Settlement types,territorial discourse and the symbolic production of the ‘post-industrial town’

Abstract

Couched in work on territory, discourse and symbolic power, this paper examines how dominant representations and classifications of settlement types are produced by overlapping political, media and academic discourses through an investigation of the ‘post-industrial town.’ The 2016 referendum on European Union (EU) membership and the 2019 General Election thrust England’s deindustrializing towns into the foreground. Viewed as the primary settlement type of the ‘Brexit heartlands’ and the ‘Red Wall’, descriptions, encodings and classifications of the ‘post-industrial town’ have proliferated across media, academic and political discourses. Following, the article draws on Critical Discourse Studies to trace the emergence of the ‘post-industrial town’ as a territorial production. I argue that the ‘post-industrial town’ has been produced by discursive and socio-technical practices invoking symbolic imaginaries of how places look, who lives there and their practices. In the case of ‘post-industrial towns’, a dominant symbolic production has emerged of an older, white, working-class, non-cosmopolitan and socially conservative type of settlement. Moreover, discursive and symbolic territorial productions have led to (un)wilful misrepresentation and misrecognition as places ‘on-the-ground’ are brought into political strategizing and discourse.

Introduction

This article examines how dominant representations and classifications of settlement types are produced through overlapping political, media and academic discourses. It does so through investigating representational discourses that have proliferated in the United Kingdom (UK) around ‘post-industrial towns.’ Since the 2016 referendum on European Union (EU) membership, non-metropolitan urban spaces have received sustained media coverage, political campaigning and academic analysis (Bounds, 2019; Crampton, 2020; Jennings and Stoker, 2019; Office for National Statistics, 2019; Skelton, 2019). Commentary gradually condensed into a focus on ‘towns’, distinguishing the town from other types of human settlement in the UK. This towns discourse has sought to further delineate a settlement subtype of the ‘post-industrial town’, positing that places are similar enough in terms of their cultural characteristics, demographics and economic geography to classify them together as a type of settlement, and that, in turn, these signifiers distinguish the ‘post-industrial town’ from other settlement types. This article documents Critical Discourse Analysis research tracing how the towns discourse produced the ‘post-industrial town’ as a settlement subtype through a series of socio-technical, discursive and symbolic practices, the politics underpinning such territorial strategies and the symbolic representations invoked (Luukkonen and Moisio, 2016; Meyer and Migglebrink, 2013; Van Leeuwen and Wodak, 1999).

Conjoining conceptual work on territory with that of discursive and symbolic production, the next section discusses how settlement (sub)types are territorializations produced through representational practices of classification (Bourdieu, 1989, 1991; Brenner and Elden, 2009; Elden, 2007, 2010; Sisson, 2021; Schultz Larsen and Delica, 2019). Common among state and non-state statistical bodies, territorial classification of settlement types is typically based on densities of built-up area and population, or a combination of both (Healey, 2002). Alongside such density-oriented classifications, actors in media, politics, government offices, research institutions, as well as in wider social life, engage in discursive and symbolic practices that also territorialize settlement types. Discursive territorialization here refers to where territorializations, in this case settlement types, are delimited and classified through representational strategies encompassing words and images. Discursive territorialization is notably apparent in the ways that settlement subtypes are classified, denoting, for example, economic activity (fishing village, mining town), functions or features (cathedral city, arts district, shanty town) or demographic conditions (working-class town, black neighbour, affluent suburb). Moreover, instructive work on conceptualizing cities documents how groups propose different conceptions of what constitutes and delineates ‘the city’ from other settlement types, doing so by drawing on a range of quantitative and qualitative resources (Healey, 2002). As Sisson (2021: 409) writes, ‘[r]epresentations of territory’ are rendered ‘cartographically, diagrammatically, statistically, discursively, and otherwise symbolically.’

Classifying and naming of settlement (sub)types in more qualitative ways is often taken-for-granted as innocuous shorthand description. However, urban scholarship highlights the symbolic and stigmatizing meanings invoked by the naming of certain settlement types (Slater, 2018; Sisson, 2021). As described below, representational territorial practices work to ascribe how settlement types look, the people who live and belong there and the ways of living taking place (Bourdieu, 1989). This, in turn, has the potential to shape wider symbolic associations and valuations of actual places depending on how they come to be territorialized and classified as certain types of settlement (Wacquant, 2008). In some cases, positively valenced settlement subtypes are used toward ‘place branding’, such as with the practices of tourism boards or property marketing (Andersson, 2014). In other cases, symbolic production may be an unintended effect resulting from tangential work. Indeed, territorialization may itself not be the intended purpose. The argument made here is that territorial production is also pursued intentionally by actors mobilizing settlement classifications toward certain (political) ends (Brenner and Elden, 2009).

This is the case with the ‘post-industrial town’, which has emerged as a politicized settlement type where meanings are invoked to suit different purposes. The towns discourse evaluates the ‘post-industrial town’ as paradigmatic of realigning political geographies, abandoned by their representative political party, Labour, and, in turn, a key electoral battleground for the rival Conservative Party (Payne, 2021). Following, discursive and symbolic territorial productions have led to (un)wilful misrepresentation and misrecognition as places ‘on-the-ground’ are brought into political strategizing and discourse. This has immediate relevance to spaces that have undergone similar trajectories to deindustrializing towns. ¹ The peripheral deindustrializing areas central to political debates across Europe and North America, for example (Gordon, 2018; MacKinnon, 2020). Contributing to literature on the territorialization of spatial forms, the understandings also have significance beyond deindustrializing spaces by elucidating how places become discursively and symbolically produced as one type of settlement (or territory) and not another.

