2.1Contesting Brexit in language

In 2016 the UK held a referendum on its membership of the European Union. The result favoured leaving (52% leave and 48% remain). This highlighted political division as the UK negotiated its exit from the EU, also known as ‘Brexit’. There is a wealth of political research on Brexit, from tracking causes, such as key political issues (Clarke et al. 2017Clarke, Harold D., Matthew Goodwin, and Paul Whiteley 2017 Brexit: Why Britain Voted to Leave the European Union. Cambridge: Cambridge University Press. ), distrust of the ‘establishment’ (Abrams and Travaglino 2018Abrams, Dominic, and Giovanni A. Travaglino 2018 “Immigration, Political Trust, and Brexit-Testing an Aversion Amplification Hypothesis.” British Journal of Social Psychology 57 (2): 310–326. ), to the production of discourse, for instance the role of social media (Hänska and Bauchowitz 2017Hänska-Ahy, Max, and Stefan Bauchowitz 2017 Tweeting for Brexit: How Social Media Influenced the Referendum. In Brexit, Trump and the Media, ed. by J. Mair, T. Clark, N. Fowler, R. Snoddy, and R. Tait. 31–35. Suffolk: Abramis Academic Publishing.), and political conceptualisations of Brexit (Krzyżanowski 2018Krzyżanowski, Michał 2018 “Social Media in/and the Politics of the European Union: Politico-Organizational Communication, Institutional Cultures and Self-Inflicted Elitism.” Journal of Language and Politics 17 (2): 243–266. ). The protracted process of the UK withdrawing from the EU has caused much division across British society, and it has cut across traditional political dividing lines (Meredith and Richardson 2019Meredith, Joanne, and Emma Richardson 2019 “The Use of the Political Categories of Brexiter and Remainer in Online Comments about the EU referendum.” Community and Applied Social Psychology 29 (1): 43–55. ). However, how people actually discuss Brexit in everyday encounters is difficult to access (but see Demasi 2019Demasi, Mirko A. 2019 “Facts as Social Action in Political Debates about the European Union.” Political Psychology 40 (1): 3–20. ; Meredith and Richardson 2019Meredith, Joanne, and Emma Richardson 2019 “The Use of the Political Categories of Brexiter and Remainer in Online Comments about the EU referendum.” Community and Applied Social Psychology 29 (1): 43–55. ), yet it is crucial for a full understanding of Brexit as a socially divisive phenomenon.

The apparently common division surrounding Brexit (Bowman and West 2020Bowman, Jonathan W. P., and Keon West 2020 “Brexit: The Influence of Motivation to Respond without Prejudice, Willingness to Disagree, and Attitudes to Immigration.” British Journal of Social Psychology 60 (1): 222–247. ) and the political landscape has been reflected in people’s everyday talk, as reported hate crimes such as racism, homophobia, xenophobia and misogyny have soared (Clarke and Newman 2019Clarke, John, and Janet Newman 2019 “What’s the Subject? Brexit and Politics as Articulation.” Journal of Community & Applied Social Psychology 29 (1): 67–77. ). Meredith and Richardson (2019)Meredith, Joanne, and Emma Richardson 2019 “The Use of the Political Categories of Brexiter and Remainer in Online Comments about the EU referendum.” Community and Applied Social Psychology 29 (1): 43–55. collected comments on newspaper items and focused on what the political identities of ‘Brexiter’ and ‘Remainer’ mean to those who use them. They mapped those two opposing sides and documented how they sit together within the broader ‘voters in the Brexit referendum’ device, and are mutually exclusive opposites, with one being implicitly defined by using the other (Stokoe 2003Stokoe, Elizabeth H. 2003 “Mothers, Single Women and Sluts: Gender, Morality and Membership Categorization in Neighbour Disputes.” Feminism and Psychology 13 (3): 317–344. ; Leudar et al. 2004Leudar, Ivan, Victoria Marsland, and Jirí Nekvapil 2004 “On Membership Categorization: ‘Us’, ‘Them’, and ‘Doing Violence’ in Political Discourse.” Discourse & Society 15 (2–3): 243–266. ). Yet Meredith and Richardson (2019Meredith, Joanne, and Emma Richardson 2019 “The Use of the Political Categories of Brexiter and Remainer in Online Comments about the EU referendum.” Community and Applied Social Psychology 29 (1): 43–55. , 49) note that “for both Brexiters and Remainers […] it can be seen that the precise nature of these categories is contested”. The contestability of Brexit is also taken up in Demasi’s (2019)Demasi, Mirko A. 2019 “Facts as Social Action in Political Debates about the European Union.” Political Psychology 40 (1): 3–20. analysis of EU debates; he demonstrates various ways that ‘facts’ may be challenged, notably how context under which the ‘fact’ is being debated may be altered so that it may be contested. With the nature of categories and ‘facts’ shown to be contestable in talk about Brexit, it is worthwhile investigating how one alters their position to control the interaction. This is where the present paper seeks to contribute.

2.2Membership categorisation analysis

We build upon the toolkit of membership categorisation analysis (MCA) to explore how interactants – or ‘members’ – categorise themselves and others. MCA is an ethnomethodological approach to explore members’ reasoning practices and how members organise themselves and objects through their talk. MCA was first described by Eglin and Hester (1992)Eglin, Peter, and and Stephen Hester 1992 “Category, Predicate and Task: The Pragmatics of Practical Action.” Semiotica 88 (3–4): 243–268., who built on Sacks’ (1995)Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. original membership categorisation device analysis (see Francis and Hester 2017Francis, David, and Sally Hester 2017 “Stephen Hester on the Problem of Culturalism.” Journal of Pragmatics 118: 56–63. for a fuller account of the history of MCA). MCA has subsequently been established as an approach (e.g. Housley and Fitzgerald 2002Housley, William, and Richard Fitzgerald 2002 “The Reconsidered Model of Membership Categorization Analysis.” Qualitative Research 2 (1): 59–83. ; D’Hondt 2013D’Hondt, Sigurd 2013 “Analyzing Equivalences in Discourse: Are Discourse Theory and Membership Categorization Analysis Compatible?” Pragmatics 23 (3): 421–445. ) with the analytic power to show how culture is produced in action (Hester and Eglin 1997aHester, Stephen, and Peter Eglin 1997a Culture in Action: Studies in Membership Categorization Analysis. Washington, D.C.: International Institute for Ethnomethodology and Conversation Analysis and University Press of America.), and how people assemble the ‘who-we-are’ and ‘what-we’re-doing’ (Butler et al. 2009Butler, Carly W., Richard Fitzgerald, and Rod Gardner 2009 “Branching Out: Ethnomethodological Approaches to Communication.” Australian Journal of Communication 36 (3): 1–14.) in social interaction.

