Shared spaces, separate places: desegregation and boundary change in Northern Ireland’s schools

ABSTRACT

Education in Northern Ireland continues to be organised along denominational lines, with more than 90% of pupils attending separate Catholic or de facto Protestant schools. Since 2007, an initiative known as ‘shared education’ has operated in the region to provide opportunities for pupils from separate schools to meet and learn together on a regular basis. This involves the formation of collaborative partnerships between Catholic and Protestant schools to deliver joint classes and activities for mixed groups of pupils. One of shared education’s objectives is to create more porous boundaries between schools and thereby provide the conditions for relationship-building between pupils. Mindful of this aim, the current study explores to what extent, and how, shared education alters social and spatial practices that sustain division in educational settings. To do so, it adopts Tilly’s (2004) typology of social boundary mechanisms as a framework for analysing qualitative data collected with 60 pupils in two shared education partnerships. The research identifies instances where boundaries are formed or intensified through shared education, as well as where they are relaxed and reduced, and examines in particular how the emplacement of encounter contributes to this variation in social boundary change.

KEYWORDS:

• Shared education
• Social cohesion
• Social boundaries
• Geographies of encounter

Introduction

While it is over two decades since the Belfast/Good Friday Agreement brought an end to sectarian conflict in Northern Ireland, the division between Catholics/nationalists and Protestants/unionists continues to dominate social life in the region. One area in which these boundaries are particularly pronounced is education: more than 90% of pupils attend schools that are predominantly Catholic or predominantly Protestant (DENI 2019). Amid concerns about the contribution of this parallel system to social relations, the most recent and arguably most ambitious of a series of initiatives to mitigate divisions between pupils attending separate schools is ‘shared education’, introduced in 2007. Through shared education, schools collaborate across denominational lines on a range of activities, most notably the joint provision of classes for mixed groups of pupils. These occur on a regular basis over a year or more, with pupils travelling between schools to learn together. Explicit in the rationale for shared education is the intention to create ‘porous boundaries’ between separate institutions and build new connections among pupils, teachers and parents (Gallagher 2017, 44). Over time, it is hoped that this will ‘change the nature of relationships between communities … and so help to secure the peace in Northern Ireland’ (Gallagher 2017, 44).

This paper considers shared education’s progress towards this aim of altering institutional and, consequently, social boundaries in Northern Ireland. Previous studies suggest that it has had some success, with participating pupils and teachers reporting a higher number of cross-group friendships and more positive attitudes to the other group than their non-participating peers and colleagues (Hughes et al. 2012, 2020; Reimer et al. 2020). However, while this research has focused on shared education’s outcomes in terms of individual psychological change, it has been less explicit on other change brought about by shared education, particularly the spatial transformation associated with newly-mixed environments. Given the novelty of bringing pupils together in schools that have for so long been associated with one community or the other, this spatial change and its implications for social boundaries merit more focused attention. This study aims to address this by centring its lens on boundary processes – that is, the social and spatial practices that sustain the separation of groups – and the impact of shared education on these. To do so, it applies Tilly’s (2004) typology of social boundary mechanisms to the analysis of qualitative data collected with 60 Catholic and Protestant pupils in two shared education partnerships.

The development of shared education in Northern Ireland

While denominationalism had been a feature of Irish education from the mid-19^th century, the first government of Northern Ireland aimed to create a single, unified system without direct religious influence (Gallagher 2005). Opposition from Catholic and Protestant church authorities to plans to transfer to schools to state control meant this aim was unrealised (though the Protestant churches subsequently transferred their schools, with conditions, in the 1930s) and there remains a largely parallel system of education, albeit now with full state funding (Byrne and Donnelly 2006). From the early 1970s, as hostilities in Northern Ireland increased, concerns grew among educationalists and observers that denominational schooling might foment social divisions (Gallagher 2004). Though this was contested by church authorities (see, for example, Catholic Bishops of Northern Ireland 2001), there was general support for educational initiatives aimed at attenuating division and enhancing mutual understanding. These initiatives took two forms: those that sought to challenge separation within existing educational structures; and those that sought to disrupt the very structures themselves.

Initiatives in the first group included 1) curricular interventions to increase pupils’ understanding of the histories, traditions and beliefs of the two main groups in Northern Ireland; and 2) school-based contact programmes to bring Catholic and Protestant pupils together for short-term projects or visits (Gallagher 2004; Richardson 2011). In the second group, offering an alternative to separate education, were integrated schools. These were established from the early 1980s, largely through parental efforts and with philanthropic support, to educate pupils from Catholic, Protestant and other backgrounds in the same institutions (Fraser and Morgan 1999). State-funding was extended to the integrated sector in 1989 and there are currently 65 integrated schools in Northern Ireland. Collectively, these initiatives have had variable success. Curricular and contact schemes generated some good practice, but were often of short duration, poorly managed and a low priority in schools (ETI 2000; Richardson 2011; Smith and Robinson 1996). The integrated sector, by comparison, is well established and notionally popular (Hansson, O’Connor-Bones, and McCord 2013), but its growth has slowed in recent years and it remains small.

