Humiliating and dividing the nation in the British pro-Brexit press: a corpus-assisted analysis

ABSTRACT Since the United Kingdom’s referendum on European Union (EU) membership in 2016, a new political cleavage of Remainers and Leavers has developed (Kelley, N. [2019]. British social attitudes survey: Britain’s shifting identities and attitudes. (36). National Centre for Research). This paper explores how five pro-Brexit newspapers discursively construct political division in Britain in relation to two key events in the final year of Britain’s EU membership: the extension of the withdrawal process past the original date of March, and the introduction of the Benn Act in September. The paper reveals two primary discursive constructions of division in Britain: a divide between incompetent and arrogant political officials and an innocent, suffering public, and an identity cleavage between pro-Remain ‘elites’ and ‘ordinary’ Leave-voting citizens. The study argues that the construction of these divisions threatens a collective national identity in Britain at a time when it is most required. It concludes that by apportioning blame for socio-political divisions, the newspapers obfuscate their role in contributing to disunity in the UK.


Introduction
In 2013, David Cameron announced the possibility of a British referendum on membership of the European Union (EU). In his so-called Bloomberg speech, he claimed that Britain had the 'character of an island nation' which rendered it distinct from its mainland European counterparts (Cameron, 2013, n.p.). This initial speech made clear that Britain's withdrawal from the EU would be tied to questions of British identity (Wenzl, 2019(Wenzl, , 2020. It is not surprising, then, that in the seven years between Cameron's speech and Britain's withdrawal from the EU in January 2020, Britishness has been repeatedly ideologically contested. That there have been disagreements over what Britain stands for is evidenced by the construction of Britain as divided in national newspapers the day after the 23 June 2016 referendum (Koller & Ryan, 2019). The Telegraph Online went so far as to claim that 'there is no shared understanding of what the country is or should be' (Samuel, 2016, n.p.). Although leaving the EU was pitted as the solution to Britain's problems by Leave campaigners (Charteris-Black, 2019), the referendum was just the beginning of a period of national introspection. This paper explores how pro-Brexit British newspapers discursively construct Britain as humiliated and divided in relation to two key events in the Brexit timeline: the original withdrawal date of March 2019, and the introduction of the Benn Act in September 2019. Through key semantic domain analysis (Rayson, 2008), the paper uncovers a populist rhetoric of disunity between a national public and politicians, and a political cleavage between Remain and Leave voters. The article contends that by constructing socio-political divisions, and thereby exacerbating and reinforcing political cleavages, the newspapers threaten to create a deep-rooted sense of indignation that potentially destabilises a collective national identity. The study concludes that by blaming politicians and Remain voters, the newspapers obfuscate their role in contributing to socio-political disunity in the UK.

Brexit: a tumultuous process
On 24 June 2016, the British public awoke to the news that by a slight margin, the country had voted to leave the EU. The UK's nations were divided: while Wales and England voted to leave, Scotland and Northern Ireland opted to remain. As the UK began a process that had yet to be attempted, Scotland's First Minister Nicola Sturgeon declared that Brexit could precipitate another Scottish independence referendum. Following the referendum result, and contrary to his promise to remain in office, prime minister David Cameron resigned. He was succeeded in July 2016 by Home Secretary Theresa May.
May's Brexit plans, which included a controversial Irish border 'backstop', were set back in January 2019 when the Government suffered a substantial defeat on a meaningful vote on her Brexit deal. She presented her 'Plan B' deal to the House of Commons on 29 January 2019 but lost again when Members of Parliament (MPs) voted against the deal on 12 March. An extension to the withdrawal process became inevitable. The first extension was announced on 15 March 2019; a day later, May wrote in the Sunday Telegraph that it would be patriotic for MPs to vote in favour of her deal. Despite her pleas for patriotism, the Prime Minister requested an extension to Article 50. On 10 April, the UK and EU27 agreed that extension of Article 50 would last until 30 October 2019, with an earlier withdrawal if both sides ratified an agreement before then. With this extension guaranteed, May announced her resignation. In her departing speech, she told the public it would 'always remain a matter of deep regret' that she had 'not been able to deliver Brexit' (May, 2019, n.p.). After winning the Conservative Party leadership race in July 2019, Boris Johnson became the UK's third prime minister since the EU referendum was announced. Johnson generated controversy when he sought to prorogue Parliament to avoid extending Brexit, a decision that was ruled unlawful by the Supreme Court.
