Stuck between the EU ‘rock’ and UK ‘hard place’? Northern Ireland as a liminal space after Brexit

ABSTRACT

Northern Ireland hardly featured as an issue in the historic 2016 Brexit referendum campaign in the UK. Subsequently, however, it became the main zone of contention between the EU and UK. This article examines Northern Ireland's experience since the Brexit vote through the theoretical lens of liminality. It does so, specifically, by focusing on the three strands of relationships at the heart of Northern Ireland's 1998 peace agreement, and how Brexit impacted on each of these relationships. It argues that Brexit proved profoundly unsettling to the existing political order and triggered significant angst and fear in Northern Ireland.

KEYWORDS:

• Northern Ireland
• Brexit
• liminality
• peace process

Introduction

One of the most ironic elements of the difficult aftermath of the 2016 Brexit referendum was the fact that the referendum campaign had been almost devoid of any discussion of Northern Ireland, and the impact UK withdrawal from the EU might have on all three ‘strands’ of territorial relationships at the heart of the 1998 Good Friday Agreement (GFA). These strands consist of: (1) the relationship between the nationalist and unionist communities in Northern Ireland (the internal strand); (2) between Northern Ireland and the Republic of Ireland (the ‘North–South’ strand); and (3) between the United Kingdom and Ireland (the ‘East-West’ strand; Cochrane, 2020; O’Brennan, 2019). In Murphy’s (2021) assessment, the referendum campaign was ‘an Ireland-free zone’. The UK’s decision to leave the European Union led to a renewed contestation of spatial and territorial relationships which carry profound implications for the political geographies and constitutional futures of both the United Kingdom and Ireland (Walsh, 2019). The vote almost immediately brought the fragile peace process in Northern Ireland back into sharp focus. The ensuing negotiations, intended to achieve a durable Brexit settlement proved extremely unsettling to Northern Ireland and its fragile territorial-constitutional arrangements.

In evaluating the impact of Brexit on Northern Ireland, this article adopts as its key frame that of liminal territorial politics. Brexit shone an uncomfortable light on the ‘in-betweeness’ of Northern Ireland: the ‘constructive ambiguity’ associated with the Good Friday Agreement gave way to what some scholars regard as a ‘destructive ambiguity’ (Coulter et al., 2021) associated with the Northern Ireland Protocol (later replaced by the Windsor Framework). Where the Agreement softened the edges of territorial-constitutional contestation, Brexit significantly amplified anxieties and encouraged renewed patterns of re-tribalization and siloization. Disputes about the desired constitutional endpoint of the peace process returned to complicate politics and community relations. The set of impulses around ‘de-bordering’ which accompanied and were accelerated by the 1998 Agreement, mutated into calls for ‘re-bordering’, as the Brexit negotiations rumbled on (McCall, 2018). This article interrogates the nature and importance of this return to liminal spaces and identities, and how the re-appearance of ‘ontological dread’ has complicated the search for a satisfactory post-Brexit settlement in Northern Ireland. While the Good Friday Agreement succeeded in achieving peace, it is much more doubtful whether it has achieved reconciliation, its overarching aim (Tannam & Kelly, 2022). As Coulter et al. (2021, p. 164) remind us: ‘Northern Ireland is not a society at ease with itself’.

To understand Northern Ireland as a peripheral and disadvantaged region throws light on the nature and significance of its liminality. Indeed, it could be argued that Northern Ireland suffers from multiple and inter-locking peripheralities: it is peripheral within the UK, peripheral on the island of Ireland, and peripheral within the European Union. It is both a place and a ‘non-place’. It is both a ‘place apart’ and ‘integral to our precious Union.’ It ‘is’ and ‘it is not’. It is ‘part of’ and ‘apart from.’ It is ‘of’ Ireland but not part of the Republic. It is both static and moving, paralyzed and yet evolving, defined by spectral imagery and ghostly avatars of the past (Coulter, 2021, 2022). It was both inexcusably absent and manifestly present in the Brexit negotiations. The nascent contentiousness of Brexit meant there was no escape from the political even for those who consciously choose to define themselves as neither nationalists nor unionists. Life is increasingly difficult to envisage in linear or sequential terms. Instead, the worst of the past has returned to occupy space within individual and collective imaginaries.

Thus, the relatively positive equilibrium of the post-1998 period in Northern Ireland gave way, after 2016, to an increasingly precarious disequilibrium, perhaps even the kind of ‘schismogenesis’ defined by David Graeber and Wengrow (2021), reflected in periodic rows between Dublin and London or Brussels and London, with Northern Ireland caught in the middle.

Theories of liminality

The theoretical specification of liminality was associated initially with the discipline of anthropology. Originating from the Latin word for ‘threshold,’ liminality portrays an equivocal, in-between state which conveys the possibility of changed lived realities in specific contexts (Murphy & McDowell, 2018). Liminality is both inclusive and exclusive: it both unites and divides; it binds people together and keeps them apart. The term was first used by Arnold van Gennep in 1960/1909 in Les Rites de Passage, describing processes of social change. Later, Victor Turner (1967) expanded on van Gennep’s ideas, taking them beyond use in mere ritual spaces to a point where in-betweenness constituted a stage of transformation of social categories. Turner, however, conceived of this transition process as a positive-sum experience and one where transition to a positive endpoint could be achieved. Here liminality constitutes a ritualized process of transformation in which actors emerge from liminal spaces transformed and refashioned (Montgomery, 2010; Turner, 1967).