Given the limitations of contact schemes and the slow growth of integrated education, shared education was founded in 2007 to provide pupils across all schools with opportunities for high-quality, sustained cross-group interaction. Initially implemented through three pilot projects with philanthropic funding, it currently involves approximately two-thirds of primary and post-primary schools¹ in Northern Ireland and is managed by the Department of Education (DENI 2015, 2018). As outlined above, shared education supports the formation of inter-denominational school partnerships to collaborate on all aspects of educational delivery – school improvement, teacher professional development, the creation and sharing of resources, etc – but most notably the provision of curriculum-based classes and activities for mixed groups of pupils. These occur regularly (e.g. weekly or monthly) over the course of a year or more and are typically held at each school alternately. While the focus and structure of shared activities is left to schools to determine according to their local circumstances (Gallagher 2016), shared education policy specifies that each partnership should feature ‘significant, purposeful and regular engagement and interaction in learning between children and young people from different community backgrounds’ (DENI 2015, 5).

One of the hopes of shared education’s architects was that, by removing ‘the institutional barriers which had been erected between young people, largely for reasons of history, it might be possible to change the nature of relationships’ (Gallagher 2017, 44). Two bodies of research were influential in this aspiration: research on social networks, which emphasises the value of positive interdependencies to improve organisational working and relationship-building; and research on intergroup contact, which promotes sustained and intimate cross-group interaction, particularly via friendship, to change intergroup attitudes (Duffy and Gallagher 2017). Much research into shared education’s contribution to peacebuilding has taken a contact-based approach, examining individuals’ levels of ‘prejudice’ vis-à-vis the other group (the ‘outgroup’) as a measure of its impact. These studies have found that participants of shared education report higher levels of friendship, increased intergroup trust, decreased anxiety, reduced stereotyping, and more positive intergroup attitudes as a result of their involvement (Hughes 2014; Hughes et al. 2012, 2020; Reimer et al. 2020). However, this research has also highlighted some less favourable responses, including reluctant participation from a substantial minority of pupils, a tendency towards non-contact in some shared activities and the comparative rarity of cross-group friendships that extend beyond the school (Hughes 2014; Loader 2015).

While research of this type is valuable, the emphasis within intergroup contact theory on individual psychologies, and particularly indicators of prejudice, means that this research can offer only a partial perspective on shared education’s contribution to peacebuilding. For one thing, it can neglect the wider political and historical context in which contact occurs (Dixon, Durrheim, and Tredoux 2005), though some qualitative research in Northern Ireland – including research on shared education – has addressed this, noting how participants’ experiences of contact are shaped by their experiences of societal conflict and interpreted through that lens (Donnelly 2008; Hughes et al. 2010). Moreover, rooted in the quantitative tradition and with a focus on attitudinal outcomes, most contact research offers only partial insight into the processes by which contact produces changes in attitude. Mediating processes are typically described in terms of psychological change, such as increased or decreased anxiety or empathy, without necessarily elucidating the social or environmental features that might contribute to this.

In particular, and of most relevance to this paper, research on intergroup contact has often ‘either ignored the spatial dimension of intergroup relations or … espoused limiting conceptions of social space’ (Dixon 2001, 590), neglecting how physical settings may sustain separation, practically and ideologically. Outside social psychology, this inattention to spatial phenomena has also been acknowledged in a small body of geographical work on social encounters (Mayblin et al. 2015; Omer et al. 2018). In response, these authors argue for greater consideration of ‘the spatial organisation or geographic unit of analysis under study [and] the types of interactional space available’ (Omer et al. 2018, 59) in the study of intergroup contact. This has resonance with research into shared education: while some studies have considered the effect of the local neighbourhood or town on intergroup dynamics among students (Hughes 2014; Loader 2015), there has been limited regard for the influence of place at the micro level of the school. This is an important omission given that shared education is rare in bringing pupils together in denominational school spaces characterised by ‘internal homogeneity and clear, strong boundaries’ (Sibley 1995, 80), rather than the perceived ‘neutral’ venues more typical in contact programmes (see Drury, Abrams, and Swift 2017; Klein et al. 2018).