Like May, Johnson faced challenges on the road to delivering Brexit. One significant obstacle was Labour MP Hillary Benn's Bill, presented on 2 September 2019. The Bill, which became law on 9 September, established two deadlines. It set 19 October 2019 as the date by which Johnson would have to pass a Withdrawal Agreement, convince MPs to agree to a no-deal Brexit, or request another extension from the EU. If the latter became a reality, the proposed deadline for leaving the EU would be 31 January 2020. Independent MP Sir Oliver Letwin filed an amendment to withhold MPs' approval of Johnson's Brexit deal until the legislation to enact it was passed. This amendment would automatically trigger the Benn Act, forcing the Prime Minister to request an extension to Brexit until 31 January 2020. On 19 October 2019, the deadline the Benn Act had established, the Letwin amendment passed by 322 votes to 306. The same day, the Prime Minister wrote to then President of the European Council, Donald Tusk, to formally request an extension. He made clear that he did not want one, but the European Council granted an extension until 31 January 2020. Johnson then called a General Election for 12 December 2019, which resulted in a landslide victory for the Conservative Party. With a majority of 80 seats, Johnson was able to convince MPs to agree to a revised Withdrawal Agreement, and at 11pm on 31 January, the UK withdrew from the EU. A transition period began, which ended on 1 January 2021.
As this condensed and necessarily incomplete overview indicates, the Brexit process was characterised by disagreements over the future of the UK. This paper will explore how delays to Brexit caused by political disagreements influenced pro-Brexit newspaper coverage of British identity and incited socio-political division in the UK.

The nation and national identities
The nation is an imagined community (Anderson, 2006), 'a mental [and political] construct that is real only insofar as individuals believe in and identify with it' (Wenzl, 2020, p. 73). The imagined national community is constructed, negotiated, and renegotiated through discourse (Billig, 1995). By discourse, I mean 'linguistic and non-linguistic social practices and ideological assumptions' (Flowerdew, 2011, p. 178). National discourse is underpinned by a sense of timelessness, based on the collective memory of a shared history, the elaboration of a shared present, and the vision of a collective future (Bennett, 2018). As nation-building is a discursive act, there is no single, undisputed national frame: there are multiple, competing, constructions of the nation that differ depending on the discourse producer and production context (Billig, 1995). A national community is reified through institutional discourses and disseminated through, inter alia, mass communication and education (Wodak et al., 2009).
Just as the nation is socially and discursively constructed, so too are identities. Identities answer the question of who we are, both to each other and to ourselves (Benwell & Stokoe, 2006). A person can be many things at once; identities are multiple, contextdependent and constituted in discourse (Benwell & Stokoe, 2006). Identity construction is an ongoing discursive process of negotiation between multiple selves and others. National identities are a form of habitus, 'a common complex of ideas, concepts or perception schemes' that are intersubjectively shared and internalised through socialisation (De Cillia et al., 1999, p. 153). In other words, they are complex patterns of meanings and values related to a group whose borders are defined by the state (Duchesne & Frognier, 2008). The collective conception of national identity is strengthened by a belief in united ideals and values which differentiate one nation from others and add emotional weight to group belonging (Bellucci et al., 2012;Henderson & McEwen, 2005). Building a collective sense of national identity involves delineating what a nation is and is not (Benhabib, 1998). This process involves demarcating an Other (Wodak et al., 2009). As collective national identities are fluid and context-dependent, Others are never fixed; the construction of a group as an Other depends on socio-political context.
The national identity I explore in this paper is Britishness. Britain's collective identity emerged out of wars with France between the Act of Union in 1707 and the beginning of the Victorian age in 1837 (Colley, 1992). The British identity narrative is rooted in history and has long placed Britain in opposition to Europe. First, the Reformation, the Industrial Revolution and Victorian values helped to shape Britain as liberal, democratic, globally influential, and detached from the Continent (Maccaferri, 2019). The Suez crisis of 1956 then precipitated a post-imperial narrative of national decline (Maccaferri, 2019). By the time of the EU referendum, Britain's collective identity had become splintered due to a sense that a forgotten England needed to reassert itself amid increasing devolution (Henderson & Wyn Jones, 2021). The Brexit process undoubtedly exacerbated existing fractures, leading to calls for a second Scottish independence referendum based on Scotland's vote to Remain (BBC, 2016). Polls now suggest that Scottish support for the domestic Union is weaker than ever (Eardley, 2021).

Britain, Brexit and the news media
The British news media has a history of Eurosceptic sentiment that permeates its construction of national identity; since 1973, the British media has 'been on a journey from permissive consensus' to 'destructive dissent' towards the EU (Daddow, 2012(Daddow, , p. 1219). Britain has been consistently pitted against Europe in political, economic, and ideological terms (Anderson & Weymouth, 1999;Daddow, 2012), and Britain's history of parliamentary democracy, its military prowess, and sovereignty have been positioned as the core of British exceptionalism (Maccaferri, 2019;Marcussen et al., 1999).
