Sticking Together in ‘Divided Britain’: Talking Brexit in Everyday Family Relationships

Despite media headlines depicting the UK as a nation divided over Brexit, we know very little about how Brexit has affected personal relationships. This article addresses gaps in sociological understandings of Brexit and politics more generally by exploring how Brexit is experienced in everyday family relationships, focusing on how people orientate themselves towards discussing Brexit (or not) with their family and friends. Based on qualitative interviews, the article draws upon theories of relationality to explain the tenacity of family ties and to demonstrate how and why families ‘stick together’ despite often having different opinions about politics. It is argued that a conceptualisation of politics as relational is essential to understanding how they are experienced in everyday life.


Introduction
Since the lead-up to the 2016 UK referendum to leave the European Union (EU), Brexit has been a constant presence in British politics and media, infiltrating the everyday lives of people living in the UK and beyond. Sociological approaches to Brexit have however tended to focus on the public realm through discussions about the rise of nationalism and populism (Corbett, 2016;Pitcher, 2019) and the importance of analysing Brexit through the lens of Europe's colonial past (Bhambra, 2017;Virdee and McGeever, 2018). Others have focused on community and, in particular, the 'left behind ' -or, as Mckenzie (2017a) terms, the 'left out' -or on the ways Brexit is experienced by particular social groups, such as EU migrant families (Kilkey, 2017) or British 'expat' migrants living within the EU (Benson, 2020;Benson and Lewis, 2019). This article shifts the sociological gaze towards the domestic sphere in order to understand how families in Britain are living with Brexit in their everyday lives.
Life in 'Brexit Britain' poses particular challenges that differ from other, less exceptional, political events such as General Elections. Brexit has dominated UK news media since the referendum and, coupled with a sense of the unprecedented and unknown nature of Brexit, can feel frightening (Hervey, 2019) and exhausting (Carter and Davies, 2019). This creates a sense that more is at stake than in other voting decisions. A referendum vote forces the adoption of a binary position (Smith, 2017) and, coupled with the closeness of the result (52% voting Leave compared with 48% voting Remain) and an overriding media focus on division, 'Brexit Britain' represents a particular political climate through which families are navigating.
During this period of intense Brexit speculation and discussion, the UK media have represented Britain as a nation deeply divided over Brexit and many sociologists have sought to explain the referendum result through the lens of division. Delanty (2017) identifies European-wide cleavages between 'nationals' and 'cosmopolitans ' and Susen (2017) identifies 12 classes of division. As with much other work within the sociology of Brexit, these analyses take place at an (inter)national level. It has been in the UK media where the idea of 'Divided Britain' has been considered at the micro-level of people's personal relationships. High profile cases of families irrevocably split over Brexit have periodically made the headlines, the most notable being the 2019 public rift between Prime Minister Boris Johnson and his brother Jo over Brexit strategy, culminating in Jo resigning from his position as a Member of Parliament. Headlines included, 'Even Boris' own family don't trust him' (Crerar, 2019) and 'Blow for Bojo as bro Jo go goes' (Murphey and Cecil, 2019).
Though the idea of 'Divided Britain' has captured the public imagination, sociologically we do not know how Brexit has affected personal relationships. In this article I focus upon how families are talking about and dealing with Brexit in their everyday relationships. Drawing upon qualitative interviews conducted between February 2017 and March 2019 in a project funded by the British Academy, 1 the article demonstrates how practices of and orientations towards talking politics are deeply relational. Theories of relationality (Mason, 2004;Roseneil and Ketokivi, 2006;Smart, 2007) are employed to explore how approaches to and orientations towards talking Brexit are constituted as part of long-lasting tenacious relationships. It is often taken for granted in literature about political socialisation that political talk is a 'good thing', useful for deliberation purposes and the formation of political opinions (Conover et al., 2001;Dutwin, 2003;Zuckerman et al., 2005) and that connecting self to politics through 'talking politics' is 'civically desirable work' (Whitebrook, 2001: 8) with 'the family' seen as a key site of political socialisation (Jennings et al., 2009). However, this work tells us little about what is actually happening when people talk about politics, leaving the concept of political socialisation as something of a 'black box'. Furthermore, existing empirical studies of Brexit have thus far not focused on the importance of personal relationships. This article contributes to these literatures by arguing that it is not possible to understand how and why people approach political discussion without understanding their relationships.