The central concepts of MCA are categories that can be assembled into collections, or membership categorisation devices (Sacks 1995Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. ). For instance, ‘family’ is a device comprising categories such as ‘mother’, ‘son’, ‘daughter’; or the device ‘positions toward Brexit’ consists of categories such as ‘Remainer’ and ‘Brexiteer’. The latter illustrates that devices are culturally sensitive aggregates, assembled in situ in a particular context for a particular purpose. Categories are selected by a speaker to be heard as part of whichever device is relevant to their present environment, and they are understood to be associated with certain qualities, or predicates. Sacks (1995)Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. illustrates this with an example from a child’s story: ‘The baby cried. The mommy picked it up’. To understand the action that is described here, the ‘baby’ and the ‘mommy’ are taken to be categories in the same device ‘family’, such that the mother is not any mother, but the mother of that particular baby. The action of ‘picking up a baby’ is thus a category-bound activity (Sacks 1995Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. ), as something that mothers expectedly do, and ‘being caring’ is a quality that mothers accountably have, a predicate of the category. Predicates are a broader take on Sacks’ (1995)Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. original concept, as they include not just activities, but also “rights, entitlements, obligations, knowledge, attributes and competencies” (Hester and Eglin 1997bHester, Stephen, and Peter Eglin 1997b “Membership Categorization Analysis: An Introduction.” In Culture in Action: Studies in Membership Categorization Analysis, ed. by Stephen Hester, and Peter Eglin. 1–23. Washington, D.C.: International Institute for Ethnomethodology and Conversation Analysis and University Press of America., 5).

Whilst devices are assembled locally and can shift during an encounter, it is possible that there is a so-called omnirelevant device that is always potentially applicable throughout an entire encounter. Sacks (1995)Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. discusses the example of a group therapy session, where at any point the participants can orient to this omnirelevant device and do a category-bound activity as a category member within that device, such as introduce a new group member. Omnirelevant devices were more fully defined by Fitzgerald et al. (2009)Fitzgerald, Richard, William Housley, and Carly W. Butler 2009 “Omnirelevance and Interactional Context.” Australian Journal of Communication 36 (3): 45–64. as those which operate at the organisational level and sometimes the immediate level of the interaction – for example, in a TV interview the categories of ‘interviewer’ and ‘interviewee’ have omnirelevance. This is an example of an institutional interaction where matters of who the participants are to one another are germane to how the business of the encounter is conducted. Relevant to the analysis is how categories and devices become applied, and the consequences of this.

The application of some categories to a population is known as partitioning (Sacks 1995Sacks, Harvey 1995 Lectures on Conversation. Oxford: Blackwell. ). Partitioning refers to how interlocutors establish a categorial basis (who-they-are) for performing certain actions and managing entitlement to speak about certain topics (Nishizaka 2021Nishizaka, Aug 2021 “Partitioning a Population in Agreement and Disagreement.” Journal of Pragmatics 172: 225–238. ). It regards how interlocutors may divide themselves into different categories (for instance, interviewer/interviewee to man/woman) to belong to different ‘territories of ownership’ (Raymond and Heritage 2006Raymond, Geoffrey, and John Heritage 2006 “The Epistemics of Social Relationships: Owning Grandchildren.” Language in Society 35 (5): 677–705. ) and leverage the associated domains of knowledge and responsibilities tied to those identities, such as to make an accusation of ‘mansplaining’ based on one being seen as a woman and not as an interviewee (Joyce et al. 2021Joyce, Jack B., Bogdana Humă, Hanna-Leena Ristimäki, Fabio Ferraz de Almeida, and Ann Doehring 2021 “Speaking out Against Everyday Sexism: Gender and Epistemics in Accusations of “Mansplaining”.” Feminism & Psychology 30 (4): 502–529. ).

In the example below, taken from the final moments of the encounter investigated in our analysis,11.Please see Joyce and Walz (2021)Joyce, Jack B., and Linda Walz 2021 Picking Fights with Politicians Dataset. for the full transcription of the encounter and URL to the recording. David Davies (DD) excludes BasedAmy (BA) from the interaction on the basis that she is not a speaker of Welsh, whereas he and the Interviewer (IR) are, and therefore he is able to escape the argument.

160 BA:  =you’re a MEMber of PARliament ^Fig. 1
161 DD:  yeah well I’m not your eM Pee am I.
162 BA:  thank god [thank god you’re not]
163 DD:            [Carry on  let’s     ] do a bit in welsh 
164 DD:  [gallaf ei anwybyddu     ] ^Fig. 2
164a     [I can just ignore it    ]
165 BA:  [so this man he claims he] vote Brexit right? he      
166      claims he voted Brexit but he signed the dea:l he 
167      voted for the deal that is a total 
168      [betra:yal of brexit] 
169 IR:  [excuse me          ]  
170      could you just [turn it down       ] 
171 BA:                 [( it’s a total d- )]
172      you know what I’ve hada ju- listen >no no no< 
173      listen you guys are a disgrace as well ↑how much 
174      have you been paid- how much have the Bee Bee Cee 
175      been paid t- to promote the European union 
176 DD:  gwnewch y cyfweliad yn Gymraeg
176a     Do the interview in welsh