Earlier applications of the term, which related specifically to the rites observed within tribes and cultures, have given way to more expansive social scientific meanings of the term. Tradition, culture and anchors of identity are deemed especially worthy of investigation. These contemporary explorations and interpretations of liminality, within a range of social sciences, highlight patterns of anguish, fear and even existential dread associated with liminal spaces and beings (Horvarth, 2013). ‘Liminars’ are thought of as entities that are between allegiances (Balduk, 2008) and transition is a key frame for understanding the liminal condition. Echoing Galtung’s (1969) notion of ‘negative peace,’ Roger MacGinty (2006) describes periods of transition from conflict as an often grudging hiatus in violent conflict ‘crowned with an internationally supported peace agreement that finds little approval at home after initial enthusiasm has worn off.’ But most theorists agree that the liminal condition is not one that evolves out of specific choices of individuals or groups. Rather, it is often experienced and not the result of specific acts of localized agency.

We can thus define liminality as a condition of ‘in-betweenness’ which manifests in spatial, temporal and cognitive phenomena in contested settings. Applied to situations of ethnic conflict and post-conflict social and political orders, liminality adds theoretical value to the exploration of issues of territorial borders, constitutional provisions and models of identity change. It is especially useful as a frame for considering modes and spaces of disorder and exclusion, and the disruption caused by exogenous events on those orders and identities. The sections below explore the territorial and spatial dimension, as it has evolved in Northern Ireland after the 1998 Peace Agreement and, in particular, following the Brexit vote in 2016.

Given the fragility of many peace settlements, liminality is a device increasingly of interest to social scientists interested in the way those settlements develop, and whether they produce genuine change and transition to more settled constitutional, identity and territorial orders. Navaro-Yashin (2003), for example, engages with liminal individuals in conflict-torn Cyprus (Montgomery, 2010) In these kind of conflict settings, war zones are constituted simultaneously as sites of coming together and falling apart, of spaces turned precarious and (often) violent and predatory (Montgomery, 2010) but where different forms of co-existence can emerge. The persistent nature of the liminal experience in Northern Ireland (comparable in different ways to the experience of Bosnia and Herzegovina after the 1995 Dayton Agreement) draws our attention to the notion of liminality as an individual and collective, as well as longitudinal, experience of ambiguity and in-betweenness where there may be no ‘endpoint’ but rather an endless churn of actors, institutional breakdowns and ‘negative peace’ (Galtung, 1969; Murphy, 2018), effectively a tortured journey towards an unknowable destination (Murphy & McDowell, 2018). In both Northern Ireland and Bosnia, peace has been ambiguous, contested, fraught with periodic tensions and isolated episodes of violence, but with persistent fears of a return to the worst of the past. Victor Turner described this as a form of in-betweenness best captured as ‘betwixt and between,’ of a society no longer experiencing large-scale violence, but yet unable to maintain a sustainable peace. This framing of Northern Ireland has been increasingly evident in scholarly literature (Coulter et al., 2021; Murphy, 2018) which can be interpreted as a kind of ‘Groundhog Peace.’ Murphy (2021) asserts that although a ‘negative peace’ is not unexpected in a region engaged in transition from significant contestation and violence, ‘it is nevertheless a precarious and tenuous moment for any conflict transformation process.’

Territorial and spatial liminality

Territory is commonly understood to refer to politically bounded physical space, or, in other words, areas of land or sea under the jurisdiction of a particular ruler, state or regime (Walsh, 2019). Within human geography, the concept of liminality gained currency after Rob Shields (1991) developed ideas connecting physical space and marginality. According to Victor Turner (1967), attributes of liminality are necessarily ambiguous, since this condition and these groups within society elude conventional classifications. In this manner, people become outsiders, kept at a distance, even from their co-nationals or co-religionists in neighbouring jurisdictions (Balduk, 2008). The relationship of Northern Ireland’s nationalists with the Republic of Ireland, and unionists with the United Kingdom, could well be said to conform to such an understanding: people both belong, and don’t belong, simultaneously. Northern Ireland might thus be thought of as a place where people seek a sense of place within a ‘non-place’ (Balduk, 2008), where everyone is an insider and outsider simultaneously.

This is the sense in which liminal territory is identified as an uncomfortable place to live in, as precarious and potentially violent (Shannon, 2020). These are physical spaces that frequently exhibit contradiction, contestation, hybridity, social exclusion and vulnerabilities (individual and collective). It is in border regions that we might expect to locate finite limits to nation state prerogatives and bounded territoriality (Walsh, 2019). In this environment, the everyday activity of individuals, households and communities no longer map neatly on to nation-state borders, if indeed they ever did. Connected to this, spatial imaginaries have been defined as ‘deeply held collective understandings of socio-spatial relationships that are performed by, give sense to, make possible and change, collective socio-spatial practices’ (Davoudi, 2018; Walsh, 2019). Although European integration didn’t introduce the notion of supranational European space (think of the Hapsburg or Ottoman Empires over many centuries), it did help solidify a certain idea of Europe, one which was multicultural, cross-national, multi-faith, a transnational space where borders were more fluid than the hard borders associated with nation states. In this reading, the recent history of European integration may be interpreted as a series of concerted, if indeed partial and incomplete, efforts to frame European space in non-state-centric terms. This model imagines Europe as a space characterized by coherent functional regions, and where formal nation-state borders are increasingly irrelevant (Walsh, 2019). Despite this step in the direction of cosmopolitanism, the empirical reality is that Europe is still constituted as a territory of nation states, even if the precise form of the EU’s constitutional and institutional architecture exhibits significant ambiguity. Prior to Brexit, the ambiguity of EU integration complimented significant ambiguity in Northern Ireland arising from the region’s unique historical specificities and the institutional model put in place by the Good Friday Agreement. After Brexit, the tension between the British and EU models of regional integration became much harder to reconcile. These tensions crystallized around where the post-Brexit physical border might be located, with the choice being between: (1) a land border on the island of Ireland, separating Northern Ireland from the Republic of Ireland (unacceptable to nationalists and the government of the Republic); and (2) a sea border between Northern Ireland and Great Britain (unacceptable to most unionists). The next section examines the role of the border in the outworking of Brexit.