In recent years, Britain has been constructed positively in national news media. Events that might be interpreted as criticising aspects of British history, such as the 2012 Olympics opening ceremony, are recontextualised in praise of British creativity and courage (Thomas & Antony, 2015). Britons are envisioned as defiant in their irreverent ridiculing of EU regulations in British tabloids (Henkel, 2018), and in right-leaning national newspapers, the UK is depicted as financially superior to EU Member States (Ichijo, 2008). Within the Brexit context, representations are more nuanced: leading up to the referendum, constructions of a nation that had lost its essence could be found next to more positive representations of Britain (Khabaz, 2018). Negative discursive constructions persisted after the EU referendum: in six articles published the day after the vote, Britain was framed as a nation divided between politicians and the public (Koller & Ryan, 2019).
Political research on Brexit and the nation frames Britain as divided along socio-political and cultural lines. It posits that disunity exists between those who benefit from globalisation and those who do not (Goodwin & Heath, 2016;Hobolt, 2016), as well as those who are cosmopolitans and those who are nationalists (Delanty, 2017). Political divisions also persist between those who live in provincial backwaters and those who live in cosmopolitan areas (Jennings & Stoker, 2016). These divisions, researchers claim, crystallise in the broader political cleavage of Remainers and Leaversidentity labels that now command more allegiance than traditional political parties (Kelley, 2019).
Beyond these divisions between citizens, sociologists have identified a populist rhetoric in which the EU referendum vote is conceptualised as ordinary citizens lashing out against out-of-touch political elites (Calhoun, 2017). Populism is a political style of communication (Farrand & Carrapico, 2021;Moffitt, 2017), 'an anti-elite discourse in the name of the sovereign People' (Aslanidis, 2016, p. 97). A populist style contains appeals to the people versus the elite, as well as a narrative of crisis, breakdown, or threat (Farrand & Carrapico, 2021;Moffitt, 2017).
This paper extends research about British division and populism by examining discursive constructions of the nation in the pro-Brexit news media during the final year of Britain's EU membership. It reveals how newspapers consistently construct Britain as humiliated and divided, between politicians and the public and between British citizens with different political stances. It goes beyond the text to consider the socio-political contexts, demonstrating that division is constructed in response to, and incited as part of reporting on, political upheavals.

Data collection
As discourse about Brexit has been so diffuse, I could have analysed various types of media in this project. The diversity of genres available is evident from existing Brexitrelated studies, which range from analyses of Wikipedia pages (Kopf, 2019) to vox pops on news programmes (Miglbauer & Koller, 2018). Newspaper data has not been neglected either; entire books have been dedicated to the UK media's representation of Brexit (Buckledee, 2018). However, there has been no study of how pro-Brexit newspapers constructed Britishness in the years after the EU referendum. This paper fills the research gap by offering a corpus-assisted examination of the pro-Brexit media's construction of Britain as it approached its withdrawal from the EU. The study focuses solely on pro-Brexit publications because two thirds of British national newspapers had a pro-Leave dominance in their articles around the time of the referendum (Levy et al., 2016). Many of the newspapers that voiced a desire to leave the EU also had the highest readership levels (ibid.). Pro-Leave discourses, then, constituted a dominant, widely accessible, and widely accessed construction of Britain for readers.
Newspapers were selected based on their stances towards Brexit. The Express (2016) was unequivocal in its attitude, labelling its campaign to persuade readers to vote Leave a crusade on its front page. The Sun (2016, p. 1) was equally transparent, urging its readers to 'BELEAVE in Britain' and vote Leave.' The Daily Mail claimed on 22 June 2016 that people who believed in Britain would vote Leave. The Daily Telegraph (2016), meanwhile, pitched a Leave vote as an expression of hope for the future of the country. Finally, on the day after the referendum the Daily Star (2016, p. 1) covered its front page with a bulldog wearing a Union Jack hat and called for the public to 'make Britain great again'.
Articles were downloaded from the online news repository Nexis. News and comment articles, alongside letters to the editor, were selected for analysis as they contribute to the newspapers' constructions of Britain. I decided not to collect live blogs from the papers' websites because each time a blog was updated, it appeared as a separate article in Nexis. Collecting each iteration would skew analysis focused on phraseology, as repeated phrases would be from different versions of the same article. However, as the Daily Mail and The Daily Telegraph both frequently reported on Brexit through live blogs, these outlets are slightly underrepresented in the data.
Articles had to include 'Britain' or 'British', 'Europe' or 'European' and 'Brexit', 'nation' and 'people' somewhere in the text. Including the noun 'Brexit' ensured that the articles were related to Britain's withdrawal from the EU. The nouns 'nation' and 'people' limited the texts to those that focused on collective and individual national identities amidst broader discourses about political agreements and politicians. I manually filtered out articles which included non-relevant uses of 'people'.