At L163, David Davies instructs the BBC Interviewer to continue with the interview that was previously disrupted by BasedAmy and explicitly states “let’s do a bit in welsh”. Here and in the following lines (L164/164a, and later in L176/176a) David Davies orients to his and presumably the BBC Interviewer’s membership as Welsh speakers. This is coupled with a bodily position change to markedly change his physical orientation from BasedAmy and toward the BBC Interviewer (see Figures 1 and 2). So here, partitioning into different categories serves to escape BasedAmy’s turns by deliberately excluding her from being able to understand, which is antagonistic in itself (cf. Cromdal 2004Cromdal, Jakob 2004 “Building Bilingual Oppositions: Code-Switching in Children’s Disputes.” Language in Society 33 (1): 33–58. ). BasedAmy treats this as partitioning by, in the first instance, speaking to the recording, “so this man he claims” (L165) rather than responding to David Davies. Moreover, the BBC Interviewer asks BasedAmy to lower the volume of her speaker, “turn it down” (L170), to which BasedAmy can and does respond by targeting the BBC, “you guys are a disgrace” (L173).

The local invocation of the category ‘Welsh speaker’ partitions the interlocutors on the basis that they are or are not speakers of Welsh, and further overlays the device ‘legitimate participants’ to the interview. This restricts BasedAmy’s ability to directly respond to the turn beyond commenting on the talk itself (which she does). This example illustrates how the invocation of a new category and partitioning the population is done for the immediate purpose: to re-establish the omnirelevant ‘interview’ device and return the encounter to an interview by addressing a Welsh speaking audience. This example illustrates how partitioning manages one’s rights to speak. We return to partitioning in the analysis, but now move to a discussion of how membership categories have been explored in political encounters.

2.3Categories in political encounters

Political talk has been a fruitful site of research for MCA studies as spearheaded by Eglin and Hester (1999Eglin, Peter, and Stephen Hester 1999 ““You’re All a Bunch of Feminists”: Categorization and the Politics of Terror in the Montreal Massacre.” Human Studies 22 (2–4): 253–272. , 2003Eglin, Peter, and Stephen Hester 2003 The Montreal Massacre: A Story of Membership Categorization Analysis. Waterloo: Wilfrid Laurier University Press.) in their analyses of the Montreal massacre. MCA studies have followed significant social and political events to understand the pragmatic characteristics and interlocutors’ sense-making practices as a means of doing culture-in-action (Hester and Eglin 1997aHester, Stephen, and Peter Eglin 1997a Culture in Action: Studies in Membership Categorization Analysis. Washington, D.C.: International Institute for Ethnomethodology and Conversation Analysis and University Press of America.). Examples of such studies are a dissection of political commentary on the then imminent fox hunting ban in the UK (Housley 2002Housley, William 2002 “Moral Discrepancy and ‘Fudging the Issue’ in a Radio News Interview.” Sociology 36 (1): 5–21. ), and an analysis of public addresses made following the September 11th 2001 attacks (Leudar et al. 2004Leudar, Ivan, Victoria Marsland, and Jirí Nekvapil 2004 “On Membership Categorization: ‘Us’, ‘Them’, and ‘Doing Violence’ in Political Discourse.” Discourse & Society 15 (2–3): 243–266. ). The latter study reveals the intricate category work that establishes an ‘us’-‘them’ category pair united by opposition with individuals belonging to both ‘us’ or ‘them’ depending on how the conflict is framed. More recently, Meredith and Richardson (2019)Meredith, Joanne, and Emma Richardson 2019 “The Use of the Political Categories of Brexiter and Remainer in Online Comments about the EU referendum.” Community and Applied Social Psychology 29 (1): 43–55. discussed the precise nature of categories relating to the UK leaving the European Union.

Beyond the mapping of notable events, MCA studies of political talk endeavour to uncover how matters of cultural knowledge are leveraged (Fitzgerald and Thornborrow 2017Fitzgerald, Richard, and Joanna Thornborrow 2017 “ ‘I’m a Scouser’: Membership Categories and Political Geography in the 2016 UK Election Call Phone-in”. Journal of Language and Politics 16 (1): 40–58. ), how political accountability is accomplished or avoided (Housley 2002Housley, William 2002 “Moral Discrepancy and ‘Fudging the Issue’ in a Radio News Interview.” Sociology 36 (1): 5–21. ; Housley and Fitzgerald 2003Housley, William, and Richard Fitzgerald 2003 “Moral Discrepancy and Political Discourse: Accountability and the Allocation of Blame in a Political News Interview.” Sociological Research Online 8 (2): 18–26. ) and how interlocutors come to be seen as a victim (Clifton 2009Clifton, Jonathan 2009 “A Membership Categorization Analysis of the Waco Siege: Perpetrator-Victim Identity as a Moral Discrepancy Device for ‘Doing’ Subversion.” Sociological Research Online 14 (5): 38–48. ). These studies tackle moral discrepancies between intention and action, namely the distinction between what the authorities promised to do versus what they have or have not done, and how trust and accountability are accomplished.

Political discourse across a variety of contexts is suffused with categories, with categorisation forming the methods for formulating opposition (Robles 2011Robles, Jessica S. 2011 “Doing Disagreement in the House of Lords: ‘Talking around the Issue’ as a Context-Appropriate Argumentative Strategy.” Discourse & Communication 5 (2): 147–168. ), for characterising the nature of social and political events and actors, and for holding politicians accountable. These are, of course, interrelated, and as we will exemplify, they can be brought to bear throughout a single encounter. Indeed, Housley’s (2002)Housley, William 2002 “Moral Discrepancy and ‘Fudging the Issue’ in a Radio News Interview.” Sociology 36 (1): 5–21. analysis introduces the moral discrepancy device to see how a radio interviewer secures an account from a politician. Housley and Fitzgerald (2003)Housley, William, and Richard Fitzgerald 2003 “Moral Discrepancy and Political Discourse: Accountability and the Allocation of Blame in a Political News Interview.” Sociological Research Online 8 (2): 18–26. reproduce these findings in their analysis of a political TV news interview to show how an interviewer attempts to ascribe moral discrepancy and the interviewee evades that ascription. The crux of these studies is the discrepancy between category-bound and category-generated predicates of government agents. Jayyusi (1984)Jayyusi, Lena 1984 Categorization and the Moral Order. Routledge & Kegan Paul. distinguishes between these two, understanding bound features as conventionally occurring with a category, and generated features as connected to some category as an interactional accomplishment. Clifton (2009)Clifton, Jonathan 2009 “A Membership Categorization Analysis of the Waco Siege: Perpetrator-Victim Identity as a Moral Discrepancy Device for ‘Doing’ Subversion.” Sociological Research Online 14 (5): 38–48. clarifies this distinction of conventional occurring as an in situ achievement and not an a priori association. With these studies investigating the negotiation of categories in institutional environments, our analysis explores whether and how this applies in non-institutional political environments and what its relationship with antagonism looks like, in practice.