The border

The border on the island of Ireland itself constitutes the very definition of ‘in-betweenness’ and ambiguity. Borders and borderlands are fundamentally ambivalent, defined by deep contradictions and contestation. They are almost mystical to some (Kassabova, 2017), the focus of hopes as well as fears, thrilling scenarios and banal, everyday exchanges. In some, they inspire utopian dreams of cooperation, reconciliation and peaceable coexistence. For others, they are elegiac landscapes of hatred, regret, revenge, separation and transgression. In some instances, they signal keeping faith with essentialist projects of identity-construction and projection. In others they engender apprehension as sites of identity absolutism.

Borders are often presented as formal lines of separation and defence which distinguish between what ‘belongs’ and ‘what is foreign,’ between ‘us’ and ‘them,’ between ‘self’ and ‘other’ (Hayward, 2018). Borders create barriers to the free movement of people, goods, and services and are ‘intrinsic to everyday life’ (Popescu, 2011). As Agnew (2008) suggests, borders ‘not only limit movement of things, money, and people, but they also limit the exercise of the intellect, imagination and political will’. Borders exhibit common characteristics as much as they also demonstrate unique specificities. Even in the orthodox Westphalian meaning of the term, borders are, however, ‘both spatially and temporally uneven’ (Radil et al., 2021). For Komorova and Hayward (2019), border regimes constitute ‘networked systems of rules that regulate behaviour and are activated by the mobility of people and things at a variety of scales’.

Borders thus have to be understood as multi-faceted and multi-layered entities which emerge from specific bordering processes, everyday discursive, institutional and societal interactions and localized practices, that have the effect of constructing, maintaining and reproducing categories of spatial and cultural difference (Hayward, 2018; Scott, 2015). The summary logic of such is that borders are conceived not just in terms of brute material characteristics but increasingly as dynamic, fluid and functional processes (Hayward, 2018; Scott & Van Houtum, 2009). Rumford’s (2008, 2012) ‘multiperspectival approach’ presents borders as produced and ‘performed’ through a multiplicity of bordering practices and ‘borderwork.’ Borders can be manipulated to both transform and/or fix geographic territory, physical security, ethnic identities, political dominance, as well as communal emotions and memories (Hayward, 2018; Paasi, 2012).

Anne Norton (1988) equates liminality with border constructions which concretely identity and specify territorial demarcations (cited in Balduk, 2008), thus establishing a link between liminality and state borders. Although it could be argued that liminality might just as well be linked to nations, Balduk argues that given the central importance of the state to modern politics, and the formal prerogatives that derive from statehood, it makes much more sense to link liminality to the state as the territorial sovereign. In territorial terms, liminal entities live – literally – at/on the border, ‘betwixt and between’ different geographically contiguous state units. They are frequently situated at a physical remove from the metropolitan, and can exhibit significant evidence of poverty and impotence associated with distance from power structures and spatially concentrated economic resources. And although the European Union has often been presented as emblematic of a ‘borderless Europe,’ as Radil et al. (2021) point out, the COVID pandemic saw a very significant amount of re-bordering across Europe. While the EU aspired rhetorically to the notion of a post-Westphalian, post-sovereign space, in reality the COVID re-bordering efforts were consistent with a much longer timeframe in which borders were far from disappearing in Europe. Long-standing concerns about the effects of globalization and resistance to refugee flows are often cited as influencing this phenomenon of re-bordering in Europe. For Gaeta (2021), this is indicative of the Janus-faced character of EU borders, ‘or the simultaneous inward and outward projection of seemingly opposed political axioms of inclusion and exclusion.’

Northern Ireland occupies a unique geographical position – part of the island of Ireland, and yet politically part of the United Kingdom of Great Britain and Northern Ireland. In the post-Troubles era, however, a border which was solid, militarized and physically intimidating became something porous ‘that runs along hedgerows, across loughs, through fields, in a way that, at least locally rarely calls attention to itself’ (Carr, 2017). The border is 500 km long, and, as McCall points out, ‘has hundreds of crossing points.’ This all adds up to the most dense cross-border agglomeration of settlements in Europe. Additionally, as McCall demonstrates, ‘it is unruly: key arterial routes can cross the border more than once.’ He provides the example of the road between Cavan town (in the Republic) and Dungannon (in Northern Ireland) through the Drumully Salient, which crosses the border no fewer than five times.

About 25,000 people are estimated to be regular cross-border commuters who work or study across the border on a continuing basis (Taillon, 2018). After 1998, the peace process led to secondary roads being re-opened and those sections of a highly militarized border being gradually and incrementally de-securitized, mainly through the dismantling of British army security bases and towers along the border (McCall, 2018). And yet, as Shannon (2020) points out, for almost a century it has remained an international border, represented on official maps, each side a separate jurisdiction, indeed separate countries but ‘apart on maps … never a line’ (Carr, 2017). For Hayward (2018), the great innovation of the Good Friday Agreement was about careful and targeted compromise:

to accept the validity of both Irish nationalism and British unionism, each of which have competing views about the legitimacy of the Irish border, but to weaken the violent potential of these competing ideologies by reducing the actual significance of the Irish border in day-to-day terms.

Despite the existence and importance of formal state boundaries, cross-border governance arrangements often go beyond the limited objectives set for them. Local and regional actors, driven by functional imperatives and ‘delivering for the locality/region’ often seek to circumvent formal boundaries and rules and the constraints they impose on communities. They do so through what may begin as novel but subsequently become routinized practices and these can facilitate the development of regions as spaces of economic, social and cultural exchange, where actors seek to develop and foster functional relationships across formal state boundaries (Walsh, 2019). Komorova and Hayward argue that the ‘the best way of understanding the (Post Agreement) UK-Ireland border is as a site of integration and cooperation, not as a crossing point between jurisdictions.’ Viewed from the perspective of the island of Ireland, the border both exists and does not exist. In this fashion, McCall (2018) depicts the border as a ‘bridge’ that demonstrates:

A significant degree of cross-border contact, communication and cooperation across public, private, and Third (voluntary and Community) sectors. The outcomes of this contact, communication and cooperation include institutional and policy development, economic initiatives with mutually beneficial outcomes and community development across the border.

