they wore.

The Work of Understanding Racism
A common theme among participants, particularly when anti-Muslim racism was discussed, was the idea of understanding racism or 'not blaming people for it', which might be seen as a form of exoneration, appeasement or even empathy. This often stemmed from a recognition of the lack of knowledge about Islam among non-Muslims and the role of the media in reproducing ignorance. Public knowledge of Muslims is constructed and circulated via the media (Poole, 2002), and this coverage is overwhelmingly negative (Allen, 2012); with terrorism the most recurrent theme (CFMM, 2018). As Omar explains: So they obviously think that . . . we should get rid of these people (Muslims), they don't have the right to live here in this country . . . But I don't blame them, because they don't know. It's like, I don't know a lot about Christianity, so if I kept on seeing on the news that Christian people are terrorists, I would think yeah, it could be true. (Luton, Dyad, Amina (39, female, second generation) and Omar (18, male, third generation)) Afsana: I know they don't mean to stare and it's just something, maybe it's just fascinating to them and they're not doing it in a bad way . . . maybe they're just curious, and it's not their fault, what they watch in the media, what they heard, this is not their own judgement, they are just judging on someone else's judgement? (Tower Hamlets, Dyad, Khadijah (38, female, second generation) and Afsana (21, female, third generation)) For some participants, however, this understanding was less the result of the misrepresentation of Muslims in the media and more about taking responsibility for actions committed by Muslims; that is, accepting the blame: There is the Islamophobia, and sometimes I don't, I don't blame people . . . because . . . there's a group of people who are out there, you know, going against our religion and showing that the religion in Islam is about this, which (it) is not. (Luton, Civil Society, Karina (44, female, second generation)) Arguably this can be seen as having internalised 'good Muslim' (Mamdani, 2005) narratives. Haabeel below suggests, that through acts of terrorism committed by some Muslims, 'we expose ourselves'. As a result, he suggests a certain amount of understanding or acceptance was required: There is a reason for it (anti-Muslim racism) . . . We have proven that we are not so clean. You have seen suicide bomber, this and that, under the burkas, yes, hijab . . . I am not in support of hijabi and burkas. Okay, if you want to wear it then you stay inside. Don't go out, if you don't want to expose yourself . . . we expose ourselves. (Luton, Dyad, Leepa (19, female, third generation) and Haabeel (47, male, second generation)) For other, often older participants, this work of understanding racism involved the work of forgetting. Anti-Muslim racism was expressed in these moments as having occurred so long ago it was not worth remembering. Here Sadia calls out the anti-Muslim racism that her mother seems to want to ignore: Dipa: Well, to tell you the truth I haven't really, physically . . . I haven't really seen anything, any activity like that (anti-Muslim racism) around our area, because it's Asian population and I don't think anyone really has that . . . Sadia: Do you remember in Blackpool, those boys were calling me and Aashi terrorists? Dipa: No, that was so many years ago. I can't remember, it was nine years ago. (Birmingham, Dyad, Dipa (40 female, second generation) and Sadia (18 female, third generation)) Dipa's use of Muslim and Asian as synonyms was common, reflecting again the intersection of multi-layered racist structures and the evolving markers and processes of racialisation. Moreover, her forgetting may or may not be conscious (she clearly remembers the incident) but reveals the way that narratives of post-racial harmony and conviviality produce feelings of self-doubt. This can be seen as a form of racial anxiety when attempting to evaluate whether others' behaviour constitutes racism. In this context, the idea of 'not pulling the race card' was frequently mentioned: Interviewer: Why is it difficult to talk about (racism)?
Anisa: Because what if you're wrong, what if it's not discrimination, what if people just generally don't like you? And then it's more like you're just blaming all of your issues on being discriminated against . . . So I've genuinely never personally felt discriminated, but sometimes you think, oh . . . if I do say it's discrimination am I being paranoid, am I just blaming it on something and really it's an issue about me? Shamea: But also . . . the first job I was talking about, not getting that job, an English friend helped me write the letter. So at the time even when I said it she thought . . . 'are you sure you want to pull that card'. She said it, I said, 'but Claire, I never, ever do that' . . . So even an English person thought straight away, you're pulling that card, you know, the race card, she said it, that was a long time ago. You know, Rina was a baby but . . . when you're in the workplace you're professionals, you don't talk about those topics . . . you don't talk about those things in a professional environment.