Whereas towns ‘tended to be out of favour in policy-making and academic debate’, in recent years, substantial amounts of analysis, reportage and commentary has attempted to conceptualize and classify towns (Beatty and Fothergill, 2020b: 1239). Outlined in the third section, the research documented here applied principles of Critical Discourse Studies to a range of published materials, including speeches by key politicians, grey literatures by thinktanks and political thinkers, academic studies by prominent political scientists and journalistic accounts (Fairclough, 1993; Reisigl and Wodak, 2009). Wacquant (2008: 48) terms such groups and actors ‘specialists in symbolic production.’ Symbolic productions become dominant when produced by those deploying surpassing levels of symbolic power, the capacity to construct legitimated and normalized realities. Actors in the media, politics, thinktanks and academe are, of course, not a homogenous ensemble sharing a singular agenda. However, they have each engaged extensively in territorializing practices that have symbolically produced the ‘post-industrial town.’ The task is to trace territorializations across a variety of texts, as well as critiquing the territorial practices and political purposes underpinning these territorial productions (Bourdieu, 1991; Sisson, 2020).

Formative to dominant territorialization of the ‘post-industrial town’ was political science scholarship prefiguring the EU Referendum that conducted statistical analyses of existing survey data to interpret changing voting behaviour (Ford and Goodwin, 2014a; Jennings and Stoker, 2016). A ‘left behind’ (Ford and Goodwin, 2014b) voter was classified, naming older, white, working-class voters with socially conservative and communitarian politics predominant in ‘declining’, non-metropolitan areas, which Jennings and Stoker (2016) termed ‘backwaters.’ The notion of a ‘left behind’ population gained extensive traction within media and political discourse (often in truncated and generalized forms) following the 2016 EU Referendum. Responding to the geography of the referendum, the ‘left behind’ as a symbolic rhetorical device was applied spatially to refer to places in addition to a certain voter type, whereby the ‘Brexit heartlands’ were viewed as ‘left behind’ non-metropolitan places home to a ‘left behind’ voter type.

Between the 2017 and 2019 General Elections, ‘left behind’ places were identified in academic, media and political discourse as being principally towns, notably seaside and post-industrial towns. The emergent towns discourse was bolstered by political science work positing broad cultural, political and demographic differences between the ‘city’ and the ‘town,’ where ‘town’ was territorialized relationally as an antithetical settlement type to a supposedly diverse, vibrant and cosmopolitan ‘city’ (Brett et al., 2017; Jennings, 2017; Jennings and Stoker, 2019). In the run-up to the 2019 General Election, the ‘left behind town’ was then positioned as the primary settlement subtype of the ‘Red Wall,’ a band of historically Labour Party supporting constituencies ‘comprised of post-industrial towns and rural areas surrounding the big Labour-supporting cities of the North and Midlands’ (MacKinnon, 2020: 13). Politicians and political strategists increasingly solicited support from ‘left behind’ ‘post-industrial’ towns and, in turn, became centred in discourse on the soc-called ‘Levelling Up’ agenda (Skelton, 2019). During this phase, ‘left behind town’ and ‘post-industrial town’ became synonymous. The ‘left behind/post-industrial town’ of the ‘Red Wall’ was extensively reported on, accompanied by images of disgruntled ‘left behind’ voters in empty, rundown highstreets (Payne, 2021).

What has emerged from this process is a multifaceted symbolic production of the ‘post-industrial town’ which classifies a settlement type that is demographically white, working-class and older, in economic decline, non-cosmopolitan and socially conservative. The fourth section of the paper argues that the over-focus on ‘left behind’ indicators misrecognizes much of everyday life in deindustrializing towns in favour of partial, politicized territorial productions. This acts to symbolically marginalize the people, social practices, politics and identities of those that do not fit with dominant territorial imaginaries. It is, therefore, important to investigate the territorial practices that gave rise to dominant imaginaries, how descriptive naming structures come to dominance, and the symbolic territorial meanings ascribed to and evoked by settlement (sub)types.

Settlement types, symbolic power and discursive territorial practices

The practice of classifying material human settlements into types of settlement follows territorial strategies. Understood broadly, territory refers to a ‘bounded social space that inscribes a certain social meaning onto defined segments of the material world’ (Delaney, 2008: 14). Alongside physical control of space, territory is brought into being through a series of representational strategies. These include socio-technical practices of quantification, measurement and mapping, resulting in ‘representations of territory’ (Brenner and Elden, 2009) that look to encode, delimit, classify and name material space (Storey, 2012). Human settlements are similarly measured, delimited and classified. Often classification is relational, with settlement types ordered into size-based typologies, apparent in commonly known ‘settlement hierarchies’ of hamlets, villages, towns and cities. This practice is, of course, long standing. The English word ‘town’ originates from the Old English ‘tun’, meaning a group of dwellings, and still evident in toponyms such as Luton and Brighton. This might delineate a ‘tun’ from a ‘ham’, the root of hamlet, meaning farmstead, and evident in Nottingham and Birmingham, for example. Strategies and practices for classifying settlement types have, however, become increasingly more sophisticated, rationalised and, arguably, quantitatively driven.

Focus by state and non-state bodies has mainly been on classifying the ‘urban’ and/or ‘city.’ Thresholds of population size and/or built-up area are regularly used (UN-Habitat, 2016). Reflecting shifts to agglomeration economies, quantitative classifications serve as geospatial boundaries for the generation of statistics in the governance of urban populations and economic productivity. Local administrative and governance structures, as well as built environment and symbolic factors play additional roles in classifying cities within different state territories (Storey, 2012). For example, in the UK, places are only ‘officially’ classified as cities when granted ‘city-status.’ The criteria for city-status have changed over time, such as the previous requirement to have a cathedral. This has resulted in many places without ‘city-status’ being larger in both built-up area and population than smaller ‘official’ cities (Beckett, 2017). Such classifications sit alongside, and contrast with, those of other groups with vested interests in conceiving the city (Healey, 2002). Thus, in part, following Elden (2007), classification of cities as settlement types can be viewed as part of a wider process of capitalist statecraft and governmentality where cities play an important role in terms of population dynamics and economic growth.