Further, Komorova and Hayward (2019) understand the UK-Ireland border as ‘both an edge and a fissure,’ in a context where state borders are fundamentally traversable and where there are important ‘entanglements and contradictions in the ways in which different groups are treated within these islands of Britain and Ireland.’ Murphy (2021) states that the Agreement ‘largely settled the contestedness of the border and both unionists and nationalists were at peace with its existence, its operation, and its meaning.’ The image of state territory as internally cohesive, bounded ‘container space’ where the identifiable economic, social and political boundaries coincide at the borders of the nation state (Walsh, 2018) was one that advocates of Brexit emphasized repeatedly. The notion that the state needed to maintain the constitutional and territorial integrity of the UK was one of the most prominent tropes in the discourse of Brexiteers, and, in Northern Ireland, of the Democratic Unionist Party (Walsh, 2019). But that core Brexit idea of taking back border control to specific, formally defined and inviolate physical boundaries, coterminous with nation state frontiers, ran counter to the spatial imaginaries associated with the Good Friday Agreement that had enabled the emergence of nuanced both/and approaches rather than the either/or associated with the previous era of bounded territoriality which vexed Northern Ireland. The vehemence with which Brexiteers focused on a narrowly-defined notion of sovereignty as a core concern in the negotiations meant that, as Walsh (2019) argues, Brexit brought a return to the zero-sum territorial and identity dynamics of the past and threw the logics of the 1998 accord into severe doubt.

Northern Ireland’s spatial demarcations predate Brexit (Coulter et al., 2021). Shannon, for example (2020), argues that nationalist and unionist communities have lived an increasingly separate existence – living, learning, working and socializing in degrees of isolation from each other in what has been described as a ‘ratchet effect,’ where islands of segregated living increased sharply during periods of heightened unrest and did not necessarily decrease much in the post-conflict phase in Northern Ireland (Boal, 2008; Shannon, 2020; Smith & Chambers, 1991). In the 2011 census, for example, approximately 76% of Belfast’s population lived in ‘severely segregated super output areas’ (where the population was 90% or more ethnically homogenous). Only 6% of Northern Ireland’s young people were recorded as being educated in non-segregated schools (Gray et al., 2018). A decade later the picture had barely changed, with 93% of young people still attending segregated schools (Belfast Telegraph, 23 November 2021). North Belfast might be viewed as a microcosm of Northern Ireland. There, the very proximity of communities to each other promoted the construction of ‘peace walls’ and interfaces that are not merely boundaries between communities but important instruments in the definition of discursively marked space. These spatial demarcations, in turn, generate and/or reinforce both physical boundaries and boundaries of the mind. For Shirlow and Murtagh (2006, p. 58), the ‘spatialisation of fear, violent resistance and the desire to promote discourses of internal unity while under threat, has redefined, rebuilt and delivered more impassioned forms of space-based identity.’ In this sense ‘fear, anxiety, hatred and aggressiveness’ feature centrally in their account of life in Belfast (Cash, 2017).

The territorial relationships

As a function of the Good Friday Agreement, the island of Ireland emerged as a common, shared space, one that implicitly acknowledged and accommodated British and Irish identities. Strand One set up a power-sharing executive, based on a consociational model of decision-making (Tannam, 2021), including a 108-seat (later reduced to 90) Northern Ireland Assembly, devolved from Westminster, and a power-sharing Executive composed of major parties representing both major communities (and later the cross-community Alliance Party). These two institutions form the core of the Agreement’s intra-Northern Ireland architecture (Tannam & Kelly, 2022).

Strand Two deals with North–South relations based on an inter-ministerial model with the North–South Ministerial Council (NSMC) at its core, and includes a number of implementation bodies to discuss matters of mutual interest on either side of the border, including cross-border policy initiatives (Murphy, 2021; Tannam & Kelly, 2022). The NSMC has focused on North–South cooperation in agriculture, education, environment, healthcare, transport and tourism and brought together ministers from both jurisdictions with responsibility in these sectoral areas. Implementation bodies promote cooperation between North and South in the areas of business development, food safety, language programmes, trade, waterways and EU funding programmes (McCall, 2018). In this respect, Strand Two reflects the centrality of the border to the conflict and, in turn, to the peace process and the hope that functional cross-border cooperation could be increasingly fostered (Taillon, 2018). The operation of Strand Two is directly dependent on Strand One under the terms of the Agreement. This means effectively that the NSMC cannot function if the Assembly and Executive are not functioning (Tannam & Kelly, 2022).

Strand Three constitutes the East–West territorial axis of the Agreement and provides for regular meetings between the UK and Irish governments, the devolved institutions in Northern Ireland, Scotland and Wales, as well as the governments and executives of the Isle of Man and the Channel Islands in the British–Irish Council (Taillon, 2018). The Northern Ireland peace process was ‘built on the back of close British–Irish cooperation’ (Murphy, 2021; Tannam, 2001). A standing British–Irish Intergovernmental Conference (B-IIGC) was also established to deal with the ‘totality of relationships,’ develop formalized cooperation between the two governments and meet frequently at summit level, with both prime ministers in attendance (but normally at ministerial level) (Taillon, 2018; Tannam, 2021). The Agreement specifically acknowledges the particular interest of the Irish government in Northern Ireland. The B-IIGC was meant to constitute an institution contributing to the real stability of the peace process, ‘ensuring that both sides can act as guardians of the Agreement and protect the different community interests in Northern Ireland.’ A core aim of Strand Three is to ensure that both governments develop a neutral approach that is stable, predictable and builds on unionists’ and nationalists’ interests fairly (Tannam & Kelly, 2022). Crucially, the Agreement stated that all of the constitutional and institutional arrangements, including the Assembly, the North–South Ministerial Council and attendant implementation bodies and the B-IIGC were ‘interlocking and interdependent’ (Tannam, 2021).