To begin to develop a fuller understanding of shared education’s influence on group dynamics, and mindful of its aim to foster ‘more porous’ institutional and intergroup boundaries, the current study considers the programme’s role in effecting boundary change. This shifts the focus from individual attitudes and dispositions to the social and spatial practices that regulate relations between groups. Importantly, the study considers shared education as an emplaced phenomenon, the physical settings and its meanings shaping participants’ experiences and ultimately the initiative’s progress against its aims. The article adopts Tilly’s (2004) social boundary mechanisms as a framework through which to examine shared education’s influence on boundary processes, and these are detailed in the following section. Prior to this, the article considers further the concept of emplacement in relation to intergroup encounter.

The emplacement of intergroup encounters

This paper examines encounters through shared education as ‘emplaced’ experiences – that is, those that take place within distinctive geographical, material and social spaces (Svašek and Komarova 2018; Gieryn 2000). It acknowledges that the settings of encounter are not mere containers of interaction, but ‘significant constituent[s] of social processes and bearer[s] of meaning’ (Gunn 2001, 3). Indeed, physical environments are inseparable from the social life that occurs within them, each informing the way the other is experienced (Gardner and Mand 2012). Places contribute to the structuring of society and social relations – sustaining boundaries, differences and hierarchies that include some and exclude others; regulating interaction and behaviour; separating the acceptable from the transgressive; and reproducing norms and identities (Gieryn 2000; Benwell and Stokoe 2006). Defining (however implicitly) who belongs and who does not is important in the shaping of social space (Sibley 1995) – and, in turn, ordering the world into ‘us’ and ‘them’.

With respect to intergroup relations, a body of research from social geography has considered the role of place and space in shaping the potential for ‘meaningful encounters’ – those which alter negative perceptions of ‘the other’, reduce social boundaries, and improve intergroup relations (Valentine 2008; Askins 2016). This work is critical of a literature within urban studies that equates proximity with contact and thus with increased understanding and mutual respect (Valentine 2008; Askins and Pain 2011). Valentine (2008) cites examples of diverse environments that have engendered little social mixing or, worse, instances of incivility emerging from the reproduction of wider social inequalities within interaction. Subsequent studies of encounters as emplaced experiences reveal the importance of space and materiality in shaping interaction, both ‘opening up ways of belonging, exchange and new social relations and/or conversely closing them down’ (Neal and Vincent 2013, 918; see also Mayblin et al. 2015; Mayblin, Valentine, and Andersson 2016).

With regard to the influences on the nature of encounters, Mayblin et al. (2015) found that the prevalence of one-to-one or group-based interaction varied according to changes in the size and shape of a space, the seating arrangements, the level of background noise and the number of people present. Studies of place have also highlighted the spatially contingent nature of identity and emotion, and the implications of this for encounters. Changes in a setting, or movement between locations, may disrupt individuals’ ‘emotional and embodied sense of being “in place”’ (Askins 2016, 517), producing feelings of belonging or exclusion, comfort or discomfort that affect their disposition towards interaction (Omer et al. 2018). In addition, research has revealed the influence of placed-based norms on people’s behaviours and responses to others (Valentine and Harris 2016). This raises questions about the potential of encounters in ‘the specific time-space’ of one setting (Mayblin, Valentine, and Andersson 2016, 221) to effect social boundary change in places governed by different norms.

Boundaries and boundary mechanisms

Although the notion of the ‘boundary’ is central to much research in the social sciences (for example, defining the social and cultural groups, institutions or geographical sites that are researchers’ focus), comparatively few scholars have made boundaries and boundary processes the explicit subject of their scholarship (for example, Wimmer 2008; Lamont and Molnár 2002; Tilly 2004, in sociology; Newman 2003, 2006 in political geography). Among this group, boundaries are defined conceptually in terms of their function to separate one entity or category from another – as Tilly describes, to ‘interrupt, divide, circumscribe, or segregate distributions of population or activity within social fields’ (214) – and to prevent the entry of that which is considered undesirable. This function is similar whether the boundaries contain social groups, cultural practices, places or territories (Newman 2003), though the salience and stability of the boundary may vary (Wimmer 2008). In performing this function of categorising and ordering social life, boundaries operate as a series of social processes, though they may also have a material form. Newman (2003), for example, describes boundaries as institutions with internal rules and behaviours that govern inclusion and exclusion, the permeability of the boundary, and the potential for movement across it. Similarly, for Wimmer (2008), boundaries comprise both categorical and social or behavioural dimensions. The former involve the classification of individuals into groups while the latter provide ‘scripts of action’ that regulate interactions across the boundary (975).