It is within the context of city-led post-industrial economies that attention on towns and other non-metropolitan geographies has emerged (Rodriguez-Pose, 2018). The absence of conceptualizations of towns in the UK until recently is perhaps reflective of their marginalization as a settlement. Towns have been imprecisely conceived relationally on an urban-rural spectrum of settlement hierarchy – bigger than a ‘village’, smaller than a ‘city’. Responding and contributing to the towns discourse, numerous quantitative typologies have been formulated of small, medium and large towns based on population thresholds, and largely at the discretion of their producers (Baker, 2018; Brett et al., 2017; Centre For Towns, 2017; Office for National Statistics, 2019). For example, Centre For Towns (2017) classifies small, medium and large towns as settlements of between 10,000 and 30,000 residents, 30,000 and 75,000 residents and over 75,000 residents respectively. The House of Commons Library (Baker, 2018) adapts this, with ‘small towns’ being between 7500 and 24999, ‘medium towns’ 25000 to 59999 and ‘large towns’ 60000 to 174999. Though arbitrarily drawn, such quantitative classification indicates the primacy of statistical territorialization of the ‘post-industrial town.’

Influential research by political scientists and thinktanks tend toward practices of mapping cross-sectional survey data on cultural, political, economic and demographic factors. Quantification of qualitative factors – for example, beliefs and values – is a socio-technical practice of ‘statisification’, ‘a mode of understanding and translating socio-qualitative matters into quantitative “facts”’ (Meyer and Miggelbrink, 2013: 213). A defence of these methods could be that they merely read and represent conditions on-the-ground. However, as Meyer and Miggelbrink (2013: 213) continue in reference to regions, ‘[p]ractices of subjects are generalized and displayed in official statistics as well as in scientific and non-scientific surveys, rankings etc. characterizing regions.’ This occurs also with territorial classification of settlement types, and, as discussed below, the ‘statisification’ of cultural and political values provides the supposed empirical basis of much of the symbolic production of the ‘post-industrial town.’

‘Statisification’ of territorial dynamics ‘in turn become elements of macro-discourses about the state-of-the-society seemingly depicting a true picture of the situation’ (Meyer and Miggelbrink, 2013: 213). Participants in these macro-discourses include a range of non-state and non-research bodies, principally the media and cultural industries who often deploy more discursive and symbolic strategies of territorial representation. Work on territorial stigmatization is insightful in this regard as it is concerned with the (co-)production of representations of territory ‘by an ensemble of institutions within and outside the state: political representatives, bureaucrats, think tanks, social scientists, journalists and commentators, etc’ (Pattison, 2022; Sisson, 2021: 7; Slater, 2018; Wacquant et al., 2014). As mentioned, Wacquant (2008: 47) terms these actors ‘specialists in symbolic production’, active in ‘producing legitimate and legitimising knowledge’ (Schultz Larsen and Delica, 2019: 542). Research has investigated how these groups of actors produce territorial representations ‘through the delimitation of socio-spatial categories and boundaries’ (Sisson, 2021: 8). Though stigmatization is not the primary concern of this paper, there are notable similarities in the territorial practices between the denigration of material places and the representational and symbolic production of settlement types. This is particularly apparent with regards to settlement subtypes, such as the ‘left behind’ or ‘post-industrial’ town, where the noun modifier acts a descriptive container for symbolic meanings of social, political, economic and/or cultural indicators.

Naming practices use ‘words to describe, portray and depict objects, phenomena or social groups’ (Schwarze, 2022: 1420). In this sense, discourse – words, names, images, narratives – can seek to express social realities (Bourdieu, 1989). Media and political discourses also act to inscribe social meanings onto space through representational practices of description, naming and other discursive means (Giraut and Houssay-Holzschuch, 2016). Work in the field of territorial stigmatization has documented how stigmatizing representations of specific places in media and political discourses has rendered them ‘synonyms for social hell’ (Wacquant et al., 2014: 1273). Stigmatizing macro-discourses also operate discursively to produce euphemistic territorial types such as the ‘sink estate’, ‘problem neighbourhood’ and ‘ghetto’ (Devereux et al., 2011; Slater, 2018; Wacquant, 2008). While settlement subtypes may seem to exist through benign naming practices, macro-discourses still ascribe settlement types with certain meanings through generalized descriptions and territorial representations that then transfer into wider symbolic imaginaries.

Specialists in symbolic production, thus, use territorial practices to coalesce disparate people and places into legitimized and accepted typologies, classifications and hierarchies, including settlement types. Importantly, quantitative measurement, journalistic commentary, political statements and so on, do not merely ‘read’ already existing territorial units. They enact them – bringing them into existence discursively through classification and naming (Sisson, 2021). As Bourdieu (1989: 23) states, the ‘power of designation, of naming, brings into existence in an instituted, constituted form, what existed up until then… a collection of varied persons.’ Representations of territory emanating from a range of actors, then, are conceived rather than merely received and reproduced (Sisson, 2021; Schwarz, 2021). While such productions are mediated by ‘the degree to which the vision proposed is founded in reality’ (Bourdieu, 1989: 23), certain groups, institutions and individuals are better placed than others to propose, circulate and substantiate certain visions.

Bourdieu’s (1989, 1991) concept of symbolic power is useful to conceiving the process by which representational and symbolic territorial productions become accepted and normalized as accurate representations of space. Symbolic power refers to the capacity to construct reality through legitimating meanings, representations and imaginaries of the world and derives from symbolic capital, ‘the form that the various species of capital assume when they are perceived and recognized as legitimate’ (Bourdieu, 1989: 17). Legitimate forms of economic, social and/or cultural capital confer the bearer symbolic capital, operationalized in symbolic power to shape perception, (mis)recognition and versions of reality (Pattison, 2022). Elite groups of actors, often with access to mass communication mediums, can exercise their symbolic power to ascribe space with particular meanings (Sisson, 2021).

Moreover, territorial production by these groups evokes and informs imaginaries, assumptions and valuations of spaces (Devereux et al., 2011; Schwarze, 2022; Slater, 2018; Storey, 2012). This has potential political and material effects on how places are treated, for instance in policy interventions, electioneering and investment decisions (Painter, 2010). For example, the (co-)production of stigmatized territory can be in order to legitimize state abandonment, gentrification or to render marginalized places as responsible for their own inequality (Pattison, 2022; Slater, 2018; Wacquant et al., 2014). As will be discussed, the symbolic production of the ‘post-industrial town’ has already played a role in the outcome of the 2019 General Election and the formulation of the Levelling Up agenda. As questions of place and region are increasingly at the forefront of political and social agendas in relation to identity and inequality, the symbolic production of settlement types is likely to have future political, economic or social implications.