The liminal condition of Northern Ireland is evidenced by the fragility and even breakdown of the three strands of the Good Friday Agreement as a consequence of Brexit. Strand One experienced difficulties in relations between representatives of the two communities very quickly after the signing of the Agreement. This predates Brexit but was undoubtedly worsened by developments after 2016. The early problems revolved around demilitarization and resulted in a suspension of the institutions from 2002 to 2007 (Tannam & Kelly, 2022). Eventually, in 2006, a refinement of the Agreement led to the restoration of the Assembly and Executive, after difficult negotiations.

Arguably, the consociational model which the Agreement introduced, especially the model of power-sharing in the Executive, served to entrench polarized identities as, over time, parties representing the harder edge of ethno-chauvinism (the Democratic Unionist Party and Sinn Féin) gained significant electoral purchase from the Agreement (Walsh, 2018). The repeated suspension of the institutions led some to argue that the conflict in Northern Ireland has only been managed and contained, rather than producing any substantive ‘positive sum’ gains in the years after 1998 (Guelke, 2003; Hall, 2018). Thus, one can suggest that, although the Agreement did encourage a broadening of the horizons of local entities, the consolidation of the peace process proved extremely challenging. The decade after 2007 ‘was remarkably stable compared to what came before and after it. Notwithstanding frequent disagreements and standoffs, the DUP and Sinn Féin were able to lead the Executive for a sustained period without interruption’ (Tannam & Kelly, 2022). Brexit subsequently rendered that process a liminal one again, as the relationship between nationalists and unionists came under severe strain. Indeed, with each new crisis in the Brexit negotiations, the relations between nationalist and unionist political parties deteriorated further, often exhibiting the kind of discourse associated with the worst of the ‘Troubles era’. Northern Ireland thus reverted after 2016 to an ‘in-between’ space, a liminal boundary where fear-based discourse returned to emphasize national loyalties and identities (Walsh, 2019).

Northern Ireland’s liminal condition is also evident in the increase in the number of people identifying as ‘others’ in surveys of identity. These people identify neither as British or Irish, unionist or nationalist. The 2021 census showed that 32% of people identified as ‘British-only’, while 29% identified as ‘Irish-only’. Almost 19%, however, declared themselves as ‘Northern Irish only’. Since the previous census in 2011, the number of people identifying as ‘British-only’ declined by 8 percentage points (Carroll, 22 September 2022). In addition, the increase in the number of Polish, Lithuanian, Romanian and other non-British and Irish nationals means that there is now a sizeable cohort of approximately 5% of the population whose identity does not correspond with the binary model at the root of the Good Friday Agreement (NISRA, 2023). Increasingly, the institutions set up by the Agreement, in particular the Assembly and the Executive, look ill-equipped to reflect and provide for the increasing diversity of Northern Ireland. Those who fall into the ‘others’ category undoubtedly experienced greater insecurity after Brexit, as the protracted rows about how to solve the border problem triggered the return of the traditional identity binaries to dominate politics.

One final issue which reflects the increased liminality of Northern Ireland was the repeated failure of both the EU and UK to properly accommodate a democratic voice for Northern Ireland in the Brexit negotiations. The suspension of the institutions in Northern Ireland for long periods certainly did not help. But the arrangements put in place for democratic input by Northern Ireland representatives, in particular the ‘consent mechanism’ built into the Northern Ireland Protocol look like feeble post-hoc attempts to provide some form of legitimation to the process. For Hayward et al. (2020), this means that:

Northern Ireland is at risk of being subject to legislation coming from both Brussels and London without full sight or scrutiny of it … there is no automatic means by which Northern Ireland’s devolved institutions will either be able to effectively scrutinise and shape EU law they need to download or be able to upload views to the EU.

The problems exhibited by the Strand Two dimension of the architecture flowed from the suspension of the key institutions in Northern Ireland for long periods and also predate the Brexit vote. The interlocking and interdependent nature of the three strands and their associated institutions meant that the problems in Strand One inevitably caused problems in Strands Two and Three. The hostility of the Democratic Unionist Party to cooperation with the Republic of Ireland meant that it proved very difficult to hold meetings of the North South Ministerial Council after 2016 (Tannam, 2021). Although the remit of the NSMC and associated bodies is modest, there has been consistent Unionist resistance to the deepening of North–South relations. This resistance, it is argued, is based on a fear of institutionalizing cross-border relations to any degree than might compromise the Union (Murphy & Evershed, 2021; Tonge et al., 2014). The collapse of Stormont in January 2017 – prompted by a scandal concerning the abuse of public money deployed in a renewable energy scheme (the so-called ‘Cash for Ash’ scheme) – led to a suspension of Strand Two institutions, ‘a move which eliminated opportunities for any form of structured dialogue – including on Brexit and its implications – between representatives of the Northern Ireland Executive and Irish government’ (Murphy & Evershed, 2021).

Whitten (2022) argues that the provisions of the Northern Ireland Protocol that specifically relate to North–South relations fall into two categories. First, at the functional level, those that ‘facilitate cross-border trade in goods, thereby avoiding any need for physical infrastructure at the land border’ and, second, those that in the wider sense, aim to ‘maintain the necessary conditions for North–South cooperation.’ The EU and UK aimed to avoid any erection of physical infrastructure on the land border between Ireland and Northern Ireland. Set against this, however, the EU aim of preserving the integrity of the Single Market and customs union in the new dispensation had to be balanced against the UK’s desire to leave both the Single Market and Customs Union without simultaneously cutting Northern Ireland off from the UK internal market (Whitten, 2022). This required a very delicate balancing act.