Boundaries have been described as one of the ‘most fertile thinking tools’ in the study of social and relational practices (Lamont and Molnár 2002, 169). However, with much scholarship occurring in disciplinary silos, there have been calls for more unified theoretical perspectives on the properties and functioning of boundaries that can be applied across different settings, circumstances and scholarly fields. Lamont and Molnár, for example, have emphasised the importance of understanding the characteristics of boundaries, such as their ‘permeability, salience, durability, and visibility’ (186), and the conditions that shape these. Newman (2003), moreover, has advocated research into the processes by which boundaries are demarcated and managed, on one hand, and how they are opened and removed, on the other. Understanding how boundaries are made and unmade may involve, for example, consideration of the cognitive and discursive practices that sustain the boundary, including categorisation, comparison and stereotyping; narratives of group history, identity and territory; and representations of the boundary in the media, art and literature (Leonard 2008; Lamont and Molnár 2002; Dixon 2001). It may also require attention to the power relations associated with boundaries, specifically who benefits from the boundary and who determines to what extent it is open or closed (Newman 2003).

An important contribution to theorising boundaries is the work of Tilly (2004) on social boundary change. This outlines a series of cognitive and relational mechanisms that, Tilly argues, ‘help explain a wide variety of boundary changes and their consequences’ (214). Tilly divides these into 1) mechanisms causing boundary change, and 2) mechanisms constituting boundary change, where the ‘change’ may be the creation, intensification or removal of a boundary. Mechanisms causing boundary change in Tilly’s model include:

1. Encounter, where two networks come together and create a new boundary;

2. Imposition, where authorities create boundaries that previously did not exist;

3. Borrowing, which introduces boundaries that exist elsewhere – for example, in other organisations;

4. Conversation, whereby frequent, sustained social interaction leads to change at the boundary; and

5. Incentive shift, which creates new rewards or penalties for pursuing particular boundary strategies.

While these mechanisms precipitate boundary change, Tilly proposes four further mechanisms that he describes as constitutive of boundary change. These are:

1. Inscription, which occurs as relationships on either side of and across the boundary intensify and become more distinct, and representations of the boundary become more widespread. This may be reversed by a countervailing process of ‘erasure’, where relations become less intense and distinct.

2. Activation, whereby a boundary assumes increased salience as an organiser of social relations across the boundary or on either side of it. The converse of this is ‘deactivation’, which reduces salience.

3. Site transfer, where boundaries are maintained but the position of individuals, groups or social sites changes in relation to the boundary – for example, in religious conversion.

4. Relocation, where the boundaries that organise action and interaction change, typically though the deactivation of one boundary and the activation of another.

These four mechanisms may occur singly, in combination or in sequence. For example, Tilly notes that inscription and activation often occur together, simultaneously increasing the salience of a boundary and intensifying social relations on either side or across it.

Though described as ‘preliminary’ by Tilly, these mechanisms serve as valuable concepts for exploring boundaries and boundary change through shared education. In the remainder of this article, I apply them in analysing data collected from 60 pupils who participated in shared education in two school partnerships in Northern Ireland. This analysis seeks to identify and understand how boundary change occurs through shared education, recognising that this ‘change’ may take diverse forms – creating or strengthening as well as relaxing boundaries – and that even ostensibly minor modifications may hold significance where boundaries are salient and groups are highly differentiated (Wimmer 2008). The paper’s focus on the experiences of young people recognises that they are the intended beneficiaries of shared education and those for whom, and among whom, boundary change is most significant.

Methodology

The two shared partnerships in which fieldwork took place were located in small rural towns hereafter (pseudonymously) named ‘Whitecliff’ and ‘Bellevue’. Each partnership comprised two post-primary schools, one predominantly Catholic and the other predominantly Protestant. In Whitecliff, both schools were non-selective and were located within short walking distance of one another; in Bellevue, the partnership included a selective and a non-selective school located approximately 1.5 miles apart. Pupils travelled by bus or taxi between the schools. The Whitecliff schools had collaborated in various forms over a long period and their shared provision was extensive, comprising the majority of GCSE and A-level examination subjects. Sharing was a relatively new experience in Bellevue and their provision was more limited, encompassing two shared subjects post-16 – an accredited Personal Effectiveness course and a vocational (BTEC) engineering course. Beyond the school, the two areas also differed in demography and the nature of local relations. In Whitecliff, the population (approximately two-thirds Catholic, one-third Protestant) was comparatively well integrated and relations largely harmonious. In Bellevue, residential segregation was more marked between the town centre (predominantly Protestant) and the hinterlands (predominantly Catholic) and intergroup hostilities persisted.

Data were collected through semi-structured group interviews (2–4 per group) with pupils who had taken part in shared education over a period ranging from six months to two-and-a-half years. The group format was chosen to reduce power differentials between the adult researcher and the participants and encourage more relaxed discussion (Eder and Fingerson 2001). Groups also comprised only pupils from the same school to allow them to speak as freely as possible about sensitive issues of difference and identity. Written consent for involvement in the research was obtained from all participants, as well as from the principal of each school and from parents where pupils were aged under 16. In Whitecliff, potential participants were identified by teachers with the criterion that the group be mixed in age, gender and academic ability. In Bellevue, where the numbers involved in the programme were smaller, all participating pupils in year 13 (‘lower sixth’) were invited to take part. An information leaflet accompanying the consent form advised participants of the nature and purpose of the research, the protocols for ensuring confidentiality and anonymity of data, and arrangements for data storage. The project and all associated materials received prior approval from the ethics committee at the author’s institution.