Discourse analysis and the towns discourse

The primary intention of this research was to critically trace the production of dominant representations of deindustrializing towns over the preceding decade. This research agenda emerged from a wider ethnographic project on senses of alienation in deindustrializing towns that began in September 2019. By that time, deindustrializing towns were at the centre of the popular political landscape and a burgeoning towns discourse was active in the statistical measurement, observation and value judgement of economic, political, cultural and social dimensions of towns (Jennings and Stoker, 2019). This was in the context of an impending general election, that eventually took place in December 2019. Politicians, political strategists and campaigning thinktanks were soliciting voters in ‘left behind’ towns. The sense was that deindustrializing towns were being homogenized in their representation, that this homogenizing framing and labelling was occurring interdiscursively across distinct media, political and research discourses, and that much of the dominant narrative contrasted with social scientific research, as well as lived experiences, within specific places depicted as being typical of deindustrializing towns. It followed that conceiving the process by which this dominant representation of deindustrializing emerged and congealed into the territorial classification of a settlement type would elicit worthwhile insights on the discursive and symbolic production of territorial forms.

Research drew on Critical Discourse Studies, a broad research agenda oriented to examining relationships between power and discourse. At root, discourse emerges and functions through texts, understood in Critical Discourse Studies not only as communicative tools but ‘sites of social struggle in that they manifest traces of differing ideological fights for dominance and hegemony’ (Meyer and Migglebrink, 2013; Reisigl and Wodak, 2009: 89; Van Leeuwen and Wodak, 1999). For Critical Discourse Studies scholars, texts are inherently political regardless of what genre of discourse they are produced, with genre of discourse relating to the ‘use of language associated with a particular social activity’ (Fairclough 1993: 138), for example, scholarship or journalism.

Particularly appropriate for the research aims was the Discourse-Historical Approach, which ‘centres on political issues and seeks to integrate as many genres of discourse referring to a particular issue as possible, as well as the historical dimension of that issue’ (Van Leeuwen and Wodak, 1999: 91). The Discourse-Historical Approach focusses on issues of intertextuality, interdiscursivity and recontextualization. Intertextuality refers to how texts incorporate ‘elements of other texts (words, phrases, arguments, topics or larger elements)’ (Wodak, 2011: 630). Intertextual research aims to trace the ascendency and legitimation of arguments and knowledge through their intertextualisation into other texts. Recognizing that ‘discourses are linked to each other in various ways’ (Reisigl and Wodak, 2009: 90), interdiscursivity refers to how texts are characterised by combinations of different discourses and different genres of discourse (Wodak, 2011). Recontextualization is the ‘process of transferring given elements to new contexts’ (Reisigl and Wodak, 2009: 90) and is ‘concretely manifested’ (Wodak, 2011: 629) in the interdiscursivity and intertexuality of texts. An analytical focus on recontextualization attends to the ways that new words, phrases, labelling, arguments and so forth are introduced, alongside existing elements, into discourses and genres in potentially rearranged, reinterpreted or selective forms.

A Discourse-Historical Approach was applied to the towns discourse, analyzing processes of intertextuality, interdiscursivity and recontextualization. The towns discourse encompasses a multitude of actors operating in different fields of action that each have distinctive genres and practices of communication. Moreover, fields of action engaged in the towns discourse have competing interests and purposes, using texts to articulate and advance these agendas. However, while ‘specialists in symbolic production’ produce different texts within separate genres that may have differing intentions, all seek to legitimate their claims, agendas and knowledge (Bourdieu, 1989). Pursuant of this, there has been much dialogue and recontextualization of argument and analysis between actors, fields of action, discourses and genres.

Notably, from 2016, political science narratives forwarded by Ford and Goodwin (2014a, 2014b) and Jennings and Stoker (2016, 2019) gained ‘traction in the media and political circles’ (McCann, 2019: 265). Goodwin, a political science academic, has made regular appearances in the broadcast and print media. Jennings, also a political science academic, has closely collaborated with Lisa Nandy, a Labour Member of Parliament (MP) and foremost political voice in the towns discourse, establishing a thinktank committed to researching and advocating for towns (Centre For Towns, 2017). Other influential thinktanks and political strategists/commentators have also contributed to the towns discourse, including Renewal (2015), New Economics Foundation (2017), Resolution Foundation McCurdy et al. (2020) and Onward Tanner and O’Shaughnessy (2019). These institutions purposefully seek to impact policy formulation and political campaigns from particular ideological perspectives, either explicit or discreet, and are deeply entwined with political parties and the media. For example, Onward’s current advisory board and Board of Directors contain The Conservative Party (2019) MPs and Conservative Party members of the House of Lords (Onward website, 2023).

The process of intertextual and interdiscursive research first involved ascertaining the above prominent actors and institutions shaping and forming the towns discourse. This was an iterative process assessing multiple fields of action – research, media and politics – to identify key actors and institutions, as well as the forums, platforms and mediums through which the towns discourse takes place. Importantly, it is not always standard practice for certain genres of discourse to explicitly declare sources of intertextuality and interdiscursivity. This is particularly the case in media and political fields, where, for example, a political speech or newspaper article, will not always note where, or from which text, its argument or comments emerge. Thus, analysis required some inference of intertexuality; however, particular phrasing, terms or analysis can be traced to other texts, often produced in other fields of action.

Research was limited to published, available texts. As the research investigates how territorializations have been intertexualized and recontextualized, texts were selected from a span of actors, fields and discourses, as opposed to targeting any one specific actor or field. Analysis focused on key texts across scholarship, media representations and oral and written texts by prominent politicians. Texts analysed were: research publications produced by political scientists Jennings and Stoker (2016, 2019) and Ford and Goodwin (2014a, 2014b), often published in the journal Political Quarterly; written and video texts by Lisa Nandy MPs (2020), Nandy and Quilter-Pinner (2017), Bates (2018); speeches by the former Prime Minister (PM) Boris Johnson (2019); books by political commentators David Skelton (2019) and David Goodhart (2017); and policy documents by thinktanks Renewal (2015), New Economics Foundation (2017), Resolution Foundation (McCurdy et al., 2020), and Onward (Tanner and O’Shaughnessy, 2019), and the Office for National Statistics (2019). The print and broadcast media have also covered, and contributed to, the towns discourse extensively (Bounds, 2019; Chakelian, 2020; Crampton, 2019; Foges, 2019; O’Connor, 2017). For example, a national tabloid newspaper, The Mirror, ran a 5-day series of articles entitled Town 2020 in September 2019, which included analysis, reporting and opinion pieces by leading figures in the towns discourse. Online media texts, including news items, opinion columns and features were collated by internet search using dates between June 2016 and December 2021. Only illustrative media texts are referenced in this paper.