Brexit and Everyday Life
Though many sociological analyses of Brexit have focused on its implications at an (inter)national level, emerging qualitative work explores how Brexit is experienced in everyday life. Many such studies seek to understand the Leave vote. For example, Mckenzie's (2017a) ethnographic exploration of working class Leave voters emphasises community and the embedding of class inequality in the Brexit vote. Patel and Connelly's (2019) analysis of the narratives of Salford Leave voters offers a geographically situated account of the 'subtle and seemingly non-racial ways in which xeno-racialised narratives underpin the accounts of Leave voters' (2019: 980). The focus on community in these studies provides valuable context but familial relationships remain underexplored.
A larger body of work centres on the effect of Brexit on migrants and their families, illuminating the practical and emotional burden of living under the shadow of Brexit. Guma and Jones (2019) point to the uncertainty and insecurity experienced by EU migrants living in Wales and Brahic and Lallement (2018) explore how 'resilience' is lived by French movers to the UK following the Brexit vote. Gawlewicz and Sofkasura (2019) explore the temporalities of Finnish and Polish migrants' responses to Brexit, emphasising the 'messy and fluid' (2019: 7) ways in which practical and emotional reactions to Brexit have evolved. Benson (2020) similarly emphasises the processual nature of Brexit through her analysis of the ebbs and flows of the insecurities experienced by Britons living in France and the ways they interact with pre-existing inequalities. Benson and Lewis (2019), in their analysis of the experiences of British people of colour living in the EU, also emphasise continuity over narratives of disruption; demonstrating that Brexit is an unexceptional continuation of everyday, structural and institutional racism that, in ways similar to Mckenzie's (2017a) identification of ongoing class discrimination, was part of participants' everyday lives before Brexit.
Through themes of everyday racism, migration and community, these studies go some way towards unpacking the lived realities of Brexit. However, there remains a lack of research examining Brexit and personal relationships.

Talking Politics in Families
Much existing work exploring families and political talk is centred around 'the family' as a key site of political socialisation where children's political views are influenced by those of their parents (Jennings et al., 2009). McDevitt and Chaffee (2002: 282) point to a growing recognition that children are not merely passive recipients of their parents' politics but instead play an active role in the process of socialisation, identifying a 'trickle-up influence'. Gamson (1992: xi) suggests that it is increasingly important to understand how people, '[f]aced with a cacophony of media clatter, popular wisdom, and knowledge from their own experience' make sense of politics and that analysing the ways they talk about particular political events, concepts and legislation can reveal the complexities of these sense-making processes. Taking a more everyday life approach, Nolas et al.'s (2017) ethnographic study demonstrates how the practice of political talk is transmitted across generations. Ekstrom (2016: 2) similarly explores young people's everyday political talk, drawing upon Goffman (1959) to conceptualise this as 'a social achievement, related to the exploration, disclosure, and management of self-identities in various social settings and relationships'.
This work outlines the benefits of talking politics for increasing political participation, as an act of deliberation and political decision making and as a process of selfidentity management. It has also been recognised that talking politics can be a risky and delicate practice. Eliasoph (1998) found that people feel better able to discuss politics in intimate settings. Similarly, Morey et al. (2012) identify a freedom to disagree about politics with close intimates. These authors are influenced by Goffman's (1959) work which they have used to explore how people present themselves as political beings in different social settings. However, this often has the effect of leaving 'backstage' interactions less well interrogated, with an underlying assumption that these settings are 'low risk' and perhaps therefore less worthy of interrogation.
Some studies have looked in more detail at political talk in intimate settings. Identifying a lack of attention to home as a site of difference in political geography, Valentine et al. (2015: 281) attend to the ways that encounters with difference, particularly inter-ethnic and inter-racial diversity, are negotiated within families, arguing for a more nuanced understanding of 'socialisation' where 'family members exercise agency and try to negotiate the contradictions and paradoxes of their ties to one another over time '. Ekstrom (2016: 6) found that young people's sense of their family as a 'safe place' to talk about politics was based on their conceptualisations of their parents as 'politically interested and the political talk as expected in the family'. In order to understand the complexities of 'talking politics' in family relationships and avoid overlooking 'backstage' political discussions in intimate, domestic settings; it is necessary to consider the nature of these familial relationships.