2.4Antagonism

The central phenomenon explored in this paper is how a citizen ‘picks a fight’ with a politician, this is to say how an interlocutor antagonises a co-interlocutor through linguistic means. Despite much research which describes ‘antagonism’, there is no consensus on what features constitute ‘antagonism’. Generally, antagonism is understood as an interactional achievement that describes how an individual is hostile toward another (Dersley and Wootton 2000Dersley, Ian, and Anthony Wootton 2000 “Complaint Sequences within Antagonistic Argument.” Research on Language and Social Interaction 33: 375–406. ). Antagonism is a central aspect within an activity such as disputing, which O’Driscoll and Jeffries (2019O’Driscoll, Jim, and Leslie Jeffries 2019 “Introduction: The Origins of The Routledge Handbook of Language in Conflict.” In The Routledge Handbook of Language in Conflict, ed. by M. Evans, L. Jeffries, and J. O’Driscoll. 1–9. New York: Routledge. , 4–5) define as “any situation or behaviour involving parties (individuals or groups) who are, or consider themselves to be, instrumentally, intellectually and/or emotionally opposed”. Indeed, Housley et al. (2017Housley, William, Helena Webb, Adam Edwards, Rob Procter, and Marina Jirotka 2017 “Digitizing Sacks? Approaching Social Media as Data.” Qualitative Research 17 (6): 627–644. , 2) describe antagonism as being achieved through speech practices that “elicit oppositional and relational responses”. We may also draw parallels between antagonism and the social media phenomenon of trolling, the latter defined by Hardaker (2010Hardaker, Claire 2010 “Trolling in Asynchronous Computer-Mediated Communication: From User Discussions to Academic Definitions.” Journal of Politeness Research 6 (2): 215–242. , 237) as a user “whose real intention(s) is/are to cause disruption and/or trigger or exacerbate conflict for the purposes of their own amusement”. Building on these definitions, we understand antagonism as comprised of a range of linguistic practices which on the surface may not appear oppositional, but over the course of an interaction take a position which opposes the interlocutors and entices them to respond in a way that would be damaging to themselves.

Previous studies have explored how interactants exert control over each other in sequences of talk. First, Robles and Castor (2019)Robles, Jessica S., and Theresa Castor 2019 “Taking the Moral High Ground: Practices for Being Uncompromisingly Principled.” Journal of Pragmatics 141: 115–129. explore the consequences of taking a stand in their discourse analysis of a political issue in the US: issuing marriage licenses to same-sex couples. Robles and Castor deconstruct an unyielding dispute to show how accounts for one’s position are withheld, and how accounts are demanded in ways that presuppose no reasonable account exists. These function to accomplish incompatible ideological positions and to cooperate in the achievement of conflict. In the same vein, Reynolds’ (2011Reynolds, Edward 2011 “Enticing a Challengeable in Arguments: Sequence, Epistemics and Preference Organisation.” Pragmatics 21 (3): 411–430. , 2015Reynolds, Edward 2015 “How Participants in Arguments Challenge the Normative Position of an Opponent.” Discourse Studies 17 (3): 299–316. ) research on enticing a challengeable demonstrates the power that certain interactional practices hold in conflict. Reynolds presents cases where control over a course of action is won via fabricating a challenge against an opponent.

In certain environments the features of antagonistic speech are sanctioned, yet in others they may be rewarded. Research on adversarial talk in UK politics has explored the format of Prime Minister’s Questions, a weekly exchange between MPs and the Prime Minister, where adversarial positions are taken and needless antagonism is sanctioned (Bull and Strawson 2019Bull, Peter, and Will Strawson 2019 “Can’t Answer? Won’t Answer? An Analysis of Equivocal Responses by Theresa May in Prime Minister’s Questions.” Parliamentary Affairs 73 (2): 429–449. ). In that environment the interlocutors have a symmetrical relationship (politician–politician) with formal opportunities to strike and counterstrike, or as Goffman (1967Goffman, Erving 1967 Interaction Ritual: Essays in Face-to-face Behaviour. Chicago: Aldine., 25) describes, to score a ‘point’ at the other’s expense. Traditionally, opportunities to score points against politicians have been restricted to these formal rituals. In recent times, however, the distance between the public and politicians has decreased, for instance through social media, affording opportunities for citizens to entice politicians and antagonise them.