Forty-six texts in total, ranging in length, medium and genre, were closely analysed in terms of the arguments and claims regarding conceptualization of deindustrializing towns, how these arguments and claims were generated through discursive and socio-technical practices, as well as how arguments and claims had been recontextualized in them. This included examining the use of naming, labelling, description, phrasing, argumentation, tropes, rhetoric and subtext. Critical to the eventual analysis was the continual recontextualization of meanings and arguments ascribed to the term ‘left behind.’ This analysis, presented in subsequent sections, finally involved developing a historical understanding of how the ‘post-industrial town’ was produced territorially by the recontextualization of academic, media and political discourses over successive phases.

Territorial production of the ‘post-industrial town’

This section traces the territorial production of the ‘post-industrial town’ through the towns discourse. In sum, analysis by political scientists and thinktanks has been recontextualized by media and political fields, emplacing encodings of a white, older, socially conservative, working class voter archetype in an increasingly delimited settlement type. This territorial production was couched in a series of socio-technical and discursive practices conditioned by wider political developments. Though the EU Referendum and 2019 General Election were major propellants for recontextualization, the towns discourse was prefigured by statistical political science work proposing supposed demographic, cultural, political and economic cleavages in England.

Antecedents of the towns discourse

In 2014, the political scientists Ford and Goodwin (2014a, 2014b) explained the rise of the right-wing populist party, UK Independence Party, by conceiving a population they termed the ‘left behind.’ In Ford and Goodwin’s (2014a, 2014b) analysis, the ‘left behind’ are ‘older, working-class, white voters who lack the educational qualifications, incomes and skills that are needed to adapt and thrive amid a modern post-industrial economy’ (Ford and Goodwin, 2014b: 278). From analysis of British Social Attitudes survey data, the authors claim that the ‘left behind’ hold socially and culturally conservative views on ‘race and immigration, national identity, gender, rights for same-sex couples, Europe and ethnic diversity’ (Ford and Goodwin, 2014b: 279).

Jennings and Stoker (2016) also posited a similar thesis of ‘Two Englands’ drawn from editions of the British Elections Study surveys. ‘Cosmopolitan’ England is supposedly ‘global in outlook; relatively positive about the EU; pro-immigration; comfortable with more rights and respects for women, ethnic communities and gays and lesbians; and generally future-oriented’ (Jennings and Stoker, 2016: 372). Opposingly, ‘backwaters’ England is ‘inward-looking, relatively negative about the EU and immigration, worried by the emergence of new rights for “minorities” and prone to embracing nostalgia’, ‘more ingrained in a White British and English identity’ (Jennings and Stoker, 2016: 372 and 375). ‘Backwaters’ England is primarily coastal towns and areas, the same as Ford and Goodwin (2014) locate their ‘left behind’ population, and ‘does not tend to include declining towns in Northern England which are being shaped by ongoing processes of deindustrialisation’ (Jennings and Stoker, 2016: 373). However, Jennings and Stoker (2016: 381, added emphasis) also hypothesize ‘varieties of provincial backwaters… reflecting distinct economic and demographic trajectories – such as the economic decline of former northern industrial towns with migrant communities.’

While Jennings and Stoker (2016) and Ford and Goodwin (2014a, 2014b) were specifying a ‘left behind’ and ‘backwaters’, right-wing political strategists were beginning to perceive discontent among working-class Labour Party supporters (Isakjee and Lorne, 2019; MacKinnon, 2020). The Conservative thinktank, Renewal, for example, was founded in 2014 to broaden the appeal of the party to working-class voters, especially in deindustrializing areas in the Midlands and North of England. The edited volume Access All Areas (Skelton, 2014) is indicative of early thinking on working-class voters and their policy concerns, particularly in what later became referred to as the ‘Red Wall’. The founder, Skelton (2014: 6), argued, ‘[a]s the Labour Party becomes… ever more out of touch with its traditional, working class support base, the Conservatives can fill the gap to become the new “worker’s party.”’

Broadly, what these texts claim to identify is a feeling of economic, political and cultural peripheralization along lines of race, age, class status and cultural values among voters in non-metropolitan areas. These significations would subsequently be centralized as academics, journalists, politicians and political commentators clambered to explain the unexpected result of the EU referendum.

Interdiscursive territorializations of the ‘post-industrial town’

The EU Referendum was widely interpreted as substantiating the existence of the ‘left behind,’ their significance in determining plebiscites, as well as placed them (Watson, 2018). Aligning the EU Referendum to elections in Europe and the US, an apparently peripheralized constituency was seen as having ‘used the ballot box to rebel against the feeling of being left behind’ (Rodriguez-Pose, 2018: 190). Returned to in the next section, compelling evidence suggests this reading was misplaced, that the majority of the Leave vote – and other populist movements – did not follow the geographic and class indicators of the ‘left behind’ (Antonucci, et al., 2017; Gordon, 2018). As Sykes (2018: 155) observes, ‘[t]he “left behind” narrative and imaginary… contributes to evacuating other readings of the territoriality of the referendum vote, laying its causes firmly before the door of “Left Behind Britain.”’