The concept of relationality, the idea that our relationships shape our lives and identities, helps get to the root of what family connections mean to people. Mason (2004) identifies the presence of 'relational thinking' in how decisions about where to live are made. Though not always positive, Mason (2004: 177, emphasis in original) demonstrates how 'practices and identities [are] embedded within webs of relationships', arguing that to understand these practices it is necessary 'to keep the processes of relating in focus just as much as, if not more than, the individual or the self'. Smart (2007: 49) describes relationality as 'a mode of thinking which not only influences decisions and choices but also forms a context for the unfolding of everyday life'. Pointing to the embedded nature of personal relationships, particularly with family, Smart emphasises their long-term, tenacious nature. The long-standing nature of family relationships matters and, as Finch (1989) notes, feelings of obligation towards family members result from a shared history of reciprocity. These relationships continue to influence who we are and what we do, even in situations where people are estranged from or have negative relationships with their family. Smart (2007: 45) describes such relationships as 'sticky' because, 'it is hard to shake free from them at an emotional level and their existence can continue to influence our practices and not just our thoughts'. By conceptualising people as embedded in webs of relationships with others that remain meaningful and continue to influence everyday thinking and decision making, even when not wholly positive, it is possible to make sense of practices of 'talking politics' and understand how family relationships can survive political difference. Furthermore, a relational approach enables a richer understanding of the everyday lived experiences of Brexit. Hall (2019: 44) similarly found that adopting a relational approach enabled her to understand austerity as 'a lived and personal condition'. Previous studies exploring the relationship between family and politics often overlook what actually happens when 'political socialisation' occurs and, as Valentine et al. (2015) argue, much work exploring political views fails to conceptualise participants as embedded within long-lasting relationships. Thus, we cannot understand the ways people make sense of Brexit as individual experiences but rather as practices embedded within meaningful relationships. Emphasising the 'I' in Mead's social self, Roseneil and Ketokivi (2006) draw attention to the role of individual reflexivity and creativity in forming the self. While Smart's (2007) concept of 'sticky' relationships enables us to understand why relationships might survive political disagreement, this use of Mead provides clues about how this is negotiated through the 'agentic reflexivity of the relational person' who reflects upon their own personal actions in relation to others (Roseneil and Ketokivi, 2006: 148). Furthermore, Roseneil and Ketokivi (2006: 148) draw upon Emirbayer's transactional approach to understand relationships as inseparable from the dynamic of the situations in which they are enacted, arguing: 'Any particular family, or "individual", gains its identity and becomes "what" it is in relation to surrounding persons, places, meanings and events.' Thus, a relational lens not only helps reveal political events as constituted within personal relationships but also, how our relationships are in turn shaped by wider political contexts.

The Study
This article is based on a qualitative study conducted between June 2017 and May 2019 funded by a British Academy/Leverhulme Small Research Grant. The study explored how the Brexit referendum was experienced in everyday family relationships, focusing on how it was discussed (if at all), how boundaries around talking politics in families are negotiated and how (dis)agreement has affected family relationships. The project comprised 24 interviews with 31 participants -17 individual, six heterosexual couples and one mother-daughter pair. The sample included eight men and 23 women between 18 and 72 years old. The sample was generated roughly along generational lines to reflect popular discourse about generational division over Brexit. Sixteen participants could be described as 'baby boomers', 2 11 as 'millenials', 3 four were in their 40s and did not fit into these categorisations.
Links between social class and voting behaviour are highly complex (Dorling and Tomlinson, 2019). Participants were recruited from a range of social class backgrounds and include both Leave and Remain voters as well as some who have since changed their mind and two who did not vote. I did not ask participants to reveal how they voted, though most did. Most of the couples had voted the same way in the referendum, although one person had not voted. The mother-daughter pair voted differently in a recent General Election. All participants had experience of some sort of political difference within their family and all had at least one family member or friend who had voted differently in the referendum. Political differences were wider than voting behaviour and many participants identified different sites of political disagreement, such as about politicians, a second referendum and General Elections.