Direct encounters between citizens and politicians have been investigated on social media (Housley et al. 2017Housley, William, Helena Webb, Adam Edwards, Rob Procter, and Marina Jirotka 2017 “Digitizing Sacks? Approaching Social Media as Data.” Qualitative Research 17 (6): 627–644. ), the constituency office (Hofstetter and Stokoe 2018Hofstetter, Emily, and Elizabeth H. Stokoe 2018 “Getting Service at the Constituency Office: Analyzing Citizens’ Encounters with their Member of Parliament.” Text & Talk 38 (5): 551–573. ) or in public debates (van Schepen 2019Van Schepen, Nynke 2019 “Political Transparency Matters: Citizens Challenging Officials via ‘Have You Planned X’-Type Questions.” Discourse & Society 30 (5): 521–535. ). Except for social media encounters, these interactions tend to be governed by some institutional rules restricting how citizens may challenge politicians. For example, in their analysis of public plenary consultation meetings for urban planning, van Schepen (2019)Van Schepen, Nynke 2019 “Political Transparency Matters: Citizens Challenging Officials via ‘Have You Planned X’-Type Questions.” Discourse & Society 30 (5): 521–535. demonstrates how, when a citizen is invited to ask a question, using a confirmatory interrogative at a politician challenges that politician on the basis that they have so far failed to consider some issue. To illustrate this with an extract from our data, “have you read the [Brexit] deal?” (Extract 7, L76) implies that the politician has failed to read the Brexit withdrawal deal and challenges them for it. Yet the data explored in our analysis is different in that it features a unique encounter between an MP and a citizen in a public setting not governed by institutional rules, and which quickly turns antagonistic, as we show.

4.1Establishing omnirelevant devices, categories and their predicates

In these first examples we see how the interlocutors establish omnirelevant devices, categories and their predicates – this is where BasedAmy positions herself as in opposition so to be antagonistic. BasedAmy takes a pro-Brexit position (known colloquially as being a Brexiteer) and challenges David Davies on that basis. However, David Davies is also a Brexiteer, which BasedAmy is ostensibly not aware of initially. The first extract shows the moment that BasedAmy’s talk with her online audience and David Davies’ interview merge.

01 IR:  How does i:t #ehh >make you feel
02 DD:  well obviously what I will do is eh record
03      it so I’m actually going to turn my camera
04      on now to record (.) this la:dy and eh I will
05      put (.) anythin[g abusive on eh onto twitter           ]
06 BA:                 [what did I do I just stood here filming]
07 DD:  because [eh I think it’s- (.)  ]
08 IR:          [erm unfortunately (  )]
09 DD:  I think it’s unacceptable
10 BA:  why’s he feeling intimidate- well you’ve got a
11      bloody great camera there on him for gods sake
12      what’s the difference of my phone.
13       [hhuh]
14 IR:  w[ell ] cause when you’re tal:king-
15 BA:  °I w[as quiet and yous guys just stopped talking°
16 IR:      [there’s two: loud speakers basically blaring out
17 BA:  when I kept-

This extract represents the initial moments of the encounter, with the BBC interviewer asking David Davies how it feels to be video-recorded and challenged by a citizen. As BasedAmy approaches with her camera and loudspeaker, she is treated as an antagoniser, “this la:dy and eh I will put (.) anything abusive on eh on to twitter” (L04–05). Moreover, BasedAmy, in overlap, claims innocence before the characterisation of ‘abuse’ is made, “what did I do” (L06), contesting David Davies’ characterisation by treating it as unwarranted given the ordinary, reasonable interpretation of her activity “just stood here filming” (L06). This is later amplified, “you’ve got a bloody great camera there” (L10–11), setting up the two competing operative devices: for David Davies, this is addressing the camera and giving an interview. For BasedAmy, in contrast, describing the event through the technological categorisation of the ‘great camera’ introduces the distinction ‘professional/amateur’, and thus she can be seen as an ordinary member of the public in public space legitimately overhearing the interview. In this sense, setting up these two competing omnirelevant devices ‘interviewer-interviewee’ (wherein BasedAmy would feature as a disrupter) and ‘politician-citizen’ to configure who-they-are and what-they-are-doing facilitates a possible confrontation.

This highlights one way of achieving antagonism: explicit establishment of categories and their predicates. ‘Innocence’ is evoked as a quality that BasedAmy has, as the ‘innocent’ party engaged in a lawful activity. In contrast to herself, BasedAmy frames David Davies as guilty of being needlessly confrontational towards a citizen, which would be inappropriate for his role as MP. David Davies refutes BasedAmy’s claim of innocence, “I think it’s unacceptable” (L09), but BasedAmy reformulates this refutation as a feeling “why’s he feeling intimidate-” (L10), drawing on David Davies’ own formulation of ‘acts of minor intimidation’ prior to this extract. Consequently, BasedAmy ironicises that feeling ascription, “what’s the difference of my phone” (L12), portraying her filming of David Davies, who is already being filmed by the BBC crew, as not materially different, despite the institutional purpose of the BBC recording.

The interviewer answers BasedAmy, identifying the trouble with her assertion “cause when you’re tal:king” (L14) to pass blame for the disruption onto BasedAmy. However, BasedAmy directly rejects that turn, “I was quiet” (L15), so that the contestation of what caused the confrontation is actually progressing the confrontation – who the ‘guilty party’ is and how their behaviour provides evidence for this is a matter for categorial negotiation.

The following extract continues with BasedAmy repeating her assertion as a correction to the version of events which is transpiring, and accounting for her intentions, “>I was just gonna< listen to what he was saying” (L19) notably refusing responsibility for causing any disruption to the interview that was being conducted. She then pivots to antagonising David Davies based on his position toward Brexit.

18 BA:  I was quiet (.) right when I came here I was quiet
19      >I was just gonna< listen to what he was saying.
20      but now I can tell he’s a remoaner (.) he must be
21      a rem- did you vote remain:n
22 DD:  act[ually I’m on-]
23 IR:     [do you fancy ]
24      [not do- do you  ]
25 BA:  [are you an eMPee]
26 DD:  I [am an eMPee         ]
27 IR:    [fancy not doing that]
28 BA:  are you an eMPee did you vote to remain
29 DD:  no I voted to leave actually.