Articles were published in March and September 2019. I arrived at these dates by counting the number of relevant articles published each month across the year and selecting the two months that had the highest number of published articles. Unsurprisingly, the months with the peaks in output corresponded to key political events: an extension to Article 50 past the original withdrawal date (March), the introduction of the Benn Act, and the ruling that the prorogation of Parliament was unlawful (September). In total, the specialised corpus consists of 350 articles and 316,571 words. Table 1 shows the breakdown of articles and words per newspaper. While the corpus is not balanced, it is representative of the articles that these outlets published in both months (with the caveat that the loss of live blogs means the Daily Mail and The Telegraph are slightly underrepresented). I opted not to balance the corpus because the representative make-up illustrates the extent to which each newspaper's readers were inundated with Brexit-related articles about national identity.
Although down-sampling means this paper cannot speak to discourses across the entirety of 2019, the corpus offers the most representative image available of the discourses of national identity that appeared in Leave-backing newspapers at key points in the final year of the UK's EU membership. The data is worthy of examining because it can provide new insights into how these newspapers mobilised national identities to influence public and political decision-making at different critical junctures (Capoccia, 2016). The down-sampling makes it possible to map discourses onto political events, allowing the analysis to go beyond the texts to their socio-political contexts.
Analytical framework I employ corpus-assisted critical discourse analysis in this paper. Corpus linguistics is the study of discursive patterns in naturally occurring language use through the analysis of texts stored in an electronic database called a 'corpus' (Baker, 2006). Critical discourse analysis (CDA), on the other hand, is an analytical practice that investigates how social inequality is constructed and legitimised discursively (van Dijk, 2015;Wodak, 2001). CDA studies share the understanding that 'discourse is structured by dominance' and is 'situated in time and space', and that powerful groups, such as news organisations, legitimate dominance structures through their ideologies (Wodak, 2001, p. 3). At the core of these principles is an interest in the roles of context and power. Given the interest in power and dominance that underpins CDA, the analytical focus is typically on institutional discourses produced by those with socio-political power, such as news outlets. The combination of corpus linguistics and discourse analysis in corpus-assisted discourse studies (CADS) was first explored by Hardt-Mautner (1995). The approach is defined as 'that set of studies into the form and/or function of language as communicative discourse which incorporates the use of computerised corpora in their analyses' (Partington et al., 2013, p. 10, original emphasis). The aim of CADS is 'the uncovering, in the discourse type under study, of what we might call non-obvious meaning' (Partington et al., 2013, p. 11, original emphasis). CADS can reveal the 'incremental effect of discourse' (Baker, 2006, p. 13). That is, it can highlight discursive patterns that recur across texts but are imperceptible in a manual analysis (Baker, 2006). The CADS position I adopt subscribes to the view of corpus linguistics as a methodology; corpus techniques are used as a way into the data that extends the scope of the discourse analysis (Partington et al., 2013).
Within the CADS approach, I use key semantic domain analysis (Rayson, 2008). Key semantic domain analysis identifies salient discursive patterns by establishing statistically significant, key semantic fields in a dataset. The method involves comparing the relative frequencies of words automatically tagged as belonging to a semantic domain in the target corpus with those in a reference corpus (Rayson, 2008). 1 After establishing key semantic domains, I perform a micro-linguistic analysis of examples from the data, focusing on representations of social actors and actions. I draw on van Leeuwen's (1995) framework of social actor representation, including functionalisation, aggregation, and individualisation, and occasionally Halliday's (1985) transitivity analysis (particularly material processes, which represent tangible actions, as these are frequent in the corpus).

Data analysis
I began the analysis by running the articles through the semantic analysis tagger on the web-based programme suite, Wmatrix4 (Rayson, 2008). Using the UCREL semantic analysis system, I compared the relative frequencies of the different semantic categories in the corpus with their relative frequencies in the British National Corpus Written Informative Sampler. I chose the Informative Sampler because it is one of the only pre-loaded corpora on Wmatrix4 that is similar to the target corpus in terms of genre. The Sampler consists of 779,027 words, just over a third of which is about public affairs (UCREL, 1998). I selected the Informative Sampler instead of the BE06, which is also available on Wmatrix4, because it includes only informative language. By comparing the news corpus to informative writing, I could identify rhetorical or evaluative language. As the BE06 includes fiction, where evaluative language is more frequent, a comparison with this corpus could obscure more evaluative uses of language in the target corpus.
I organised the domain analysis results by effect size and selected as candidates for analysis semantic domains that had a log-ratio value above 1.0 and a log-likelihood value above 6.63 (p < 0.01). I used both log-likelihood and log-ratio measures as they are complementary (Pojanapunya & Watson Todd, 2018). That is, the effect size measure (log-ratio) examines the size of the difference in relative frequency between the target and reference corpora, while the statistical significance measure (log-likelihood) indicates the confidence with which I can say a difference in relative frequencies is not due to chance (Gabrielatos, 2018).