Efforts were made to include participants from a range of ethnic backgrounds though this was difficult to achieve in such a small study. Of the 31 participants, one was British Chinese, one was British and Libyan, one was Somali, two were British Pakistani, one was German Pakistani, one was Polish and British and two were German. The rest of the sample were white British. In order to avoid over-sampling individuals who may have a particular interest in Brexit, and to instead reflect the everyday focus of the project, 20 participants were recruited through doorknocking in a range of socio-economic and ethnically diverse areas in the north of England (Davies, 2011). The sample was boosted through community groups, snowball sampling and personal networks. It is recognised that people who have experienced a lot of conflict in their families may have been disinclined to participate.
The interviews lasted between one and two hours and most took place in participants' homes with some in cafes. Interviews covered childhood experiences of discussing politics and current family relationships, including perceptions of family members' political views. Following Morgan (1996), a broad approach to defining 'family' was adopted and participants were encouraged to talk about those who they felt constituted 'family' as well as friends. Participants were then asked about the period surrounding the Brexit vote and to recall conversations they had about Brexit, including on the day of the result. Images from media coverage of the referendum were used as an aide-memoir in some interviews though most participants remembered details well. Most participants were able to narrate instances of political discussions surrounding Brexit and their narratives of the political views of their family members were revelatory of the importance of tacit understandings of the political views of others. The limitations of asking people to recall conversations in a qualitative interview are acknowledged though it has been fruitful to explore how people narrate the role of relationships in their -necessarily partialaccounts and it was possible to observe some political discussion and disagreement 'live' in joint interviews. The one-off nature of the interviews also meant that it was difficult to capture the ebbs and flows of Brexit as its aftermath rumbled on throughout the period of fieldwork and beyond. 4 Interviews were recorded and professionally transcribed and thematic and narrative analysis used to identify themes and typologies. All interviews were coded thematically using Computer Assisted Qualitative Data Analysis Software. Themes were identified according to predefined areas of interest stemming from literature and media coverage (for example the theme of 'generation'), as well as from the data (for example the theme of 'biting your tongue'). Narrative analysis was also used to draw out the role of biography in shaping practices of talking politics and typologies of styles of talking politics were identified. Though a detailed analysis of the raced nature of Brexit is beyond the scope of this article, these themes are important and warrant further research.

Talking Brexit
Despite differences in attitudes towards, and practices around, talking politics, the data indicate that decisions about if, how, when and why to talk politics are deeply relational. Here I outline some of the ways in which participants' narrations of how they talk politics are experienced as part of the long-lasting relationships within which they are embedded.

Recognition and Understanding
Despite experiencing differences of opinion about Brexit, many participants described attempting to explain their voting decision and thoughts about Brexit politics to others in their family. These conversations were rarely easy and had the potential to lead to conflict or to reignite past arguments. However, for many participants it was worth persevering, not because they thought they could bring their family member around to their way of thinking but because it mattered greatly that their reasoning was understood.
Brian, a 67-year-old white British Leave voter described several attempts to explain his views to his Remain voting son, Josh. Brian was particularly worried that his son would think poorly of him due to what he perceived to be media bias against Leave voters who he felt were depicted in the media as, 'blue-collar, working class and council estate, and the retired, who didn't care, and the ignorant'. Brian said of his conversations with Josh: I felt defensive. I felt . . . I'm sure Josh wouldn't have thought, you know, that his dad was ignorant or had not thought about the implications of being in the EU or out of it. So yeah, I felt a bit defensive and I felt the need to try and explain that it wasn't . . . that when we voted for Brexit, it wasn't any kind of kneejerk reaction and it was based on . . . you know, I had to be so careful because it's very easy for us to say, 'listen, we're 67 and we've got . . . we literally have a lifetime of experience of seeing the comings and goings of various political parties' . . . So I felt obliged to try and explain it as objectively as I could to him . . . Because I wanted him to see my point of view.
Brian is striving to explain himself to his son and his 'defensiveness', as he terms it, has led him to worry about his son's opinion of him. It is also clear that Brian feels he has a sense of generational expertise derived from his perceived greater life experience. It is significant that Brian speaks about being careful in how he applies this; he knows that over-using this perceived generational capital will result in Josh rejecting his message and the stakes are high, as Brian puts it: because he's my own kid, I wanted him to understand, you know, why we voted that way, that it wasn't done out of idiocy or a kneejerk reaction. I wanted him to understand. And also, I wanted to be able to sort of bring to his . . . mind what I'd experienced, you know. (my emphasis) In talking politics with Josh, Brian is both explaining and defending his decision to vote to leave the EU because he loves his son and he cares about his good opinion but also to try and educate his son whom he sees as less experienced. Both of these motives are compounded by the wider context of Brexit, where generational divisions in voting behaviour and attitudes towards Brexit have been emphasised (Norris, 2018) and where Brian perceives a media bias against Leave voters.