In total, 60 pupils were interviewed for this study, 32 in Whitecliff and 28 in Bellevue. Twenty-seven pupils in the sample attended controlled/Protestant schools and 33 attended maintained/Catholic schools. All were aged between 14 and 18, and the numbers of female and male students were 32 and 28 respectively. Interviews took place in an unused classroom during the school day and, with participants’ permission, were audio-recorded and transcribed for analysis. Data were analysed using a thematic approach that involved close reading and coding of data and the subsequent combination of codes into themes that reflected patterns within the dataset. Following Fereday and Muir-Cochrane (2006), both inductive and deductive codes were used, the latter informed by Tilly’s social boundary mechanisms. All coding and analysis was undertaken by the author, who is resident in Northern Ireland but hails from elsewhere in the UK and tends to be positioned outside the Catholic/Protestant binary. In this regard, the research process and analysis were subject to the benefits and disadvantages of ‘outsider’ status (Smyth 2005; Hermann 2001). While the benefits of this can include a perception among participants of researcher ‘neutrality’ that encourages confidence and frankness in interviews, the limitations include the risk that the researcher fails to grasp geographical and cultural references or nuances that would be evident to one with ‘insider’ status. To mitigate this, I capitalised on another potential advantage of the ‘outsider’ – the tendency for participants to assume limited knowledge on the researcher’s part – and asked questions to clarify my understanding that might have sounded strange coming from an insider. I also discussed findings of the research with close colleagues who could confirm or correct my understandings of particular symbols, traditions and place-based associations.

Findings

‘Incentive shift’ and participation in shared education

To encourage schools’ involvement in shared education, the programme’s originators have explicitly promoted its educational benefits (sharing resources and expertise, enhancing teachers’ skills, and delivering a broader curriculum through collaboration) alongside its reconciliation objectives (DENI 2015). This recognises that, in a system where performance is measured by academic outcomes, reconciliation activity is typically a lesser priority. Among pupils, this study found that, where they had a choice over involvement in shared education,² the programme’s perceived educational benefits similarly provided the main inducement to participate – what Tilly terms the ‘incentive shift’. The majority of interviewees described their wish to pursue a particular exam course, experience a new subject, or gain points for university entry as their primary motive for involvement.

Interviewer

So what made you choose these classes?

Female student

Well, I want to be a social worker, so that’s why I chose Health and Social Care. And Personal Effectiveness because it’s worth 70 UCAS points, and I wanted to do religion, but it was in the same bloc as a different subject.

Interviewer

Right, OK. Yeah. What about you?

Female student

I was just interested in Health and Social Care. And I just thought the 70 UCAS points were useful and I wasn’t doing a fourth subject. (Catholic school, Bellevue)

Female student

Yeah. I was, like, first I was going to pick BTEC sport, cos it was taught here, and I knew the teacher here, but then it’s more like theory-based, so you don’t get much practical … And our PE teacher said that we’d be better doing the GCSE PE over there, cos it’d be more suited to us and we’d be able to do it, no bother.

Interviewer

Mmm. OK, so initially, did you think you’d prefer to stay at this school?

Female student

Yep. (Protestant school, Whitecliff)

As the second quote demonstrates, the opportunity to study a preferred subject could provide a powerful incentive to traverse the physical boundaries of another school, even where students were apprehensive about doing so. In comparison, few spoke of the social benefits of shared education – for example, meeting a new group of peers, making friends, or learning about another religious or cultural tradition – as a motive for participation. Where potential social benefits were discussed, students focused on the instrumental rather than intrinsic value of these advantages. They emphasised the opportunity to develop skills, such as the ability to interact in a culturally mixed setting, that they could convert to capital in the higher education or employment markets.

I think mixing with different schools gives more of an experience and you can do it for work later on in life. (Female student, Catholic school, Whitecliff)

For a minority of students, however, the opportunity to gain these ‘soft’ skills or study a new subject proved an insufficient incentive when weighed against the requirement to cross the school boundary. This group had strategically selected subjects that were delivered through shared classes but only at their own school. Anxiety rather than hostility appeared the main influence on these pupils’ decisions, particularly apprehension at the prospect of attending an unknown school where they would be surrounded by unfamiliar people.

Male student

I picked most of the subjects here, cos if I went over to [Catholic school], I’d get lost or something.