Strand Three (British–Irish relations) has also been deeply impacted by Brexit which created a ‘vast array of unintended and complex consequences’ for the British–Irish relationship (Tannam, 2021), In particular, the removal of the EU scaffold of shared UK-Irish membership of the EU between 1973 and 2016 proved deleterious to UK-Irish cooperation. Indeed, Ireland and the UK cooperated closely on a range of important issues in Brussels over the span of 50 years of shared membership, including competition policy, tax and economic affairs more generally (O’Brennan, 2019, 2021; Tannam, 2021). The intensity of British–Irish exchanges in Brussels is reflected in the testimony of one former Irish ambassador that, during the period of shared membership, there were on average 25 meetings a day in Brussels where Irish and UK representatives encountered each other. Brexit ‘turned such common interests into competing ones’ (Tannam, 2021).

The negative subsequent trajectory of relations between Dublin and London has been all the more disquieting as it followed a period of pronounced transformation in UK-Irish relations. Queen Elizabeth II’s deeply symbolic visit to Ireland in May 2011, when she laid a wreath at the graves of the IRA leaders during the Irish War of Independence struggle against British rule in 1919–1921, was one important source of this transformation. President Michael D. Higgins’ reciprocal state visit to the United Kingdom in April 2014 was equally significant (Gillespie, 2021; O’Brennan, 2019; Tannam, 2021). Brexit entirely upended this shared EU relationship which had militated in favour of a strong East–West (British–Irish) relationship and multiple levels and forms of cooperation.

Tannam (2021) argues that existing weaknesses in the Strand Three structure were made significantly worse by Brexit, even if Brexit was not the sole cause of the weakness of Strand Three. She argues that there were four pre-existing problems within Strand Three: (a) different perceptions of the 1998 Agreement; (b) less attention paid to Northern Ireland and the B-IIGC by both Dublin and London; (c) less attention paid to Northern Ireland, despite the regular suspension of the Agreement’s institutions and, (d) after 2016, the vulnerability of successive UK governments to pressure from Brexiteers. Tannam (2021) argues that a sense of complacency crept into both the British and Irish approach to Northern Ireland after the peace agreement was achieved. That complacency combined with the instability within Strand One of the Agreement combined to mean that Dublin and London placed less priority on Northern Ireland and ‘mistook an absence of core conflicts of interest for cooperation.’ In addition, a reduction in staff numbers working in the B-IIGC, combined with the loss of institutional memory, didn’t help (Tannam, 2021).

The B-IIGC was ‘the obvious forum for both governments to discuss Brexit’s challenges for Northern Ireland and their relationships.’ Yet, after 2007 the B-IIGC did not meet until 2018 and only met three times from 2018 to 2021 (and not at all in 2016, the year of the referendum). In contrast, the British–Irish Council (BIC) met at regular intervals of six months – as stipulated by the Agreement. But this could well be because it is perceived as more of a talking shop than the other institutions (Tannam & Kelly, 2022). It didn’t help that under Boris Johnson (2017–2022) and Liz Truss (2022) there was great volatility in the UK position on Brexit and this inevitably bled into British–Irish relations. The UK government’s announcement that it would alter an important legacy agreement which was intended to underpin the investigation of Troubles-era British army violence in Northern Ireland also caused serious problems with Dublin. In short, Brexit saw the British–Irish relationship hurtle from one crisis to another amid an extraordinary and unprecedented depletion of trust between Dublin and London (Tannam, 2021).

The repeated threats by the UK government to resile from the Northern Ireland Protocol caused great friction between Dublin and London. An early indication of the direction of travel of the Johnson administration came in the autumn of 2020 when it published the Internal Market Bill which contained controversial measures relating to the way trade would function between Northern Ireland and Great Britain. These measures would have overridden the Withdrawal Agreement which the UK signed with the EU in 2019. Article 4 of the Agreement made it clear that the provisions of the treaty took precedence over UK domestic legislation. Commenting on the move, Northern Ireland Secretary of State Brandon Lewis asserted that the Internal Market Bill would break international law in a ‘very specific and limited way’. Boris Johnson didn’t help matters when he went to Northern Ireland and assured unionists there would be no border in the Irish Sea (when it was perfectly clear that this was exactly the outcome which the Agreement necessitated).

Senior Conservative Jacob Rees-Mogg, leader of the so-called ‘European Research Group’ (ERG), the hard-line Brexit grouping within the Conservative party, conspiratorially declared the Irish border problem did not really exist – it was an ‘imaginary problem’ created by Dublin and Brussels to prevent Brexit happening (The Irish Times 26 August 2018; O’Brennan, 2019). This led Simon Coveney, Ireland’s foreign minister at the time, to warn that ‘reckless shouting and sloganeering has consequences for the people of Northern Ireland.’ He went on to say, ‘psychologically, it (peace) has transformed the landscape and allowed identity to breathe more freely. Protecting this precious achievement, a backbone to our hard-won peace, is the only motivation in prioritizing Northern Ireland in the Brexit negotiations’ (O’Brennan, 2019; O’Carroll 2018). Around the same time, Conservative minister Michael Gove pronounced that the Agreement had outlived its usefulness. This ambivalence ‘further weakened British Irish relations’ (Tannam, 2021). For Ireland, every new iteration of such gesture politics meant that trust in the UK government was further reduced. As ever, Northern Ireland was caught in between the two sovereigns. Tannam (2021) argues that ‘successful resolution and confidence building necessitates firm guardianship by both governments under Strand Three’. Brexit made that dual guardianship almost impossible at times.