Interviewer: Why do you think, for example, you can't bring up the fact that there is a possibility of it being discrimination? Shamea: Because I can't prove it. That was the very first job . . . That was a long time ago. (Birmingham, Dyad, Shamea (42, female, third generation) and Anisa (21, female, fourth generation)) Shamea is talking about discrimination at the application stage but her suggestion 'that (it) was a long time ago' works to minimise the initial assertion, as though it might not happen now or is unimportant. There is significant evidence, however, that this discrimination is commonplace. Research suggests that people with Asian or African-sounding surnames have to send in nearly twice as many curriculum vitae (CVs) to get an interview (Di Stasio and Heath, 2018). Discrimination in recruitment impacts employment rates. The unemployment rate of Bangladeshi women (at 19%) is almost four times that of white British women and discrimination is likely to be a key factor (Runnymede Trust, 2020).
Visible, gendered signifiers of 'difference' such as veiling also provoke responses in others that appeared to demand the work of understanding or exoneration. Indeed, Layla jokes about this, suggesting that she, as a hijab-wearer 'has an advantage' when it comes to getting a seat on the bus: Layla: I use a lot of public transport because I don't have my own car, so I do have an advantage of being a Muslim, no one comes and sits by you. They see me wearing a hijab, they might think I have a gun in my bag.
Interviewer: So people don't come and sit next to you?
Layla: There are certain (people) that do, but there are some people I can tell that are like okay, they will go to a seat that is further back. I am like thank God in that sense, no one has to come and sit by you . . . but then in certain ways I feel as if probably you need to talk to me first. (Birmingham, Dyad, Layla (19, female, third generation), Fatiha (22, female, third generation) and Shamirun (38, female, first generation)) Faghira also highlights the impact of the 'hijab' in public space, and again she does not 'blame them' for their suspicion of her, even if she experiences its injustice: They target you if you're wearing a hijab, you know. And I have seen a difference . . . And I, I don't blame them, because of what the media has shown . . . Obviously they're brainwashed into thinking what they see but what I find is, don't treat me like it's my fault. (Tower Hamlets,Dyad,Faghira (31,female,third generation) and Naba (53, female, first generation)) For Amerah, racist abuse, also associated with her niqab, has been normalised to the extent that she accepts it and shows compassion towards abusers: Amerah: Yes, I choose to wear the niqab as well and I had the same. I had people coming up to me, saying stuff. Even very recently . . . there was this white guy, bless him, I think he was homeless but he was walking around and then shouting at me 'oh, you're going to blow us up'. He must have had mental health issues, but that is such a normal thing. You kind of just let it go.
Interviewer: Why do you let it go?
Amerah: Because it's just so normal, especially who would you tell and what would they do? . . . There is nothing you can do, unless you actually physically get hurt . . . but just words. It seems so normal. (Tower Hamlets, Dyad, Amerah (24, female, second generation) and Hasina (42, female, first generation)) Here we see the diversity of responses to 'the burden of conviviality'. Some participants downplayed anti-Muslim and anti-South Asian racism and did not want to label them in that way, while others challenged them. Fatiha, below takes full personal responsibility for the behaviour of others: I think it is all down to the person. If you're an outgoing person, no one would actually think wrong of you . . . I work with a lot of white people, Indians and I work with a mix. I find they are all friendly, it depends how you are . . . I think it is always down to the person. If you're nice to someone else, and if you are loving and stuff . . . It is like I have all of these white people and they actually come and say I haven't seen you in ages, they hug me and stuff like that . . . It depends on how you are. (Birmingham, Dyad, Layla (19, female, third generation), Fatiha (22, female, third generation) and Shamirun (38, female, first generation)) Fatiha's intervention in sanitising her workspace through an emphasis on the interpersonal ('it is all down to the person') highlights the way encounters with potential racism must be managed by those Othered so as not to disrupt white privilege. She has to do the work of getting others on side, if she is to pre-empt racist encounters. The narrative of individual responsibility obliterates the structural injustices involved. Her reference to Indians in this context links back to the hierarchically indexed ethno-racial positions through which conviviality plays out (Valluvan, 2016). It also speaks to the contradictions between her positioning as both 'Asian' and 'Muslim', which could be rooted in linguistic, religious or class-based differences. Here, conviviality in the sense of 'being at ease in the presence of diversity' 'does not seem well equipped to address the question of . . . who is in a position of accommodating whom or being "affectively at ease" with those differences' (Lapina, 2015: 39). Gilroy (2006: 40) explained that 'conviviality is a social pattern in which different metropolitan groups dwell in close proximity but where their racial, linguistic and religious particularities do not . . . add up to discontinuities of experience'. In the absence of conviviality, then, the burden is on those marked as 'different' to conceal aspects of themselves: I don't understand what I'm supposed to do as a British citizen and then in all of that I'm bringing up my kids as well who are equally confused . . ., you know, there was a joke the other day . . . (a) conversation I had with my older daughter, she's run out of books to read in the house and I was saying to her . . . 'why don't you read Tariq Ramadan's book which is what I've got at home?' and she said, 'Oh, imagine that mum, me taking that book to school and opening it up and reading it. Imagine what the teachers will say' and it hit me that oh my god, my kids have to be so careful of what they read in public and it's not visible to me but it's going through their heads as well that they have to hide certain aspects of what they're doing from the school because they'll be seen as extremist. (Tower Hamlets, Civil Society, Mahjabeen (33, female, second generation))