Regardless of its veracity, the ’left behind’ and ‘backwaters’ theses conjoined with spatial imaginaries of the ‘Brexit heartlands’ to inform research that further conceptualized geographical distinctions along lines of demographic composition, economic trajectories and political culture (Mackinnon, 2020). A prominent articulation was the opposing tribes of ‘somewheres’ and ‘anywheres’ forwarded by David Goodhart (2017), with ‘anywheres’ standing for the metropolitan, cosmopolitan, liberal elite and ‘somewheres’ encoded as rootedness and communitarian (Rogaly, 2020). It was ‘left behind’, though, that emerged as the preferred rhetorical phrase deployed in media and political discourse. The intensity and widespread support of these discourses, often evidenced by widely circulated colour coded maps of the EU Referendum result, was such that an ‘image of the left-behind Leave voter has taken hold: a poorly educated inhabitant of a faded seaside town or the grim, post-industrial North’ (Surridge et al., 2021: 3).

Work increasingly sought to narrow the geographical extents to which these ‘left behind’, ‘somewheres’ and ‘backwaters’ could be more accurately placed, with ‘towns’ more generally emerging as an early settlement type. In this research, towns were delimited relationally to cities. For example, the New Economics Foundation thinktank (Brett et al., 2017: 16) claimed that the ‘differences not only in economic opportunity but also in demographics, cultural preferences, and lifestyles, are startlingly pronounced.’ Similarly, Jennings and Stoker (2019: 156), in a more territorially defined development of their ‘Two Englands’ thesis, proposed that cities are ‘becoming younger, more ethnically diverse, more educated and more socially liberal – while towns are aging, are less diverse, more nostalgic and more socially conservative.’ Jennings and Stoker (2019: 159) argued that there was a correlation between economic inequality and cultural values, whereby ‘places that have experienced relative decline have become more “closed” on the “open-closed,” or “cosmopolitan-communitarian,” dimension.’ However, as the ONS’s (2019: 3) typology recognizes, ‘some towns are among the most prosperous places in the country for residents to live, while others are sometimes described as “left behind.”’ The emergent towns discourse was not proposing a city-town distinction, but cities in relation to deindustrializing towns and some coastal towns.

In sum, scholarship emanating from academics and thinktanks produced a settlement type of towns encompassing interrelated cultural, social, economic and demographic dimensions in relation to cities: more white, more post-industrial working-class, more communitarian, more socially and culturally conservative, more economic decline. This research was concerned with tracing empirical trends, no matter how dubious or flawed the methods used for generating their claims. This research provided evidence for more homogenizing and politicized territorial productions in political and media discourse (O’Connor, 2017; Skelton, 2019).

Intertextual recontextualization and popular political territorialization

After the June 2017 UK General Election, political constituencies containing deindustrializing towns were increasingly viewed as decisive to electoral victory. The so-called ‘Brexit heartlands’ mapped neatly onto the so-called ‘Labour heartlands’ of historically Labour supporting constituencies in non-metropolitan industrial areas of the Midlands and North of England. In this context, fears within the Labour Party grew that support was being lost among the ‘left behind’ voter type in its ‘heartlands’, and the Conservatives began to increasingly electioneer these Labour constituencies (Bailey and Baston, 2018).

The supposed polarization between cities and towns originating in academic work filtered into this political, as well as media, discourse. For instance, Nandy MP (Nandy and Quilter-Pinner, 2017: p. n) suggested that the EU Referendum ‘exposed the depth of the divide between our towns and cities.’ Supposedly ‘[d]iverse, liberal, fast-changing cities have often been at odds with the stability, (small-c) conservativism and communitarian values that are much more prevalent in towns’ (Nandy and Quilter-Pinner, 2017: p. n). This polarization rhetoric is echoed by Skelton (2019: 61), a prominent Conservative thinker, who expressed that the ‘town-city divide encapsulates this sense of two nations.’ Skelton (2019: 20 and 133) describes towns as having ‘much more of a sense of place, community and belonging than many of those who live in major cities’, arguing that ‘many people in post-industrial towns… still hold a strong sense of patriotism and pride in their country.’

Also, in political and media discourse, rhetorical use of ‘left behind’ phrasing was increasingly detached from its original demographic meaning and recontextualized and applied as a territorial descriptor (Watson, 2018). In a case of interdiscursivity, the territorial use of ‘left behind’ translocated to the nascent towns discourse and reference to ‘left behind towns’ began appearing in media and political texts. Importantly, concurrent to the emergence of the ‘left behind town’ label, was an interchangeable referencing of ‘left behind town’ and ‘post-industrial town’, such that both terms evoked conflated meanings of a white, working-class, social and cultural conservative settlement type. A common practice, particularly in media and journalistic texts, was to detail experiences and conditions in one deindustrializing town and frame the narrative as being representative of a settlement type (Mirror, 2019).

This process of recontextualization and synonymic use of ‘left behind town’ and ‘post-industrial town’ is, in part, explained by electioneering and interparty debates taking place between the June 2017 and December 2019 General Elections. Regards the Labour Party, the towns discourse was mobilized in contestations over the party’s political direction between its left-wing that had ascended to leadership under Jeremy Corbyn and its recalcitrant Blairite right-wing (Watson, 2018; Evans, 2017). In tones more measured than much of the acrimonious exchanges, Nandy (2020: 329) argued that ‘the new “cosmopolitan” radical left in the Labour Party has… led it to downplay the decades of relative decline and acute material concerns of working-class voters in towns.’ For Nandy, city-centred renewal, backed by Labour when in government, has not been equally shared and ‘postindustrial and coastal towns have been hit hardest’ (Brett et al., 2017; Nandy, 2020: 328; Rodriguez-Pose, 2018). A sign that Labour Party leadership began taking the towns discourse seriously, the party launched the ‘Our Town’ campaign in late 2018, which sought to decentre the ethnic nationalist subtext of much of the ‘left behind’ discourse toward a multiracial class experience in the deindustrializing peripheries (The Labour Pary, 2018).

The Labour Party’s ‘Our Town’ campaign looked to challenge the Conservative Party’s efforts to appeal to deindustrializing areas. Conservative rhetoric and clientelist policies subsequently coalesced into the Levelling Up agenda of Boris Johnson’s Government. While often vague, the Levelling Up agenda sought to address, or arguably placate, supposed cleavages seen to underpin Brexit. A precursor policy was the Stronger Towns Fund programme developed during the Theresa May Government. This policy was relaunched as the Towns Fund when Boris Johnson PM took over as leader in July 2019. In an August 2019 speech relaunching the Towns Fund as a flagship policy of the Levelling Up agenda, Johnson (2019, added emphasis) stated that the programme was for ‘towns that feel left behind.’ Johnson (2019) intended to state that the Towns Fund would ‘start answering the pleas of some of [England’s] left behind towns.’ However, the line was omitted in the delivered speech, though it appears in the published official transcript (The Telegraph, 2019).