Extract 3 begins with BasedAmy trying something different to antagonise David Davies, “now I can tell he’s a remoaner” (L20). BasedAmy here is seeking to establish category membership of them both, thus who-they-are and what-they’re-doing (Butler et al. 2009Butler, Carly W., Richard Fitzgerald, and Rod Gardner 2009 “Branching Out: Ethnomethodological Approaches to Communication.” Australian Journal of Communication 36 (3): 1–14.): whether he is an MP and what political position he holds. The use of ‘remoaner’ is antagonistic, as it is not simply about voting remain, but it introduces particular predicates for people who actively complain about the referendum result and are positioned as unable to accept the democratic process. BasedAmy treats David Davies’ membership of ‘remoaner’ as obvious (“he must be”, L20), but the contestable nature of ‘remoaner’ leads to BasedAmy seeking confirmation of a predicate which would definitively allocate David Davies as a ‘remoaner’: “did you vote remai:n” (L21). David Davies indicates some contrast to BasedAmy’s assertion “actually I’m-” (L22), but both the BBC interviewer and BasedAmy interject. BasedAmy takes up David Davies’ possible turn for an answer with a new line of questioning, “are you an eMPee” (L28), pushing forward her antagonistic trajectory with these polar questions which prefer short responses that David Davies cannot reasonably refuse.

In these first two extracts which constitute the initial part of the encounter, BasedAmy has disrupted an ongoing interview. This disruption occurs as she antagonises David Davies and the BBC crew in two ways that both draw on categorial resources: firstly, she negotiates what is (un)acceptable behaviour for an encounter between an MP and citizen, implying that an MP should not be intimidated by a citizen who is ostensibly not disrupting, but just listening. BasedAmy thereby claims innocence and lays the blame for the disruption onto David Davies and the crew. Secondly, she antagonises David Davies based on his presumed category membership as a remain supporter. Her use of the term ‘remoaner’ (L20) at the same time positions her as a leave supporter who disapproves of people who she sees as ‘bemoaning’ the outcome of the referendum. Ascribing – and criticising – opposing category membership in the invoked ‘Brexit’ device here serves BasedAmy as a means of pursuing her antagonistic project.

Yet this fails at the end of Extract 3, when it transpires that they are both leave voters and as such co-members within this device: “no I voted to leave actually” (L29). David Davies’ turn-final ‘actually’ acts as counter-positional to BasedAmy’s dividing of the ‘Brexit’ device. This failure to produce herself and David Davies as members of opposing categories could possibly close this antagonistic sequence, with them both agreeing. Instead, BasedAmy engineers further antagonism by challenging David Davies on the basis of their category membership.

30 BA:  well I don’t know. >why are you acting like a
31      snowflake then< (.) do[you know what a snowflake?  ]
32 DD:                        [I’m not acting like a snowfl]ake
33      but I-
34 BA:  weh you are cause you got a >little bit
35      intimidated< by me I’m jus-
36 DD:  I’m not intimidated by you=
37 BA:  =we[ll you said it]
38 DD:     [but you might ] make a: false eh:
39      allegati[on agains]t me
40 BA:          [of ↑whhat]
41 DD:  well I don’t know so um [I’m recording]
42 BA:                          [have you done]
43      something to be accused o↑f?
44      (0.4)

Here, BasedAmy initiates a new way of antagonising David Davies, “>why are you acting like a snowflake22.‘Snowflake’ is a derogatory term that describes a person who is easily offended or overly sensitive. then< (.) do you know what a snowflake” (L30–31). This turn is designed in two parts: firstly as an insult to David Davies that he is ‘making a fuss’ and is easily offended, and secondly as an interrogative as to whether he knows what the term means. Note also the shift to a ‘why’ question and its preference for a more extended response. David Davies responds in overlap with the second part “I’m not acting like a snowflake” (L32) to contest his membership of ‘snowflake’. Moreover, his rejection marks this as a dispute about what it means to be a ‘snowflake’. Indeed, BasedAmy’s assertion “you got a >little bit intimidated< by me” (L34–35) describes what predicate of David Davies’ ostensible ‘snowflake’ membership he is furnishing: that being easily intimidated is a predicate of the category. BasedAmy’s turn is neatly produced before David Davies can claim further speakership and is confirmatory with “weh you are” (L34), and upgrades her claim that David Davies is intimidated with “by me” (L35). In this sense, it is not only that David Davies is intimidated, but that he is specifically intimidated by BasedAmy, who through her physical actions and talk is claiming membership as an ordinary citizen, as suggested by the following minimisation, “I’m jus-” (L35). This antagonises on the basis that he might reject the stigmatised label, but not the basis for the accusation (Joyce et al. 2021Joyce, Jack B., Bogdana Humă, Hanna-Leena Ristimäki, Fabio Ferraz de Almeida, and Ann Doehring 2021 “Speaking out Against Everyday Sexism: Gender and Epistemics in Accusations of “Mansplaining”.” Feminism & Psychology 30 (4): 502–529. ), namely that he is intimidated.

David Davies responds in kind, “you might make a: false eh: allegation against me” (L38–39). Yet BasedAmy handles David Davies’ response by neither accepting nor rejecting this characterisation of her as someone who would make false allegations, but instead she reverses the inference of his turn. The reversal “have you done something to be accused of?” (L42–43) puts David Davies in a position of justifying himself. Indeed, his next turn does not follow immediately (see L44), and when it is produced, it does not respond to BasedAmy’s question, but seeks to conclude the altercation and move back to the interview, as seen in Extract 5.

These first extracts illustrate one way that contributes to the achievement of antagonism: establishing categories and their predicates, and in the process introducing competing omnirelevant devices. BasedAmy positions herself as an ordinary citizen who is lawfully filming, thereby characterising David Davies’ apparent offence as him being intimidated. The aim of establishing predicates in these terms is to position oneself and one’s interlocutor in opposing categories and to control which omnirelevant device is operative. Here, establishing what features they possess contributes to the encounter being antagonistic as, for example, it lays the blame for the current confrontation onto David Davies for unreasonably claiming offence. In the latter two extracts we can also observe the repeated questioning – not as queries needing answers, but as a vehicle to keep the interaction moving. The following section explores another strategy BasedAmy employs to sustain her antagonism: she challenges David Davies’ position as a Brexiteer and thus his category membership.