Not all the statistically key domains were involved in discursively constructing Britain. I manually examined the concordance lines for the words in each key semantic domain and removed domains that did not include representations of Britain. I then ranked the semantic categories according to the number of concordance lines that mentioned Britain. I selected the most key domain, 'No respect', for a close reading. The close reading entailed examining each concordance line to distinguish the social actors and actions. That is, I analysed who was constructed as disrespecting whom and then categorised these representations into groups, such as 'British politicians humiliating Britain'. I analyse the most frequent representations below.

Findings and discussion
Of the 36 key semantic domains, nine contained representations of Britain. The semantic domains and their log-ratio and log-likelihood values appear in Table 2.
As Table 2 indicates, 'No respect' has the highest log-ratio and log-likelihood values of all key semantic domains. In the target corpus, the relative frequency of the domain is 44.9 per 100,000 words, in comparison to the relative frequency of 2.6 per 100,000 words in the reference corpus. The high log-ratio and log-likelihood values indicate that the observed frequency difference between the corpora is highly unlikely to be due to error or chance (p < 0.0001).
There are two key themes realised by the 'No respect' semantic domain: The humiliation of Britain and Contempt within Britain. The underlying sentiments of these themes are humiliation and division, which afflict the relationship between politicians and the public, and interactions between citizens. Below, I dissect the themes through close linguistic analysis.

The humiliation of Britain
There are 16 examples of the 'Humiliation of Britain' theme. The most prolific instigator of humiliation is Theresa May, whose Brexit deal and leadership style appear as sources of British humiliation in six examples from March 2019 (when Britain was supposed to leave the EU). As shown in Extracts (1) to (3) below, concordance lines depict May as an intranational Otheran enemy to her country and a threat to Britain's international standing ('self-respect', 'reputation'). The threat is envisioned as an ongoing spectacle through the present perfect tense of the repeated predicate 'has been a national humiliation' (1, 2) and the clause 'the humiliation this country has suffered' (3). The lexicogrammar evokes a dual sense of public helplessness and indignation in the face of political ineptitude, creating a populist narrative of national crisis (Moffitt, 2017): (1) The entire handling of the Brexit negotiations has been a national humiliation. Even calling them "negotiations" is a stretch. The Prime Minister's Withdrawal Agreement hands over our country's self-respect along with control of our future (Telegraph Online, March 2019).
(2) Oh and 90 per cent of Brits think her [May's] handling of Brexit has been a national humiliation. What's been sacrificed here is Britain's reputation on the world stage (The Express, March 2019).
(3) Whatever emerges from the ashes ahead of this week's Commons votes, nothing can conceal the humiliation this country has suffered under Theresa May's lock-jawed leadership (The Sun Online, March 2019).
In Example (1), Theresa May is referred to by her official title as 'the Prime Minister' to emphasise her responsibility as the foremost political representative of the nation. By individualising the prime minister and collectivising the public through the first-person possessive determiner 'our' (van Leeuwen, 1995), the journalist places May outside of the national in-group. Through possessivation (van Leeuwen, 1995), the Telegraph positions the Withdrawal Agreement as May's sole responsibility. This responsibility is enhanced through the metaphorical material process 'hands over' (Halliday, 1985), which constructs the loss of the nation's international standing as a deliberate and deliberated actiona consequence of the political ineptitude of the prime minister and her (mis)handling of Brexit negotiations.
By positioning May's handling of Brexit as a national embarrassment, the Telegraph Online generates distrust in her ability to secure a stable post-Brexit future for Britain. In constructing May as an actor that cannot be trusted, the newspaper establishes the grounds for her removal as prime minister. May resigned three months after Extract (1)'s publication, following Telegraph articles which included calls for her resignation and a headline referring to her as 'desperate, deluded and doomed' (Telegraph, 2019, p. 1). The argument against May in Extract (1) is predicated on the logic that if she is the source of national humiliation, replacing her as prime minister could produce a deal that gives Britain greater 'control' over its post-Brexit 'future'.
Across all three examples, the newspapers employ lexis from the related semantic fields of death ('sacrifice', 'ashes') and emotion ('suffer', 'humiliation'). (Note that this is my categorisation, not domains assigned by the semantic tagger in Wmatrix4). These lexical choices, alongside the first-person plural determiner 'our' in the metonymic noun phrase 'our country', simultaneously personify the nation and conceptualise it as the sum of its people. By depicting the 'death' of the nation's reputation, the journalists reinforce the populist narrative of national crisis (Moffitt, 2017). They imply that what is at stake in negotiations is not just a successful Brexit but the existence of the UK as a national powerhouse. Through the frame of national crisis, the newspapers create a dichotomy between an incompetent prime minister and a suffering population whose future as a collective is hanging in the balance.
The pro-Brexit newspapers also construct a group of pro-European MPs as inflictors of British humiliation because of their 'infighting' and 'refusal to bow to the common good'. These Remain-backing social actors constitute a second intranational Other and appear in five concordance lines, of which Extracts (4) to (6) are representative examples: (4) Britain has suffered almost three humiliating years of infighting, weak leadership and giving away constant concession to the EU (Daily Star Online, March 2019).