The idea of wanting to explain and defend one's position was common and other Leave voters in the study expressed similar concerns. Take Anita's (a 69-year-old white woman) account of her attempts to explain her vote to her granddaughter. Despite her use of language akin to 'postracist racism' (Patel and Connelly, 2019), Anita is satisfied that she has convinced her: Anita: I think a lot of people think if you voted out that we're racist and we don't want other coloured people coming to this country. It's got nothing to do with race at all . . . Interviewer: Were you able to explain that to her? Anita: Yeah. She said, 'all right, I know you're not racist, but I bet a lot of them are'.
Anita's account indicates that her granddaughter might also be trying to work to maintain her good opinion of her grandmother, differentiating her from other Leave voters whom she expects will be racist. These concerns echo Mckenzie's (2017b: 207) finding that working class Leave voters were hurt by media portrayals of them as 'backwards' and 'racist'.
Common to these accounts is a sense that people cared what their friends and family thought of them and they worked to ensure mutual respect of one another's opinions and views. This echoes Shipman and Smart's (2007) identification of the effort expended by their lesbian, gay, bisexual, transgender and queer or questioning (LGBTQ) participants in achieving familial recognition for their Civil Partnerships, even from relatives who were unsupportive of their relationship. Thus, family relationships can be conceived as 'sticky' in terms of their centrality to people's lives and the fact they cannot be easily abandoned, even when difficult (Smart, 2007). Furthermore, these 'sticky' relationships implicate the self and, in seeking to explain their political views, participants are reflecting upon their own habits, actions, practices and opinions through the perceived gaze of others (Roseneil and Ketokivi, 2006). Here talking politics involves self-reflections and reflections of the significant other which take place within the long-lasting tenacious relationships in which participants are embedded. As Valentine et al. (2015: 289) argue, 'familial love . . . is a crucial emotion in bridging "difference", by creating an emotional connectivity that brings the distant closer'. This familial love is evident when Brian said he wanted to explain his views to his son quite simply, 'because he's my own kid'.

Intimate Knowledge and Embedded Practices
The embedded nature of familial relationships did not always result in a desire to talk about Brexit and many participants spoke of drawing upon their intimate knowledge of others, built up over long periods of time, when deciding whether or not to engage in political discussions. This knowledge was used to decide if and how to talk politics. Take the following reflection from 67-year-old Debbie on why she avoided debating politics with her nephew Leo last time the subject of Brexit arose: Oh, we just changed the subject and left it where it was, because he'd never change his mind. It wasn't possible to change his mind. Never ever would you change it. Whereas that's just the way Leo is, he's got very fixed views. When you start discussing something with him he wants to pursue it to the Nth detail, and I think that's something to do with his Asperger's more than anything.
It is Debbie's intimate knowledge of Leo, and her past experiences of their discussions, that led her to change the subject. Caitlin (a 28-year-old white woman who voted Remain), similarly avoided talking politics with her dad because she knows from experience that he 'would never budge'. Caitlin also articulates how knowledge of the other person informs her 'relational thinking' (Mason, 2004) about whether to pursue political discussions more generally: 'I think it depends on the person as well. With my dad, I'm just like ignoring him and not really speak, but with others, like immigration and like stuff like that, I would maybe speak up.' These relational decisions are similar to Ekstrom's (2016) Goffmanian argument about the contextual nature of political talk and to Eliasoph's (1998) identification of the relative risks of talking politics in different settings. However, it is not enough to understand these decisions in terms of the differences between public (front stage) and intimate (back stage) settings because practices of talking politics are shaped by complex, sticky relationships in which participants are embedded with others. Following Mason (2004: 177), in order to understand why people talk politics, one must understand their 'processes of relating'. Shared histories and knowledge of one another built over years informed how participants engaged in discussions about Brexit and how they dealt with political disagreement.