Interviewer

Right, okay. Yeah.

Female student

Aye, I picked them most here, because if you go over there, you don’t really know anybody or know where to go or anything like that.

Interviewer

Okay. So did you choose the subjects that that let you stay in this school?

Female student

Yeah. I did, anyway. (Protestant school, Whitecliff)

For shared classes to be feasible, the incentive for participation must be sufficient to encourage at least some students to cross the boundary from one school to another. While both partnerships were largely successful in this respect, this minority of (particularly younger) pupils resisted the incentives for involvement. This response reveals the ‘emotional geographies’ of encounter (Askins 2016) that pupils negotiated in making decisions about participation in shared education: faced with the prospect of being ‘out of place’ in the partner school, and the accompanying feelings of uncertainty and fear, these students opted to remain on familiar ground. From a reconciliation perspective this was minimally disruptive as these pupils could participate in a shared class at their own school – in itself a form of social boundary crossing given the religious homogeneity of most classes in Northern Ireland. However, this reluctance to cross school boundaries meant that the full range of subject options was unavailable to these pupils, potentially limiting their educational opportunities in comparison with their more confident peers.

‘Encounter’ – students’ experiences of the partner school

For the majority of pupils, involvement in shared education entailed attending the partner school for at least some classes and activities. Often, this was the students’ first experience of visiting an institution associated with the other religious group and was accompanied by feelings of uncertainty and trepidation. As pupils described their attendance at the partner school, a clear distinction emerged between experiences of contact in the transitional spaces of the building (entrances, corridors) and those that were possible in the classroom. Transitional spaces were less subject to surveillance by teachers and, given the transient nature of contact, behaviour that flouted norms of respect for difference had fewer social repercussions than in the classroom where encounters were more sustained. These areas also offered little opportunity for interpersonal engagement and differentiation compared with classroom-based encounters. Consequently, contact in these transitional spaces was typically brief and described in group-based terms – as ‘us’ encountering ‘them’. This was consistent with Tilly’s mechanism of ‘encounter’, whereby bringing two networks together could introduce or sharpen a boundary that was not previously present, in this case in religiously homogeneous schools.

Alluding to this boundary, a large proportion of pupils, including the majority of those in Bellevue, described feeling uneasy and ‘out of place’ as they moved around the other school. This was especially the case in the programme’s early stages, these feelings diminishing for some (though not all) pupils over time. Interviewees also made reference to the actions of outgroup pupils that, from their perspective, reinforced these feelings of difference, and which underline the emplaced and embodied nature of interaction. These ranged from subtle physical behaviours, such as the staring that pupils perceived from others as they arrived at the school, to more overtly hostile actions, including name-calling and the singing of sectarian songs. Notably, virtually all the negative incidents reported during shared education occurred during these transitory encounters.

Other different years and stuff stand outside the classrooms and you sort of have a feeling like you don’t belong there and you shouldn’t be there. (Male student, Catholic school, Whitecliff)

You feel a bit intimidated, like, getting out of the bus and everyone just kind of staring at you. It was intimidating at the start. (Male student, Protestant school, Bellevue)

Interviewer

Have you ever had any comments as you were walking through?

Female student

One or two. I have had a few comments.

Interviewer

What sort of things have they said?

Female student

Just sort of sectarian kind of things, but you know, it’s probably junior school. Immature kind of kids, you know? (Catholic school, Bellevue)

As the fourth quote indicates, pupils often attempted to negotiate and ‘soften’ the boundaries that these negative encounters had sharpened. Most commonly, they sought to minimise the significance of the experience, dismissing the culprits as ‘immature’ and attributing their behaviour to ignorance or a desire to impress peers. This appeared to naturalise these actions as the behaviours of youth, thus stripping them of malicious intent and reducing the force of their impact.

It was evident from the data that the impact of ‘encounter’ on the Catholic/Protestant boundary was intensified by two features of shared education, the first being the requirement that pupils wore their school uniform during shared activities. Informed by contact theory, this measure was intended to increase identity salience (as school uniform in Northern Ireland in most cases indicates one’s community background) and ensure that any positive effects of contact generalised beyond the encountered outgroup member to the group as a whole (see Hewstone and Brown 1986 for the rationale for this approach). For participants, however, the presence of the two uniforms emphasised the social boundary between them by privileging religious identity above other characteristics and making it visible on pupils’ bodies. The perceived incongruence between this embodied identity and that ascribed to the institutional setting, particularly in a wider context of conflict, caused visiting pupils to feel vulnerable. Again, this heightened sense of identity and associated vulnerability was particularly pronounced in the school’s transitional spaces, where other student’ responses were less predictable, highlighting the dynamic and spatially contingent nature of identity and emotion during shared education.