(5) Insulated in their Westminster bubble, are they incapable of grasping that their refusal to bow to the common good is destroying faith in democracy, turning the country into a humiliating international joke? (Daily Mail, March 2019).
(6) Perhaps the pro-Brussels majority in Parliamentso obscenely out of step with the people they are elected to represent -have already ensured the greatest national humiliation in our history (The Sun, March 2019).
Examples (4) to (6) offer different representations of what is humiliating Britain and why. The Daily Star, for instance, uses a rhetorical triplet to disparage 'humiliating years of infighting, weak leadership, and giving away constant concession to the EU'. The reference to 'concession' alludes to the resignation of the Secretary of State for the Department of Exiting the European Union, David Davis, who left due to fears of 'giving away too much to the EU' (Walker, 2021, p. 32). Alluding to Davis's resignation provides evidence to readers of May's poor leadership and highlights the problems associated with a soft Brexit. In contrast to the Daily Star, the Daily Mail envisions a populist scenario in which all politicians are 'insulated in their Westminster bubble' (Example 5). The adjective 'insulated', with its connotations of protection and isolation, indicate that politicians are sheltered from the consequences of international humiliation; the victims are 'ordinary citizens' who are discursively constructed through the metonymy of country for persons (Wodak et al., 2009) in the noun phrase 'the country'. This discursive differentiation constitutes a populist rhetoric of elites versus citizens and incites indignation towards (allegedly Remain-backing) political actors who are empowered to make national decisions but protected from their consequences.
Notably, even though May's approach to Brexit is disparaged by the pro-Brexit press as constantly conceding to the EU, politicians are demonised for not accepting her deal in the nominalisation 'refusal to bow to the common good'. The nominal group 'common good' personifies the nation, drawing on collective well-being to portray this political stance as an unpatriotic, selfish act of rebellion that sacrifices the nation's international reputation. The argument that selfishness 'destroys faith in democracy' rests on the idea that politicians are failing to respect the (supposedly homogenous) public opinion and therefore are not fulfilling their responsibility as political representatives in a democracy. As parliamentary democracy is often framed as the cornerstone of British identity in political discourse (Marcussen et al., 1999), a loss of democracy entails the loss of a collective sense of Britishness. Once again, then, the imagined unity of the UK is said to be threatened by the undemocratic behaviours of the largely Remain-backing political class.
The rhetorical effect of Examples (4) and (5) is the same: both articles diminish public confidence in MPs' ability to agree a Brexit deal that will benefit the country. The argument that the prime minister and MPs were incapable of agreeing a strong deal for their country resonated with the British public: in the 2019 Audit of Political Engagement, conducted by Ipsos MORI for The Hansard Society, only 25% of the 1,198 adults surveyed had confidence in British MPs' handling of Brexit (Hansard Society, 2019). As the relationship between discourse and social practice is dialectic (Fairclough, 2015), the media depiction of political ineptitude likely both reflects and reinforces this distrust among readers.
In other words, the newspapers do not just discursively construct a national crisis, they threaten to induce one by cultivating public distrust in political representatives.
In Example (6), the nation is constructed through the dichotomy of British 'people' and a 'pro-Brussels majority in Parliament'. The pre-modifying adjective 'pro-Brussels' provides a more explicit version of the implied Remainer politician -Leaver public divide that underpins Example (5). The metaphorical phrase 'obscenely out of step' presupposes that the only ideological position among Britons is anti-EU sentiment. By linguistically omitting 'Remainer' Britons from its construction of the public, The Sun implies that only Leave-voting Britons exist (see Koller & Ryan, 2019). Given that only one million votes separated Leave and Remain in the 2016 referendum, this discursive strategy excludes almost half of the voting population in the UK. Discursively erasing the existence of Remainer identities from a British in-group threatens social exclusion by legitimising intolerance towards political positions that diverge from the (slight) majority.
Perhaps more ideologically remarkable than the socially divisive rhetoric, however, is the construction of ongoing political support for the EU as 'the greatest national humiliation in our history'. Through this claim, The Sun rewrites a version of national history in which the ideological divergence between politicians and a small majority of the public is more problematic for the nation's reputation than, for example, the Civil War of the 1600s, division between pro-and anti-slavery stances in the eighteenth and nineteenth centuries, or contemporary calls for Scottish independence. By ignoring historical examples of British division, the journalist frames Brexit as the most critical juncture (Capoccia, 2016) in British history. This construction of national crisis, which displays a manipulation of British history to serve a Leave-backing agenda, commands a sense of urgency and encourages an outraged public reaction towards the present state of the nation.