Some participants described friends and family who were very outspoken about politics meaning that, even when they may have wished to avoid a discussion, this was often not possible and debates and even arguments were a common feature of their relationship. In these situations, many participants drew upon their knowledge of a person to understand their behaviour. Take the following example where Amy (a 25-year-old white Remain voter) is talking about her Leave voting friend's approach to talking politics: Like my boyfriend's best friend . . . [is] very outspoken about it and unapologetic . . . And I know that he's like that because his dad's very much like that and they're very similar. And they almost bounce off each other in a way. From looking at how they talk on social media and tag each other in posts and stuff, they're very much . . . He is the way he is because his dad's like that. And his dad's like that because of his granddad. And I can literally look at their family and you just get it.
Amy is able to draw upon her knowledge of her boyfriend's friend's family to understand his behaviours and in so doing uses what Edwards (2000: 206) terms 'kinship expertise'; her understanding of what relatedness means and how it works in people's lives, particularly what and how things are 'passed on' in families. Amy knows that attitudes towards politics and how they are expressed can be 'passed on' and this understanding comes, not only from what she knows of this particular family, but of how families work more generally derived from her own experiences. This understanding ultimately helps her to maintain the relationship despite differences of opinion about both Brexit and practices of discussing politics.
In addition to knowledge gleaned through long-lasting relationships, participants, particularly women, also spoke of drawing upon habitual ways of doing family, sedimented over many years, when dealing with political discussions. Morgan (1996) emphasises that 'family' is constituted through regularly doing everyday 'family things'. These seemingly trivial, mundane practices are given meaning in their label as 'family practices' and are intertwined with gendered family roles. Many women in the sample mentioned resorting to everyday mothering practices to keep the peace and avoid conflict when politics was discussed. Philippa (62 years old) describes herself as a 'regretsit' because she now regrets her Leave vote and talked about how her 28-year-old son has a 'fiery temper' and often clashes with her husband. Philippa describes adopting her longstanding role as peacekeeper when discussions about Brexit arise: Philippa: the youngest one, he's very intense and he won't give way . . . And he feels as if he has to have the last word . . . you think, 'right, time to stop now', but . . . it's just because he feels things very intensely. He wants to get his viewpoint across, I mean, he'll storm out and everything. Within half an hour, he'll come back. I mean, you might not speak for the rest of the day or the next day, but sooner or later he'll come round . . . He was the regretsit, like me . . . It's mainly him and [his father] that will clash. I mean, sometimes he'll clash with [his brother] because [his brother] will say, 'well, you voted to leave', and that will really rile him because he knows he did . . .
Interviewer: How do you handle it when that happens, then, what's your tactic?
Philippa: I then go into my try and smooth it down mode or I just stop joining in the debate, and then there's one less voice and hope it'll eventually just wind down.
Interviewer: And what's smooth it down mode? What sorts of things?
Philippa: Try and change the subject . . . I mean, it goes right back to when they were babies really, distraction. 'Ooh, look, Countryfile's on.' 'Look at that robin.' Totally irrelevant, but yeah. Or, 'oh, what time's so-and-so on?' Yeah, 'shall we do the dishes?' 'Try and smooth it down mode', as Philippa terms it, involves mundane everyday distractions (talking about the television and household chores) which are part of how Philippa does motherhood. These mothering practices were established when her children were babies and have continued well into their adulthood. The embedded nature of these practices comes across strongly in Philippa's account -she has a 'try and smooth it down mode' because it is part of how she mothers her children and she understands when and how to instigate it because of her intimate knowledge of her son's temperament, gleaned over many years. This understanding also helps her to make sense of his attitude to political discussion in a similar way to Amy's kinship expertise; Philippa knows that her son's intensity can be coupled with a bad temper and this knowledge helps her and her family to overcome these difficult moments when they occur. By focusing on Brexit as relational, it is possible to explore the entwining of the everyday habitual practices which define 'family' with deep personal knowledge of individuals sedimented over time as well as a sense of how kinship works more generally. These facets of relationality work together to constitute the ways Brexit is lived and practised in people's everyday lives.

Silence as Care
Remaining silent about politics was often used as a way of maintaining relationships and caring for others. Kate is a 64-year-old white woman who was interviewed with her husband Clive. Both Kate and Clive voted Leave and, despite regularly engaging with political news media and having clear opinions about many political issues, Kate said that she tried to avoid talking politics. In the following example Kate and Clive are discussing an incident at a dinner party involving themselves, their son-in-law's ('John') parents, Clive's sister and her French partner: Clive: And so John's father just opened the conversation. We'd all had a lovely dinner and a few glasses of wine, and a few beers, and he said, 'well what do you think about the Brexit?' . . . and he kind of said, 'well which way are you going to vote?', or something like that, and that opened a real debate about the subject . . . We realised we'd got different opinions and we realised it would start a bit of a debate . . . 'cause I know that they are all pro-Europe, with living in France.