Within 3 weeks of Johnson’s speech, the political data analyst, James Kanagasooriam, used Twitter to posit the ‘Red Wall’, a ‘cluster of 42 constituencies in Labour’s heartlands which were likely to turn [Conservative]’ (Kanagasooriam and Simon, 2021: 10). The following month, 101 towns were invited to apply to the Towns Fund after a selection process ostensibly based on an index of economic need (Housing, 2019). Following media scrutiny, an official audit determined that towns in marginal electoral constituencies and Government Ministers’ own constituencies were favoured over others ranked as higher need (Chakelian, 2020; National Audit Office, 2020; MacKinnon, 2020).

In October 2019, a General Election was announced for 12^th December 2019. Debates surrounding the ‘Red Wall’ characterized the run-up. Though limited to 42 constituencies, the ‘Red Wall’ was recontextualized into the towns discourse, used to signify any non-metropolitan Labour constituency in the North and Midlands encompassing a deindustrializing town (McCurdy et al., 2020). Alongside Kanagasooriam’s analysis, various texts appeared that retrenched the territorial production of the ‘post-industrial town’ using similar language to that of the ‘left behind,’ such as ‘broken’ and ‘forgotten.’ One such text was Skelton’s Little Platoons (2019), referenced above, which focused on Britain’s ‘forgotten’ towns. Suggesting that towns ‘had become peripheral’ (Skelton, 2019: x) to mainstream politics and policy, Skelton (2019: ix) claims that, in the EU Referendum, ‘[p]eople in the so-called post-industrial towns had made it quite clear that the economic and political settlement was not working for them.’

Similar tropes are reproduced in the Onward policy document, The Politics of Belonging, published in October 2019 (D). For the authors, the 2019 election would be ‘fought around a new voter archetype’ (Tanner and O’Shaughnessy, 2019: 19) labelled ‘Workington Man.’ ‘Workington Man’ and the ‘left behind’ are indistinguishable as:

… likely to be over 45 years old, white, does not have a degree and has lived in his home for over 10 years. He voted to Leave the EU in 2016 and thinks the country is moving away from his views both economically and culturally (Tanner and O’Shaughnessy, 2019: 19).

In identifying the predominant constituencies of ‘Workington Man’, the authors note that ‘there is a high preponderance of towns – rather than rural or metropolitan constituencies – in the North of England’ (Tanner and O’Shaughnessy, 2019: 38).

When the Conservative’s won the 2019 General Election it was widely read as substantiating the towns discourse thesis outlined above, serving to entrench the now-dominate territorial production of the ‘post-industrial town’ (Crampton, 2019; Cutts et al., 2020; McCurdy, 2020; Payne, 2021). In the aftermath, texts premised on understanding the ‘fall of the Red Wall’ appeared that once again centred the symbolic imagery of the ‘left behind’ signifiers (Mattinson, 2020; Mirror, 2019; Payne, 2021). For example, the journalist Richard Crampton (2019: n) writes of a 4-day tour of ‘broken towns’ with ‘grim, miserable spaces’:

I met dozens of voters... If I fed their characteristics and conversation into a computer to come up with an archetype, it would yield a person in his or her mid to late sixties, white, working class, ancestrally Labour but detached — or about to be — from that party.

If territorial stigmatization is predicated on ‘the proximity in physical space of various discredited social categories, and their physical inscription in this space’ (Schultz Larsen and Delica, 2019: 551), the academic, political and media focus on deindustrializing towns analysed above vias toward territorial fetishization. In respects to the historical economic and political peripheralization of deindustrializing towns, this could be tentatively welcomed. However, the ‘post-industrial town’ in its dominant symbolic production is one of a ‘left behind’ imaginary: white, male, older, working-class and socially and culturally conservative. While this symbolic production is not entirely disconnected from social conditions, there are some apparent and potential misrecognitions that often belie competing realities on the ground in deindustrializing towns.

Territorial politics of misrecognition

The towns discourse has helped draw attention to the issue of entrenching socio-spatial inequality and peripheralization rooted in processes of deindustrialization. Economic and labour geographies are more complex and variegated than the dominant towns discourse suggests, however. Numerous studies have highlighted how the most economically disadvantaged areas are in major urban geographies with multiethnic populations (Furlong, 2019; ONS, 2019; Pike et al., 2016). Additionally, Beatty and Fothergill (2020a, 2020b) provide an analysis of labour markets in English local authorities containing deindustrializing towns. Though we should be cautious of local authority-level analysis for reasons discussed below, Beatty and Fothergill (2020a: 11) suggest that despite challenges in earnings and job growth, particularly higher-paying jobs, labour markets in deindustrializing peripheral areas are ‘not as distressed as might have been feared’ and not ‘locked into a spiral of decline.’ Moreover, industrial employment still accounts for a higher proportion of employment in these local authorities than the national average (Beatty and Fothergill, 2020b). The narrative of declining towns and affluent cities proposed by the towns discourse, then, is ‘basically untrue in terms of the economic features of the UK’ (McCann, 2019: 265).

However, while highlighting the economic marginalities of deindustrializing peripheries, more urgent misrecognitions of class, race and cultural politics are foregrounded within territorializations of deindustrializing towns. These misrecognitions have arisen from a combination of the research agendas and methods used to measure and delimit towns, which contain critical limitations, and the political and media recontextualization of the data elicited in partial and generalized forms. The intention of political science research on the issue has mainly been to trace changing electoral landscapes and voting trends along demographic, cultural and socioeconomic lines. This is distinct in important ways to political and media discourses that have often elided the purposes and qualifications of political science narratives to generalize the political cultures of deindustrializing towns based on recent voting trends. This overlooks the historical complexity of placed-based political cultures (Boswell et al., 2020).