The pro-Brexit newspapers sow distrust between British politicians and an already sceptical public by undermining Theresa May's ability to handle Brexit and blaming a group of MPs for indecision over a future Brexit deal. The populist dichotomy of politicians and the public engenders a pejorative self-reflection on the state of the nation; the country's proud international standing is increasingly jeopardised in this narrative of national crisis by the incompetence of political officials. The populist discourse both reflects and reinforces public dissatisfaction with the state of the UK: 56% of participants in the 2019 Audit of Political Engagement believed Britain to be in decline (Hansard Society, 2019). The image of division expands beyond the politician-public relationship; the newspapers presuppose that all members of the British public harbour anti-EU sentiment, erasing Remainers from the collective British identity. This discursive exclusion precipitates social exclusion by implicitly encouraging intolerance of pro-EU views while strengthening pro-Brexit readers' sense of belonging to a collective political identity of Leave-backing Britons.

Contempt within Britain
The 'Contempt within Britain' theme depicts a nation divided along two simplified, populist axes. The first division is, once again, between politicians and the public. While the nation is seldom explicitly mentioned in these concordance lines, the dichotomisation of political elite versus the public contributes to a broader comment on the state of the nation in 2019. A binary construction of elite versus citizen appears in seven concordance lines, a representative selection of which I reproduce in Extracts (9) to (12): (9) What will have been revealed in a devastating, unforgettable form is the contempt which those who govern feel for those who elect them (Telegraph Online, March 2019).
(10) The electorate is being treated with undisguised contempt. If they get away with itwhich they probably will -Britain will have ceased to be a proper democracy (MailOnline, March 2019).
(11) The courts place themselves above democracy. MPs shun a general election. Our freedoms are betrayed by an elite who think they always know best. For over three years, the establishment has treated our people with an unfettered, sneering contempt (The Sun, September 2019).
(12) SIR -The Speaker and the Remainer Members of Parliament have betrayed the people and trashed the constitution. A bewildered and furious nation will eventually get its chance to show its contempt (Telegraph Online, September 2019).
In Examples (9) and (10), Britons are represented through the functionalisation (van Leeuwen, 1995) of the noun 'electorate' and the noun phrase 'those who elect them'. The functionalisation emphasises the power of choice that citizens have to change their political representatives (and by extension, the direction of Brexit). This public empowerment is juxtaposed with the 'contempt' (9, 10, 11) which politicians supposedly feel for citizens, an emotion that is pejoratively intensified through the adjectives 'undisguised', 'unfettered' and 'sneering'. The juxtaposition between public power and politicians' contempt frames the hostility as arrogance, inciting and legitimising anger towards political representatives.
What is at stake in Examples (10) and (11) is once again the UK's status as a democracy ('ceased to be a proper democracy'; 'place themselves above democracy'). Although these examples do not explicitly comment on the existence of the nation, in the British context loss of democracy entails the loss of national identity and thus constitutes a national crisis (Marcussen et al., 1999). An implicit call for a reaction to the threat of crisis underpins the letter in Example (12). This extract reveals that contempt between politicians and the public is experienced mutually. The format of public expressionthrough a citizen's letter rather than a quotebolsters the newspapers' broader claim that Britons are dissatisfied with the state of British politics. The expression of political disappointment is reinforced by the citizen through the emotion adjectives 'bewildered' and 'furious' and the metonymy of country for persons (Wodak et al., 2009) in the noun phrase 'a bewildered and furious nation'. Overall, the example legitimates anger and confusion and supports the broader implicit argument that citizens should demand change from their political representatives to secure Brexit and protect the country.
The second populist division is between a pro-EU, metropolitan elite, labelled 'Remainers', and a hardworking, disillusioned public referred to as 'Brexiteers' or 'Leavers'. There are 86 instances of Remainer* per 100,000 words and 71 instances of Brexiteer* per 100,000 words in the corpus. Leaver* is comparably less popular, appearing only 0.9 times per 100,000 words. That the terms 'Remainer' and 'Brexiteer' persist after the EU referendum indicates that the identities have become what Zürn and de Wilde (2016, p. 284) term a 'cleavage'. A cleavage refers to a situation in which several conflicts are subsumed under a single dimension between opposing groups, 'reinvigorating a new sense of identity politics' (Zürn & de Wilde, 2016, p. 284).