Interviewer: And so then, how did the debate unfold . . .?
Kate: I think you agreed to disagree in the end?
Interviewer: And was it a bit uncomfortable or did you feel it was ok?
Clive: Yeah, it was slightly uncomfortable yes, at one time. Here both Clive and Kate are regretting that the issue of Brexit was raised. However, while Clive feels that expressing one's political opinions can be a 'good thing', Kate seeks to avoid political discussion, saying 'what people want to hear' to avoid confrontation. While one might assume that Kate is simply not interested in politics or that, following Eliasoph (1998), avoidance of political talk has led to apathy; her reluctance to engage in political discussion seems to come instead from a belief that there is nothing to be gained from spoiling a social encounter by challenging others' views. Later in the interview, I asked Kate to expand on the reasons why she sought to avoid political discussion and she replied by emphasising the importance of time with family and friends: 'I wouldn't like [to] go out and be having a meal and somebody bring something up that's, and then it just ends on a poor night, know what I mean? It's not worth it is it?' Ultimately, it seems for Kate, talking politics is something to be avoided because the potential for conflict distracts from valued time with loved ones. While Kate was an extreme example in attempting to avoid talking politics completely, the idea of avoiding 'talking Brexit' in order to keep the peace was a common one. One Remain voter for example said of her brother, who voted Leave: 'We could have had a massive falling out, but I think you think, what's the point? . . . At the end of the day, he's my brother and we get on very very well.' Rachel's sentiment that people are entitled to their different opinions was a common one: Myself and my partner, we voted to remain, and his dad voted to leave. So yeah, it, they've spoke about it, you know, the reasons why and things like that, so yeah. I mean I'm not into . . . I don't like confrontation, so I won't argue with people. It's people's opinion, and that's their opinion, and you know. That's up to them.
A number of participants talked about making a concerted effort to stay quiet and the idea of 'biting' or 'holding' one's tongue was raised often, evoking the idea that people wanted to talk politics but were able, with effort, to restrain from doing so. Mona (a 30-year-old mixed race woman who often debated politics with her father) stated she bit her tongue: many times, just the other day actually. I was thinking, my dad, I don't want to seem like I'm bossy or overpowering. Sometimes I think, is that how I come across, do I get a bit aggressive when I'm debating?' Here Mona describes avoiding political discussion to preserve her relationship with her dad and as part of reflecting upon her own actions and how they might be perceived by others. Holding back from engaging in talking politics can constitute an act of care for a loved one. Take Sarah and David's (a white couple in their 50s who both voted Remain) discussion of holding their tongue with David's Leave voting father. They describe changing their practices around talking politics with him now that he is getting older and lives alone: Sarah: I think you hold your tongue a lot with your dad actually, don't you? I wouldn't say there's many conversations.
David: Yeah, because it would end up . . . historically, it's ended up in a lot of rowing and he's at the age now where I think it's not worth it.
Sarah: I think as well as your parents get older, your dad's not . . . you know, he's on his own, isn't he? He doesn't really have anyone other than you, so there's an element of . . . I don't think your dad wants to go there either, does he . . . According to Scott (2018: 13), 'we make our feelings known by saying nothing'. Here, the silences practised by participants are not a mechanism for making their personal politics known, indeed Kate spoke about actively concealing her political opinions. Rather, they serve to preserve relationships. Thus, in addition to understanding the avoidance of Brexit talk as a performance of nothingness, we can understand this as a practice deeply rooted in relationality -as an act of care undertaken to ensure the maintenance of relationships akin to Brownlie's (2011) identification of the emotional support exchanged through 'being there' in ongoing relationships or Hall's (2019: 201) 'quiet politics . . . where relational spaces of togetherness and solidarity can be built'.
Not all participants had experienced this care. Aisha (a 20-year-old British Pakistani woman) talked about her mother's carelessness when discussing politics: she always turns a conversation to, 'oh, you don't really know much about it' and 'you're just immature and you don't really understand' and all this stuff. And then she just talks about how like I'm lazy and stuff like this. And then she starts turning it into a personal attack on me, I kind of feel like, I kind of feel like all those conversations tend to kind of go, tend to spiral off and talk about the personal instead then and less about the politics.