The Work of Appearing Unremarkable
Here we see how 'difference' is couched in the banality of state surveillance embedded in the everyday workings of the UK's counter-terrorism strategy, Prevent, whereby teachers are encouraged to act as informants looking for signs of radicalisation among their students (Faure-Walker, 2019) and the Muslim home turns into a pre-crime space (Fernandez, 2018). Prevent initiatives focused on women (including directives about speaking English) are arguably about ensuring they are able to 'spy' on their children (Rashid, 2016). In this context of surveillance, participants discussed the assimilative pressure to conform to the behavioural norm (Puwar, 2004: 150); 'appearing unremarkable' is about 'blending in' as much as possible: Ayana: As a Muslim, as a Bengali, I just represent myself as how I believe I should be represented. I act normal, I act like everybody else, and I try to blend in because at the end of the day, we're all human. Being a Muslim is not something I can control and . . . you just live how you are and blend in as much as possible.
Mohammed: I think . . . we need to just reduce the gas a little . . . We need to look inwards as a community. We have got some serious flaws. (Tower Hamlets,Dyad,Mohammed (47,male,second generation) and Ayana (19, female, third generation)) Ayana's comments are ambiguous. She claims to act 'normal', to behave as everyone would/should, but still aims to 'blend in', recognising that she is not a part of the 'norm' and therefore has to adapt to it. 'Blending in', 'calming down', 'reducing the gas' all speak to earlier discussions about 'not playing the race card' in order to fit in and Mohammad's comment about 'having serious flaws' chimes with the earlier reference to 'taking the blame'. They also reflect a wider phenomenon whereby 'the parameters of discussions about living with diversity have instead been reframed through suggesting "benign" or "de-racialised" discourses' (Harries, 2014(Harries, : 1110. Again, this is gendered, and Afsana below (like Fatiha in the previous section) shows how for Muslim women this labour can involve expressive emotion work too: Afsana: I feel like I have to act a certain way so people don't think I'm bad, I have to be more smiley, I have to be more nice. Why do I have to do that? . . . when I'm on the train going somewhere far and I see all these people and they're looking at me strangely, and I have to . . . I open up my smile more, and stuff like that . . . Whoever I'm with, I'm like, oh my god, no, just talk English when I'm outside, so they know I know their language, and I'm one of them . . . I'm like don't talk Bengali because they're going to think . . . is she plotting something? . . . All these things come to my head, that's why I feel like I have to be not myself when I'm outside sometimes. That's what I mean by freedom, a little bit of my freedom is gone. (Tower Hamlets, Dyad, Khadijah (38, female, second generation) and Afsana (21, female, third generation)) As Lapina (2015) notes, a challenge of using conviviality in the sense of being 'at ease in the presence of diversity' rests on the fact that being affectively at ease is a matter of interpretation. Afsana feels she has to be 'more smiley' on the outside, but it causes her pain and frustration on the inside. Whose perspectives or experiences disappear in, or behind, this conviviality (Lapina, 2015)? As Amerah notes below, the pain and frustration took a heavy toll, affecting her experiences at university. Yet, in contrast to some of the other participants she acknowledges these instances as the result of racism: Amerah: In Tower Hamlets, I went to (school) just down the road, it is 99% Bengali probably. So, I have never really felt different . . . then you go to (Russell Group University) and it's almost all white . . . and there was very everyday racism, but . . . I would brush it off and I was known as a good brown person because I'd get along with everyone . . . I didn't know that when people . . . ask if you shower with your hijab . . . I didn't realise that was othering and racism . . . That was an everyday experience, and it grounded down and I used to come home and cry and tell my mom I didn't want to go to uni anymore . . . I just know how much it used to bring me down . . . it's just how much energy I have to invest in being someone who I am not . . . having that mask on . . . You shouldn't have to be like that, but it's survival really. ( Space and location regulate and define what is perceived to be acceptable to 'show', and even experience, in public (Rajina, 2018). The size of the British Bangladeshi community in Tower Hamlets appeared to provide some protection from 'the burden of conviviality' (see Ahmed, 2005) but frequent movement in and out of the borough may also have made this burden more apparent. Moreover, the borough's prominence within the British Bangladeshi imagination, as the heartland of the Bangladeshi community (Alexander, 2011), as well as its geographic proximity to the metropole, may heighten consciousness among British Bangladeshis around how to navigate and 'manage' the way they are Othered.