In the corpus, the 'Remainer' identity indexes a group of upper-middle-class, highly educated people who have benefited from globalisation. 'Leavers' and 'Brexiteers', in contrast, index a politically aware but oft-disdained group of 'ordinary' citizens. The relationship between these two identities is what the 'Contempt within Britain' theme captures and mobilises; the representation is particularly prevalent in the Daily Telegraph, as Extracts (13) and (14) demonstrate: (13) At Tate Britain, the Brexiteers were booed by visitors to the gallery who stood on the steps and signalled a thumbs down. Here was seemingly London's metropolitan elite jeering the largely out-of-town Leavers as they walked past. Perhaps nothing so much summed up the nation's divide. When the Remainers booed, the Leavers shouted back: These concordance lines reveal that the political identities of 'Remainer ' (13, 14), 'Brexiteer' (13,14) and 'Leaver' (13) do not solely reflect attitudes towards the EU. Rather, they symbolise a range of national and social divisions in the pro-Brexit press, particularly in terms of political ideology and socio-spatial divisions (Bachmann & Sidaway, 2016). Through the proper noun 'London' and the adjective 'out-of-town' in Example (13), the Telegraph Online employs the spatial imaginary of Left Behind Britain (Sykes, 2018) to discursively separate London from the allegedly rural rest of the country. The cosmopolitan-rural division (Jennings & Stoker, 2016) alludes to the EU referendum vote, in which London opted for Remain and most of England's remaining counties voted to Leave. This context suggests that the gulf between the two camps is related solely to the way people voted in the EU referendum. As the nominal group 'the nation's divide' indicates, though, the city-countryside dichotomy also symbolises an intranational problem: the perceived injustice of UK economic policies that facilitate the prosperity of London and its 'metropolitan elite' at the expense of other regions (Calhoun, 2017;Jennings & Stoker, 2016). In this discursive frame, geography symbolises socioeconomic status: the national 'divide' is between those who have and those who have not (Goodwin & Heath, 2016;Hobolt, 2016). By emphasising a geographical and socio-economic distinction between Remainers and Brexiteers, Telegraph Online frames the 'jeering' of the metropolitan elite as an insult to disadvantaged Brexiteer readers. Once again, then, the newspaper perpetuates division in the UK by its selective reporting, which incites bitterness towards a socalled self-serving 'elite' and produces a narrative in which Leavers are disadvantaged.
The Telegraph Online's use of 'Remainer' and 'Leaver' identities goes beyond merely blaming the EU for national problems. Leave and Remain stances have become new 'political and social fault lines' in the UK, and the associated identities now command a stronger allegiance than traditional political parties among Britons (Kelley, 2019, p. 13). Framing social divisions within the context of the EU referendum harnesses the strength of the public's emotional attachment to these identities to shape the political landscape. These emotional attachments can be aroused by emotive lexis, such as the adjective 'nasty' and the nouns 'contempt ' and 'bigotry' in Example (14), which cultivates a sense of passionate indignation. Concerns about the breakup of the UK commonly cite Scottish, Welsh, and Northern Irish nationalisms as likely catalysts (Cochrane, 2020). However, the conflation of national and political identities in the pro-Brexit press reveals the perception that there is as much division within the UK's constituent countries as there is between them. According to pro-Brexit newspapers, then, the threat to the existence of the UK is socio-political division.

Conclusion
This paper reveals the construction of a humiliated and divided Britain in the pro-Brexit press in 2019. Pro-Brexit newspapers argue that the ineptitude of MPs constitutes a national crisis that threatens Britain's long-standing reputation for democracy. A populist dichotomy of political elite versus ordinary citizens underpins the argument; the politician-public relationship seemingly radiates mutual contempt. This divisive rhetoric captures and reinforces the public mood; 50% of people surveyed in the 2019 Audit of Political Engagement felt that the main political parties do not care about people like them (Hansard Society, 2019).
Depicting a fractured relationship between politicians and the public supports the newspapers' political agenda to shift the balance of power towards MPs who support a hard Brexit. The rhetoric cultivates public dissatisfaction with the existing Parliament, a sentiment Theresa May could not afford to ignore after facing a vote of no confidence. The pro-Brexit press does not only depict division between politicians and the public. Under the EU-related guise of the nouns 'Remainer' and 'Brexiteer', it envisions a public cleaved apart by socio-economic and geographical disparities. The Daily Telegraph persuades readers that Remainers consider them inferior. There are consequences to this divisive rhetoric. The Remainer-Brexiteer dichotomy nurtures a belief in an unjust gulf in British society between those who do and do not benefit from globalisation, which risks creating a deep-rooted bitterness in the UK at a time when collective national identity is paramount.
In sum, the pro-Brexit press constructs a nation that is more splintered than ever. The labels 'Remainers' and 'Brexiteers' reduce various political and socio-economic positions to a bitter binary of 'elites' and 'ordinary people'. The divisive rhetoric mobilises EU-related identities for political gain but risks fomenting a resentment so deep that it may be difficult to harness a collective post-Brexit national identity after the transition period. By blaming politicians and Remain-voting citizens for national divisions, the newspapers obscure their role in inciting dissatisfaction through their reporting of Brexit-related upheavals. Note 1. The semantic tags are determined by an underlying lexicon that has been manually compiled by researchers over many years.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by Midlands4Cities Doctoral Training Partnership.

Notes on contributor
Tamsin Parnell is a doctoral student in Applied Linguistics at the University of Nottingham. She is interested in the intersection between discourse and politics, and is the co-organiser of the Discourse and Politics Seminar Series.