All of these fundamental territorial relationships were underpinned by shared UK and Irish membership of the European Union (Murphy, 2018, 2021; O’Brennan, 2019; Tannam, 2021). Shared Irish-UK membership of the EU accommodated and reinforced the Belfast/Good Friday Agreement’s conception of the border as fluid. Of crucial importance here is the constructive ambiguity which characterizes both the EU as a post-sovereign, cross-national peace project, and the GFA as a framework for accommodation of otherwise opposed communal identities and traditions. The Agreement privileged a move to decouple national identities or loyalties from claims to national sovereignty (Walsh, 2018; Murphy & O’Brennan, 2023). The EU architecture, similarly, is built on a combination of inter-governmental and supranational logics and decision-making mechanisms. It is a hybrid model of political community and inherently ambiguous as to its endpoint. That ambiguity is crucial in that it facilitates the coexistence of, and cooperation between, often opposed national positions in the Council. Its macro purpose is to preserve the peace between national components with a bloody history of conflict. For Fintan O’Toole (2018), the link between the EU model and the GFA is obvious: ‘the Good Friday gamble was that people could live with complexity, contingency, ambiguity.’ Essential to the Agreement is that:

people born in Northern Ireland have an absolute right to be Irish or British or both as they may so choose. We have here, written down in an international treaty, a recognition that national identity is not a territorial or genetic imperative, and is not necessarily a single thing. It is chosen and, therefore, open to a change of mind. And it can be multiple: those six letters – ‘or both’ – are the glory of the agreement, its promise and its challenge.

Cathal McCall (2018) points out that the Irish border has been the subject of ‘Europeanization’ via the North–South provisions of the GFA and a wealth of cross-border initiatives for two decades, resulting in a border that has been re-configured ‘from control barrier to connecting bridge,’ from a line of sometimes violent contestation to one of increasing cooperation. The crucial link here between European integration and the peace process in Northern Ireland is relational – individuals and states can be ‘both/and’ rather than ‘either/or.’ It is in this sense that European integration enabled national sovereignties to be viewed as complementary rather than threatening, Kantian rather than Hobbesian (Coakley, 2002; Hayward, 2009, 2018; Murphy & O’Brennan, 2023).

The prospect of ‘re-bordering’ what had been ‘de-bordered’ (McCall, 2018) loomed into view as a logical consequence of the desire of Brexiteers to ‘take back control’ over the UK’s borders and this impulse did not go away after the Brexit vote. This kind of sentiment has not been confined to the UK in recent years. McCall demonstrates that the European Union, although championing a long-term process of ‘de-bordering’ since the Single European Act, has recently engaged in repeated patterns of ‘re-bordering.' The EU’s migration ‘crisis’ of 2015 was understood as a crisis, precisely because some member states (Hungary, for example) put up actual physical borders to prevent migrants reaching their territory and some countries, such as Hungary and Poland, stated they would never accept Muslim migrants. The COVID crisis saw a similar process of EU member state (temporary) ‘re-bordering’ on the grounds of the public health emergency. Similarly, the hardening of the southern border of the United States did not begin or end with the presidency of Donald J. Trump. In these developments within the EU, we see a simultaneous unfolding of de-bordering and re-bordering with multiple contradictions and policy externalities in evidence.

The border was propelled to the sharp edge of the negotiations because of both its symbolic and real-world significance (Ward, 2018). As Anderson (2018) puts it, ‘GFA cross-border arrangements would be disrupted or severed, and the successful de-politicization of the border would be thrown into reverse with unpredictable consequences. Border infrastructures and customs personnel could provide choice targets for dissident republican paramilitaries,’ which, in turn, could trigger a response from Loyalist paramilitary groups. The 1998 Agreement acknowledged that, as well as being contentious within Northern Ireland, these boundaries are ‘fuzzy’ and interrelated across the context of the island of Ireland (Hayward, 2018). Brexit ‘has undermined the extent to which critically important characteristics like fluidity and fuzziness’ can be maintained in relation to the border (Murphy & Evershed, 2021; Murphy & O’Brennan, 2023).

The territorial border on the island of Ireland has also given rise to another kind of border – what Gormley-Heenan and Aughey (2017) term the ‘border of the mind.’ Here the emphasis is on identity and cognition. Citing J.C. Beckett’s History of Modern Ireland (2008/1966), they argue – as Beckett did – that the real border in Ireland is not on the map ‘but in the minds of men.’ In this framing, the Good Friday Agreement constitutes a way forward towards the ‘agreed Ireland’ that John Hume championed, a way of ‘de-commissioning’ the mindset of conflict and division. Crucially, in such a reading, this did not signal a path to unification but ‘a progressive lowering of the political temperature and the fashioning of a new modus vivendi on the island’ (Gormley-Heenan & Aughey, 2017).

There is a broad spectrum of social science scholarship which engages with conflict and post-conflict societies and how identities are impacted by both endogenous and exogenous events and developments. In a disordered world, people can often seek out a return to the familiar and comforting. Group identities frequently offer a safe harbour in the midst of chaos and division. Moshin Hamid’s (2007) epochal novel, The Reluctant Fundamentalist captures how identities can be powerfully impacted by seismic events. A cosmopolitan identity that is taken as given one minute can be subject to significant – even extreme – revision in the aftermath of specific developments (in this case, the 9/11 attacks).

Liminal identities follow as a logical consequence of liminal territorial political contestation. The unmooring of identities is greatly aided by macropolitical polarization. Brexit undoubtedly fits the bill: it constitutes just the kind of ‘critical juncture’ described by Ben Rosamond’s (2016) notion of an ‘intensive period of fluidity and crisis that brings forth revised institutional equilibria.’ We might go further and suggest the ‘critical juncture’ here has produced a return to a familiar form of identity disequilibrium, leading to institutional drift and a fracturing of established norms of ‘doing politics’ across all three territorial strands associated with the Good Friday Agreement.