Conclusion
This article addresses the work required of those marked as inferior vis-a-vis white normativity in everyday interaction, which may otherwise be deemed convivial. By focusing on British Bangladeshi communities in paradigmatic multicultural spaces, we draw attention to the variation in power-relations active in different spaces, which has been missing from some of the conviviality literature. In particular, we demonstrate the burden of 'putting other people at ease' on those racialised as South Asian and Muslim, and how broader discourses of anti-Muslim racism, building on earlier and parallel forms of anti-South Asian racism, are both internalised and resisted in everyday encounters. We argue that three main forms of labour emerged in this process: the work of education and explanation, the work of understanding racism, and quite simply the work of 'appearing unremarkable'. Interviewees described how they felt pressure to explain or defend their religion or overcome negative stereotypes by offering alternative representations to mainstream narratives. They described how they felt obliged to avoid 'playing the race card' so as not to appear 'unprofessional' and to 'let it go' when they experienced racist abuse. Most strikingly of all, many took for granted the need to appear unremarkable in public space: to smile on the bus, to 'blend in' with those around them, and to speak English loudly to show they belonged. We argue that the literature on conviviality and multiculture has focused on 'ease in the presence of diversity', obscuring the background labour, required of some but not others, to ameliorate the interrogation of difference.
Moreover, we argue that the literature on anti-Muslim racism needs to be brought into conviviality debates. What we have termed the 'burden of conviviality', with its hidden labour, is an expression of established racist structures that are multi-layered; the outcome of different forms of racialisation, as both South Asians and as Muslims. Our participants' responses to this burden varied. Some sought to challenge it while others attempted to avoid it. However, we argue that the burden echoes the narrative that racism is principally about individual encounters and thus can only be addressed through interpersonal adaptation and accommodation. Through this narrative normative whiteness is upheld, and established racist structures ensure that the burden of conviviality is not borne by everyone.
The work of education and explanation was discussed by men and women in all field sites, but it became a particular focus of conversation among young women concerning the need to justify choices around dress and veiling. In a similar way, the work of understanding racism came out in the narratives of both men and women, in all three field sites, however, because 'not blaming others' for anti-Muslim racism was often expressed with reference to visibility, women discussed this labour of empathy or 'letting it go' more frequently. There was a generational dimension too, as younger women called out experiences of anti-Muslim racism that older participants were inclined to ignore or forget. Finally, all our data on the work of 'appearing unremarkable' came from women, and all of those women were between the ages of 18 and 46. Moreover, these data emerged from Tower Hamlets. We suggest that these participants possibly became more aware of the burden of conviviality when they left the relatively safe East London borough, and proximity to the metropole may heighten consciousness around how to navigate and manage the way they are Othered. Gender, generation and space are clearly important in determining who is confronted more often by anti-South Asian and anti-Muslim racism, how that is experienced, and who is in the position of 'accommodating' whom.
The 'burden of conviviality' is the burden placed on the shoulders of those racialised vis-a-vis white normativity (as South Asian and Muslim), to educate, understand and put at ease those not racialised as 'different'. The efforts of the participants to 'blend in' is instructive of the powerful ways they are positioned against the dominant culture. This becomes the blueprint for surviving the burden of conviviality: disappearing into normative whiteness. Our data suggest that until diversity itself is seen as unremarkable, this burden is sometimes accepted, sometimes resisted, but always unevenly distributed.