Aisha's experience stands out in the data as an important indicator of the sometimes onesided nature of the work that is expended in avoiding conflict. Here Aisha has to suppress her own feelings of frustration or anger in ways akin to Hochschild's (1983) emotion work, whereby actors work to induce, suppress and display the correct 'feeling' for a social context, even though her mother does not do this. It is also notable that Aisha's mother does not put the work into understanding Aisha's opinions about and orientations towards discussing politics that were identified earlier in the article, and, rather than deploy her knowledge of Aisha's temperament to understand her daughter, Aisha describes her weaponising this knowledge. The problems caused by this lack of silence illuminate the work that silence does in maintaining relationships across political difference.

Conclusion: Why and How Families Stick Together
This article has shifted the focus of sociological understandings of Brexit towards the sphere of personal life, bringing empirical data about people's everyday interactions with their families to a field hitherto focused largely on macro-analyses or on the experiences of particular groups such as EU migrants or working class communities. Arguing that it is not possible to understand how or why people talk about politics in the ways they do without understanding the relationships within which these encounters take place, this article points to the importance of the domestic sphere as a realm where Brexit is discussed and lived. In exploring the ways people have dealt with Brexit within their personal relationships, the article has countered popular narratives of Brexit as divisive, pointing instead to the tenacity of family relationships in the face of political differences.
The article has drawn upon qualitative interview data to explore the relational underpinnings of discussions about Brexit and this means that the empirical evidence presented here has necessarily relied on participants' recollections and narratives. Further research is required to explore the everyday ways in which these interactions are lived and this study has demonstrated that it is not enough to think about 'talking Brexit' only in terms of 'front stage' versus 'back stage' conversations. Rather, Brexit talk is situated within ongoing relationships often characterised by verbal and non-verbal acts of care (Brownlie, 2011) that form the tapestry of everyday family life. This was evident in how Kate spoke of not wanting to end an evening on a 'poor night', Brian's wish to explain his views to his son, 'because he's my own kid' and in the ways Philippa drew on everyday mundane practices of mothering to keep the peace.
By drawing upon theories of relationality, the article has brought new theoretical frames to the sociology of Brexit and politics more broadly, demonstrating that voters are relational beings, embedded in webs of connection with others. This conceptualisation moves current thinking about the formation of political opinion and voting decision making away from a focus on individuals. The article has offered an in-depth analysis of how and why people talk about politics with family, demonstrating the daily work undertaken to secure the understanding, respect and recognition of family members and the role of everyday practices and intimate knowledge, sedimented over time, to deal with and avoid confrontation. Furthermore, the article has established how not talking about politics is also used to maintain relationships, often constituting an act of care. Central to these observations is the tenacity of family relationships and this article has attended to the 'sticky' nature of people's intimate relationships which Smart (2007) describes as integral to people's sense of who they are and what they do in the world even if not experienced as positive. Some of these less than positive relationships are evident in the data presented here, for example in Caitlin's relationship with her 'stuck in his ways' father and Aisha's mum's lack of care for her feelings when talking politics. Following Smart (2007), it is possible to see why work might be expended in avoiding arguing about politics, even within difficult or conflictual relationships.
In making the argument for a relational understanding of Brexit, this article has slipped between discussions of how families have dealt with Brexit, other recent political events as well as more broad political ideas, ideals and memories. This is because these events are entwined in people's lives. Discussions about Brexit are negotiated, aborted and avoided because of the interactions about politics that have gone before. This embedded understanding is essential to making sense of the ways Brexit is lived by families and in demonstrating how and why families 'talk Brexit'. In turn, following Roseneil and Ketokivi's (2006) emphasis on transactional theory, the article has demonstrated the importance of situating understandings of relationality within the wider political context in which relationships are practised, challenged and repaired. It has been demonstrated that it is not enough to understand if, when and how political discussion and disagreement occur without exploring the work that goes into maintaining diplomacy. As such the article avoids a simplistic 'black boxing' of these processes under the moniker of socialisation, contributing new ways of conceptualising the role of politics in families through an unpacking of the complexities involved in talking politics within 'sticky' relationships.