I don’t like the idea that people can judge us because of the uniform. The uniform is such a difference. Because it’s [colour] over there and it’s [colour] here, people can definitely see that you’re Catholic, I suppose. If you saw someone down the corridor of a different uniform, you instantly think ‘different religion’, do you know what I mean? (Female student, Catholic school, Whitecliff)

Female student

I don’t feel uncomfortable walking around, but it’s still sort of like, ‘Oh, I don’t know if I want to walk there or not’ because of the uniform …

Interviewer

Yeah, yeah. What were you worried might happen, then?

Female student

Mmm, just people might, like, say things and look at you. It gets very uncomfortable. (Protestant school, Bellevue)

Also enhancing this sense of boundary among participants, particularly in Bellevue, were the protocols for pupils’ arrival and departure from the partner school and the location of shared classes. Pupils in Bellevue typically attended registration at their own school and were then taken to the other school by bus, arriving shortly after the start of lessons. Due to differences in school finishing times, visiting pupils left before lessons ended to return to their own school. Although these arrangements appeared to be driven by administrative concerns, visiting pupils perceived their intention as keeping the two groups of students separate and ‘isolated’ from one another outside the classroom. Similarly, where teachers escorted visiting pupils around the school, interviewees perceived this as an act of protection, guarding the boundary for their own safety.

Female student 1

Once we’ve finished our class, [teacher] would walk us to the door again.

Female student 2

Back, yeah, back down.

Female student 1

She would always walk us down. Which is good in a way, too, because you can see she’s looking out for us, just in case anything was to happen. (Catholic school, Bellevue)

Practices of meeting visiting students at the rear entrance of the school (which remained locked until a teacher arrived) and locating shared classes in outbuildings had a similar effect. These served to reinforce the identity and ‘ownership’ of the school by permitting visiting pupils entry only through an act of admittance from a member of the other group (i.e. unlocking the door) or by locating them away from the main premises, which also limited contact with the rest of the student body. From interviewees’ responses, it was clear that these practices, however unintentionally, had communicated to pupils that interactions among group members – at least outside the controlled environment of the classroom – were undesirable and potentially risky. If ‘encounter’ had created social boundaries within the school, these actions served to define and entrench these boundaries, in line with Tilly’s mechanism of ‘inscription’.

‘Conversation’ in the shared class

While the transitional spaces of the school were most often sites of ‘encounter’, classroom-based activities aimed to foster what Tilly describes as ‘conversation’ – the frequent, sustained interaction with the other group that can lead to boundary change. The nature and extent of conversation in the partnerships under study was, however, mixed: while pupils reported that interaction (where it occurred) was largely positive, it was not widespread. The analysis found that descriptions of classes characterised by infrequent interaction were almost twice as common as those where interaction occurred regularly. A key influence on this was the provision of in-class opportunities for pupils to interact and consolidate burgeoning relationships. While shared education programmes had typically been preceded by some form of ice-breaking activity, pupils reported difficulty continuing this interaction once they moved into the classroom, with some classes including little time for either topic-related or social discussion. This could result in the reversal of gains made at the introductory stages of the programme; conversely, where shared activities required and encouraged interaction, pupils typically reported more frequent engagement and closer relationships.

Female student 1

The five boys always come down to us from [Catholic School] and we just, you know, work together. We all get on really well. We just work at Personal Effectiveness and-

Female student 2

Just sort of chit-chat.

Female student 1

Just sort of chit-chat and steady work. (Protestant school, Bellevue)

The spatial organisation of the classroom could also shape opportunities for interaction in ways similar to those described by Mayblin et al. (2015), particularly where the arrangement of furniture prevented communication beyond one’s nearest neighbour and impeded mixed-group work (see Loader 2017). This could be exacerbated by scheduling issues, one example being where pupils from the host school arrived at the classroom first and took seats on one side of the room, and visiting pupils arrived some five or ten minutes later and took the remaining seats at the other side. Without intervention from the teacher, this physical arrangement of pupils within the classroom space could result in classes that were nominally mixed but involved little mixing.

When you go over there, it’s very separate, though … They’re at one side of the classroom and you’re at the other. (Male student, Catholic school, Whitecliff)

A third influence on ‘conversation’ was the location of shared classes and the place-based norms associated with this. Pupils reported that encounters on school premises could lack the fluency of those that occurred in other settings, such as the local activity centres where ice-breakers and occasional social activities were held and where classroom norms that discouraged social interaction did not apply. As Mayblin et al. (2015) similarly reported, this movement from the formal, rule-bound environment to alternative meeting places could facilitate interactions more easily. A small number of pupils also reported that, in a school associated with the other group, they felt less comfortable to initiate interaction and less willing to offer an opinion. Such proactive behaviours and expressions of identity were viewed as potentially unwelcome outside one’s ‘own’ place.