4.2Challenging category membership

45 DD:  so anyway that’s [the-
46 BA:                   [huhhuh
47 DD:  that’s the kind of- funnily enough I think this
48      this lady is actually on the brexit si:de which
49      is my si[de   of  the  arg]ument.
50 IR:          [think she is yeah]
51 DD:  but basically it’s the same-
52 BA:  well you’re not acting like a brexiteer
53      °you’re being very-°
54 DD:  [how do brexiteers act]
55 BA:  [no you see actually  ]
56      >no honestly< if you are a brexiteer then this is
57       how,- this is what brexit[eers have b’n-      ]

Extract 5 begins with David Davies pivoting away from BasedAmy’s question (Extract 4 L42–43 “have you done something to be accused of?”) with “so anyway that’s the- that’s the kind of- funnily enough I think this this lady” (L45–47). His turn responds to BasedAmy’s line of argument but not directly to BasedAmy by moving to a meta level to talk about her (“this lady”) to a different audience. Indeed, such ‘going meta’ is a method through which speakers are able to avoid the content of the prior turn and speak as an ordinary person by pointing at some non-ordinary conduct (Clift and Pino 2020Clift, Rebecca, and Marco Pino 2020 “Turning the Tables: Objecting to Conduct in Conflict Talk.” Research on Language and Social Interaction 53 (4): 463–480. ).

Not only does David Davies’ meta turn handle the antagonistic question, it also underlines the struggle between the two operative membership devices by addressing the overhearing audience while still talking about BasedAmy. The turn seeks to bring the interaction to a close by drawing on their shared membership “on the brexit si:de which is my side of the argument” (L48–49). BasedAmy’s pursuit of antagonism occurs in L52, blocking the closing attempt by challenging the legitimacy with which David Davies can claim to be a member of the category ‘Brexiteer’. This is built as an accusation of him as not displaying the right predicates, “you’re not acting like a brexiteer” (L52). David Davies’ retort “how do brexiteers act” (L54) exposes BasedAmy’s challenge as empty because it invites BasedAmy to identify specific actions and behaviour which, once listed, would make her argument vulnerable because any of these would be contestable and deniable by David Davies.

At this point, the tables are turned, and it is David Davies who produces the next antagonising move, as evident in Extract 6.

58 DD:                            [can I just say (   )]
59      cause >I actually-< [hang on a minute-      ]
60 BA:                      [(   ) I’m still talking]
61 DD:  I campaigned for Brexit where were you?
62      >where were you people< when I was out campaign-
63      for brexit. You were nowhere_ you behind your
64      keyboards and now you’ve come out right- you are not a
65      brexiteer=
66 BA:  =you obviously haven’t watched [my f]ootage then
67 DD:                                 [no- ]
68 BA:  [before you say anymore-   ]
69 DD:  [I- I actually was campaign]ing for Brexit
70      and have been for years so I don’t need to
71      be- given lectures by people like you.

Here David Davies challenges BasedAmy’s membership of ‘Brexiteer’ (and her rights to challenge). His category membership is emphasised based on his past behaviour, “I campaigned for Brexit” (L61). His claim of being an authentic member of ‘Brexiteer’ rests on historicity (Coupland 2003Coupland, Nikolas 2003 “Sociolinguistic Authenticities.” Journal of Sociolinguistics 7 (3): 417–431. ), on displaying category predicates over a period of time (Widdicombe and Wooffitt 1990Widdicombe, Sue, and Robin Wooffitt 1990 “ ‘Being’ versus ‘Doing’ Punk: On Achieving Authenticity as a Member.” Journal of Language and Social Psychology 9 (4): 257–277. ; Walz and Fitzgerald 2020Walz, Linda, and Richard Fitzgerald 2020 “A Stranger in a Foreign Land: Identity Transition in Blogs about Transnational Relocation.” Discourse, Context & Media 36: 1–9. ). Furthermore, he categorises BasedAmy as “you people” (L62), which is subsumed with the category ‘keyboard warrior’ (L64), and challenges BasedAmy’s membership as an authentic leave voter based on her inaction, “you were nowhere” (L63), during the Brexit campaign. BasedAmy defends herself by rejecting his assertion with “haven’t watched my footage then” (L66), but this does not prevent David Davies from reusing that category, “people like you” (L71), to dismiss the legitimacy with which BasedAmy can claim to be a proper Brexiteer. At this moment, by David Davies orienting to BasedAmy’s omnirelevant device (as politician-citizen) rather than maintaining his own (interviewer-interviewee), he loses – the operative device has shifted, fulfilling BasedAmy’s interactional task.

In Extracts 5 and 6, we can observe challenges to category membership, namely, whether an interlocutor is an authentic member of the category to which they purportedly belong. After discovering that they share membership of both being Brexiteers, BasedAmy challenges David Davies’ membership for not acting how a Brexiteer ought to act. Similarly, David Davies challenges BasedAmy’s membership as inauthentic. Their respective challenges to membership seek to discredit one another’s position and thus their rights to argue about Brexit, despite their apparently shared membership. The direct challenges to their membership do not allow the encounter to be anything other than antagonistic – they cannot reconcile if they disagree about who they are.

4.3Creating ingroup and outgroup through partitioning

The following extracts show partitioning, which functions as a third way of achieving antagonism. After discovering that they both share membership as leave voters, and the challenges to discredit one another’s membership, BasedAmy and David Davies move to a different strategy: separating one another such that, despite being leave voters, they appear as members of different types of leave voters.