In a post-conflict context, where identities are contested and binaries rife, an exogenous ‘event’ like Brexit can prove extremely challenging. Identity then becomes amplified, intensified and magnified by renewed political contestation. In some ways, one might argue that, even without the Brexit vote, Northern Ireland experienced significant ontological insecurity after the signing of the Good Friday Agreement. The repeated suspension of the institutions provides enough evidence of the lack of trust between the political parties. Thus, for Cash (2017), the ‘shift away from the violence of the past has increased rather than reduced, the ontological insecurity of its citizens.’ These insecurities arise from an array of different things, including: the ghosts of a divisive and violent past, that continue to haunt political life, from battles for predominance within the new political institutions, from the ambiguities intrinsic to a cultural transformation that is neither universal nor thoroughgoing, or from ‘externalities,’ such as Brexit, that further disturb and complicate an already wickedly complicated situation (Cash, 2017). Given that post-conflict societies often exhibit a limited ‘state of becoming’ they are especially vulnerable to exogenous events that ‘disturb the peace.’ As Zygmunt Bauman (1995) argues, these are societies that combine contested meanings of the past with a nebulous set of promises about the future, often in a context of tumultuous politics in the present. Unlike the early anthropologists who posited liminality as a condition of transitioning to a definite endpoint, Bauman, Roger MacGinty and others point to a much more fluid and indeterminate model of liminality, one without obvious exit points and agreed goals. It is precisely in this kind of context that the relative ambiguity of identity landscapes can quickly change and revert to previous patterns of conflict and contestation. Here the liminal society is fractured and marked by lingering insecurity. This furthers the sense of fragility of the ‘peace’ that has been ‘accomplished’.

Conclusion

The Good Friday Agreement constituted a true watershed in the life of Northern Ireland. It brought an end to a 30 year cycle of violence that resulted in 3700 dead and many tens of thousands of people traumatized. The Agreement, however, has proved difficult to bed down. As Murphy and Evershed (2021) remind us, ‘since 1998, the Assembly and Executive have been suspended or gone into abeyance for a cumulative total of some eight years: around one third of the 25 years in which the Agreement has been in place.’ Disagreements between the parties have resulted in a further five subsequent refinements of the Agreement: St. Andrews (2006), Hillsborough (2010), Stormont House (2014), A Fresh Start (2015) and New Decade, New Approach (2020), each of which has revised to some degree the original settlement ‘in an attempt to put it on a more sustainable footing’ (Murphy & Evershed, 2021).

The Good Friday Agreement ‘sought to create the basis for peaceful democratic politics by removing the ‘existential’ things which would have been seen as turning unease into dangerous fear or a sense of being under threat’ (Coulter et al., 2021, p. 164). Brexit applied steroids to these feelings of uncertainty and fear and undermined all three territorial strands of the Agreement. In addition, UK government ambiguity, along with in-fighting within the Conservative party ‘created a climate of grave uncertainty for government, business and civil society actors in Northern Ireland’ (Murphy & Evershed, 2021). An added complication was that a clear majority of people in Northern Ireland (56%) voted to remain in the EU. In a region where the principle of consent is meant to underpin both the present and any future constitutional settlements, this provided ample grounds for those who sought to contest the result and upend the constitutional order put in place through the GFA (O’Brennan, 2019).

Brexit catalyzed a very serious deterioration in all three strands of territorial relationships and associated identities. At one and the same time, Northern Ireland was positioned at the centre of the recurring Brexit crises after 2016, and yet remained aloof from and peripheral to it. The liminal political condition of Northern Ireland was underlined again in 2022 by elections to the Northern Ireland Assembly which failed to lead to the forming of an Executive. The absence of trust between representatives of both the principal ethnonational communities was palpable. The Brexit vote, and the years of contentious bickering it set in train, sparked the re-politicization of cross border relations in Ireland and with it a concomitant increase in political sensibilities about both North–South relations and the British–Irish relationship (Hayward, 2018). As Tannam (2021) puts it: Brexit turned the common interests of Ireland and the UK into (potentially) competing interests, to the point where, after 2016, the relationship frequently looked like a zero-sum game.

Employing the concept of liminality and liminal politics, this piece argues that Northern Ireland’s experience since Brexit emphasizes the importance of thinking about what it is to be ‘in-between’ – in between territories and spatial landscapes, constitutional and institutional orders and identities. In this sense, thinking about liminality has moved far away from its anthropological roots. Rather, contemporary social scientific re-imaginings of liminality have allowed us zone in on relationships between institutions, identity and power in contested societies. Being ‘in between’ no longer means transitioning from one condition to another. It can mean being permanently stuck between different constitutional landscapes with a knock-on impact on identities and territorial relationships. Brexit, arguably, challenged the notion that liminal spaces have no ‘endpoint.’ For some nationalists, in particular, the Brexit vote brought the prospect of their desired endpoint, i.e. a united Ireland, much closer into view. In turn this perspective served to revive apprehension and fear in unionist communities. The collision of the UK’s ambiguous and tenuous constitutional order with Northern Ireland’s idiosyncratic and particularistic post-1998 governance arrangements ‘produced conditions capable of fatally challenging the resilience and durability’ of the original peace deal (Murphy & Evershed, 2021). For Northern Ireland, then, the post-Brexit landscape has been one of angst, dislocation, deep uncertainty and a return to some of the most vexed questions of the past.

Disclosure statement

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

Additional information

Notes on contributors

John O'Brennan

John O' Brennan is a professor of European politics in the Department of Sociology at Maynooth University. He also holds the Jean Monnet Chair in European integration and is Director of the Maynooth Centre for European and Eurasian Studies. His most recent works include ‘Ireland and the European Union' (with Mary C. Murphy) in Politics in the Republic of Ireland, seventh edition, Routledge, 2023.