For instance, only 30 of 42 ‘Red Wall’ constituencies overturned Labour incumbents and voted in Conservative candidates in the 2019 General Election, some of which on marginal majorities (Payne, 2021). The proportion of Conservative voters is reduced further when measured against the full electorate, including non-voters, and total population, including those not registered or allowed to vote. Similarly, the vote to leave the EU was more complex than prevailing representations in media and political discourse. Analysis by Green and Pahontu (2021: 4) combining the British Election Study and the Bank of England’s 2016-2019 panel surveys found that support for Brexit was highest among wealthy voters and ‘less wealthy Britons were less likely to support a status quo change.’ Surridge et al., (2020) find that these ‘Comfortable Leavers’ accounted for almost half of all Leave voters. Moreover, though highest in some districts dominated by deindustrializing towns, and on comparatively high turnouts, district-level Leave votes rarely constituted majorities of total electorates once non-voters are factored in, and even smaller against total population (The Electoral Commission, 2019).

There are also crucial weaknesses in the socio-technical practices of mapping quantitative data used in the territorial production of the ‘post-industrial town.’ No classification of a deindustrializing town as an areal unit exists for which statistics are collected and deployed (Beatty and Fothergill, 2020a). Research has mitigated this issue by using constituency or local authority level data, which, as the ONS (2019: 4) highlight, ‘are not suitable for investigating the topic of towns’ (See also: Jennings and Stoker, 2019). Constituency boundaries are, at least, ostensibly, determined by population thresholds. Deindustrializing towns are often included in constituencies with multiple dispersed rural places. This is also the case with local authority boundaries, which encompass a variety of settlement types.

Most importantly, much of the research relies on Likert-type and ordinal scale survey data to make claims about the social and cultural values and lives of deindustrializing towns. Jennings and Stoker (2017) are attentive to the denigrating connotations of terms used to describe what towns are and are not. Supposed prevalence based on nebulous antonymic scales, such as communitarian-cosmopolitan (Jennings and Stoker, 2019) and autonomy-belonging (Tanner and O’Shaughnessy, 2019), rests on values-based conceptualizations of the opposing poles. Ethnographic research conducted in deindustrializing towns has documented working-class participants articulating their own senses of ‘non-elite cosmopolitanism’ and social liberalism including the practices of multiethnic conviviality ranging across community arts events, workplace relations and everyday sociality documented by Rogaly (2020) and ‘the importance of [participants’] interethnic and interracial friendships and networks’ highlighted by Rhodes et al. (2019: 30).

Though levels may vary, deindustrializing towns have long had multiethnic working-class populations, with histories of international mobility throughout periods of (de)industrialization. Notable examples include former ‘mill towns’ in Yorkshire and Lancashire, as well as deindustrializing towns in the West Midlands. Electorates may vote along lines of racial identity, socioeconomic status or educational attainment, and indicators within these dimensions may be more prevalent in certain deindustrializing towns relative to cities. However, the symbolic production of ‘post-industrial towns’ as an essentialized white working-class ‘left behind’ wilfully ignores the ‘stake and claims’ (Rhodes et al., 2019: 8) of people of colour and migrant populations to such places (Isakjee and Lorne, 2019; Rogaly, 2020; Virdee and McGeever, 2018).

The effect of this territorial misrecognition is to marginalize competing cultural and political concerns among the existing populations of deindustrializing towns, rendering them unserved by clientelist policies formulated to consolidate support among archetypal ‘left behind’ voters. The symbolic production of deindustrializing towns as enclaves of a reactionary (white) working-class works to both corrode social identities and relations, disrupting new forms of class solidarity and politics from emerging, and legitimizes ethnic nationalisms at the deindustrializing margins, creating geographical frontiers along racial lines (Virdee and McGeever, 2018). These factors, in turn, risk worsening the historically rooted economic and social disadvantages that deindustrializing towns undoubtedly face.

Conclusion

Demonstrating the ‘ascriptive power of macro-discourses fuel[l]ed. by data-based descriptions’ (Meyer and Miggelbrink, 2013: 213), the territorial production of the ‘post-industrial town’ has involved discursive and socio-technical practices by ‘specialists in symbolic production’ across academic, journalistic and political fields. Much like Furlong (2019) argues, correlations between economic decline and social and political values depends on how ‘left-behindness’ is conceptualized. Couched in significations of the ‘left behind’ originating in political science work, the towns discourse has produced a dominant symbolic territorialization of a white, older, working-class, non-cosmopolitan and socially and culturally conservative settlement subtype. Such a symbolic production rests on an equally misrecognized territorialization of the ‘city’ as uniformly progressive, vibrant, cosmopolitan, diverse and affluent.

The towns discourse has been reacting to a succession of unexpected national and international plebiscites and events that, justifiably or not, propelled England’s deindustrializing towns into the centre of the UK’s political and policy arena. Territorial production of settlement types perhaps develops at a more gradual pace in other cases. Moreover, in these other cases, symbolic power is perhaps more horizontally shared, with that of ‘specialists in symbolic production’ mitigated by historically informed formulations circulating in everyday life and different fields of action. An omission here are such ground-level discourses, particularly those emanating from within deindustrializing towns (Schultz Larsen and Delica, 2019). Indeed, pre-existing valuations of places that have come to be classified as ‘post-industrial town’ have only been alluded to here, yet have likely shaped recent territorializations. Analysis has also relied upon written texts, neglecting visual and audio texts, as well as social media. Discourse is, of course, multi-modal and future work on the territorial production of settlement types might well incorporate more democratic forms of communication beyond fields exercising surpassing symbolic power and how these fields and actors may interact and contest over time.

These agendas will be advanced by paying attention to the politics of territorial(izing) productions. Though the towns discourse has advanced awareness of peripheralized geographies, the preoccupation with the concerns, values and politics of problematic ‘left behind’ significations has further peripheralized those that are not white, older and socially and culturally conservative who have equal stake in the presents and futures of deindustrializing towns. Put simply, the social, political and cultural dimensions of the dominant symbolic production are not nearly as pervasive in the places the towns discourse has sought to delimit and classify. This politics of misrecognition further circumscribes shared class experience of marginality and disadvantage across geographies in favour of narrow and misleading symbolic productions of settlement types.