72 BA:  did you- did you vote for the dea:l.
73 DD:  I did vote for the deal but.
74 BA:  Y’↑O VOTED FOR THE DE↑AL
75 DD:  ye[a↓h]
76 BA:    [OH ] MY GOD and tha- have you read the deal?
77 DD:  yeah I read the deal=
78 BA:  =and it means not to leave=
79 DD:  =and how many of eM Pees how many- [yeah  ]
80 BA:                                     [grea↑t]
81      does it- it means we do↓n’t leave=
82 DD:  =how many eM Pees=
83 BA:  =so you’re a liar you did not vote to leave.
84 DD:  °there we are then° (.) [I tell you what             ]
85 BA:                          [>>you didn’t vote to leave<<]
86 DD:  people like you- peo[ple like you al ]most make me
87 BA:                      [have you read it]
88 DD:  want to join the [eEyU again to be honest] with you
89 BA:                   [see I’ve read?         ] (.)
90      the eh withdrawal agreement unlike this man
91 DD:  [yeah ]
92 BA:  [who’s] an eM Pee and shame-
93      what is your name then if you’re an eM Pee
94 DD:  never mind what my name is I’m not talking to you

Instead of engaging with David Davies’ history of campaigning for Brexit, BasedAmy separates them into opposing categories, “did you vote for the dea:l” (L72). Her turn partitions them within the ‘Brexit’ device: Brexiteers who voted for the government’s deal (then Prime Minister Theresa May’s withdrawal agreement, which would keep the UK, to a certain extent, tied to the EU) and those Brexiteers who prefer a ‘no deal Brexit’ and thus a severing of all ties between the UK and EU. Later, voting for the deal is treated as a predicate of people who are Remainers (“it means not to leave”, L78). Dividing Brexiteers into these two populations turns the category of ‘Brexit voter’, to which they both belong, into a device, now with the opposing categories ‘people who voted for the deal’ and ‘people who did not’. This affords BasedAmy the opportunity to challenge David Davies as an inauthentic member due to his behaviour being incompatible with the predicates of Brexiteers, namely that ‘authentic’ Brexiteers would not have voted for the deal.

This reconfiguration of their relationship as (in)authentic Brexiteers respectively enables BasedAmy to categorise David Davies in terms of his character, “you’re a liar” (L83), and to challenge his membership as a Brexiteer outright, “you did not vote to leave” (L83). In fact, BasedAmy is conflating two different stages in the political process here: the initial referendum to leave or remain in which citizens could participate, and the vote for or against Prime Minister Theresa May’s ‘leaving deal’ in parliament, in which only members of parliament and not citizens such as BasedAmy could participate.

Moreover, BasedAmy works to further separate herself from David Davies into people with opinions on the deal who have either read it or have not read it, “have you read the deal?” (L76). This suspends the presupposition that David Davies has read the deal to vote on it – preferring a ‘no’, otherwise it needn’t be asked. This division between who BasedAmy and David Davies are to each other affords the subsequent challenge done by BasedAmy, “see I’ve read? (.) the eh withdrawal agreement unlike this man” (L89–90), completely ignoring his previous response, “yeah I read the deal” (L77). Partitioning Brexiteers in this manner antagonises David Davies: reading the deal is built as a predicate of a good MP, and insinuating that he has not, “and shame-” (L92), undermines the validity of David Davies’ opinion (cf. Joyce et al. 2021Joyce, Jack B., Bogdana Humă, Hanna-Leena Ristimäki, Fabio Ferraz de Almeida, and Ann Doehring 2021 “Speaking out Against Everyday Sexism: Gender and Epistemics in Accusations of “Mansplaining”.” Feminism & Psychology 30 (4): 502–529. ). As soon as blame is allotted, even if untrue, it is available for everyone – so the moment that BasedAmy says that David Davies has not read the deal, the damage to his reputation is done. Interestingly, this trajectory is paused as BasedAmy attempts to discover David Davies’ name (L93), presumably to allow her to further antagonise him as she would be able to target him as an individual rather than just as an MP. This is met with David Davies’ refusal to talk to BasedAmy any further (L94).

Dividing herself and David Davies within the ‘Brexiteer’ device enables BasedAmy to ignore their alliance as leave voters and instead emphasise their differences, ultimately positioning David Davies as not true to his political commitments. The achievement of antagonism in this case is done by BasedAmy producing David Davies as a fraudulent leave voter. This is neatly built over several turns that by way of partitioning create the dichotomy of proper leave voters (herself included) and improper leave voters, thereby targeting ostensible qualities of MPs, namely that they have integrity and represent people’s will.

99  BA:  =so he voted for the deal and this £guy’s
100      pretending like he’s a l- leaver (.) what a li↑:ar£
101      you’re a liar °shame on you° you’re a traitor (.)
102      you signed that (.) document- that is nothing-
103      [it’s not even- you can’t even say it’s leav-]
104 DD:  [you see that’s- that’s what eh              ]
105      that’s what you put up with when you’re out here
106      all the time
107 BA:  oh d[ear ahh (.) have you h-]
108 DD:      [(.                 ) th]ere’s another one
109      over here now
110 BA:  ohh look- I mean this guy- I think he picked
111      the wrong job if you can’t cope with >somebody
112      walking along with a-<
113 DD:  I can cope with it I can sit here a:ll
114      da::y to be honest with you= ^Fig. 4

Here the partitioning is successful: by making a distinction between people in favour or against the Brexit deal, BasedAmy can now say that David Davies is not a leave voter – nicely returning this to being a challenge of membership. This occurs through “this £guy’s pretending like he’s a l- leaver” (L99–100). Now that BasedAmy has established that David Davies is not a true Brexiteer (based on her earlier partitioning), she addresses the audience to which she is live streaming, “this guy- I think he picked the wrong job” (L110–111). This antagonises David Davies in two ways: firstly, by chastising him for not being an adequate MP. That is, as an MP he ought to be able to cope with, “somebody walking along” (L111–112) – a low responsibility for what is required of being an MP. This follows a similar line to the earlier challenge of him not having read the deal and thus not acting responsibly as an MP. Secondly, and returning to our first strategy of antagonism, BasedAmy has successfully enticed David Davies into her omnirelevant device by positioning herself as ordinary (i.e. not blame-worthy), and David Davies as being unreasonably confrontational – reversing what is actually happening.

This sequence preserves the ongoing attempts to achieve antagonism and provoke David Davies into a confrontation. BasedAmy challenges David Davies’ category membership as both an authentic Brexiteer and as an MP. At this point David Davies begins to disengage from the conversation. He sits down (Figure 4) and reorients to the omnirelevance of the interview by partitioning himself and the BBC crew from BasedAmy, on the basis that they speak different languages, as shown in Extract 1.