More favourably, the interaction that occurred through shared education was generally positive in nature. Pupils who reported engaging with outgroup peers described their classmates as pleasant and helpful and spoke of easy, informal interactions that permitted them to get to know one another as individuals. These more personal relationships appeared important in erasing boundaries as they reduced participants’ reliance on group-based judgements and responses. In a small handful of cases, where these friendships had continued in settings outside the school, there was also evidence of boundary ‘relocation’, with pupils finding shared identities as musicians or dancers that superseded their school or religious identities. Importantly, too, overtly negative contact was rare in shared classes, again because behavioural norms were clearer and more predictable than in other spaces of the school. Even so, a minority of pupils described intergroup dynamics in these classes as ‘awkward’ and uncomfortable, reporting that they struggled to find conversation topics or to approach a group of peers from the other school. Describing his discomfort, one pupil in Whitecliff contrasted his experience with that of peers who had existing friends in the other school.

[Pupils from the other school] just make you feel uncomfortable, so they do. But those people who know people over there, they get on quite well with people; but if you don’t, then you’re just sort of, you know, away. (Male student, Catholic school, Whitecliff)

In response to follow-up questions about the behaviours that made him uncomfortable, the pupil could offer no particular examples. Rather, it appeared that this discomfort emerged in the contrast between his feelings of being ‘out of place’ while in the company of others who were (or felt) ‘in place’ and displayed more relaxed and confident behaviours as a result. The responses from this student and others suggest that feelings of uneasiness in shared classes could entrench, or ‘inscribe’, the boundaries established through encounters elsewhere in the school. However, positive experiences of ‘conversation’ in the classroom could contribute to the relaxation or ‘relocation’ of boundaries, even where experiences in other spaces of the school were less favourable.

‘Difference’ and the (de)activation of boundaries

As an initiative with explicit reconciliation goals, shared education can provide space for the discussion of issues pertinent to division, including differences of religion, cultural expression, and national identity. Dialogue on these issues may represent a form of ‘conversation’ that can alter boundaries by increasing mutual understanding. However, it was evident from interviews that activities or discussions that drew attention to – and thus, in Tilly’s terms, ‘activated’ – the Catholic/Protestant boundary could cause discomfort among pupils, who feared provoking tensions or appearing ‘sectarian’. Consequently, only a minority of pupils reported having conversations about difference with their outgroup classmates; most sought to emphasise the similarities between the groups, acknowledging difference only in jest.

Interviewer

Do you think you could talk about [religious difference] with students from the other school?

Female Student 1

It’s just a bit serious … We don’t really want to talk about that in particular.

Female Student 2

I think if you did, like, you’d probably joke about it and nobody would take it seriously. (Catholic school, Whitecliff)

By avoiding these sensitive topics, it appeared that participants were attempting to deactivate the boundary between the groups. However, their responses elsewhere called into question the feasibility of this strategy. For one thing, it was evident that avoidance was not a passive act, but an active strategy that required continuous attention to the boundary and its norms. ‘Trying not to offend people’, as one pupil described, required constant vigilance over one’s behaviour in mixed groups and revealed the effort involved in negotiating difference sensitively (Neal and Vincent 2013). Maintaining such vigilance meant that the boundary was always activated in some form. Moreover, pupils observed that indicators of difference were apparent in the spatial and material configuration of the schools – in their architecture, the photographic displays (of British and Irish political figures, for example, or school rugby and hurling teams), religious iconography, and pupils’ uniforms. Attempts to deactivate boundaries by ignoring them seemed unlikely to succeed in environments where expressions of identity were so manifest.

Although most pupils preferred to avoid mixed-group conversation on issues pertaining to difference, a small number suggested that these discussions could be valuable in a shared setting. Their responses implicitly challenged strategies of avoidance, suggesting that social relations were best served by opportunities to learn about and learn from the other group. Several interviewees also described conversations where they had shared group-based perspectives and found the experience interesting and beneficial. These responses suggested that the short-term activation of group boundaries through these conversations could contribute to more substantive, long-term deactivation by reducing the anxiety and uncertainty around the boundary.

Female student 1

You hear them talking and occasionally say, like, ‘Mass’ or ‘chapel’ or something and … It’s actually more interesting, like. You’re like ‘what?’ And you have to [listen], you know, the way they explain it all.

Female student 2

She’d be quite a nosey person.

Female student 1

Yeah, I ask all about it.

Female student 2

She goes, ‘what’s Mass?’ and ‘why do you do this?’

Female student 1

I find it really interesting. I just find it really interesting, learning all about it. (Protestant school, Bellevue)

Female student

They wanted to know, like, what we see the Queen as, what it is she does. And then they talked about the Pope and going to Mass and stuff like that.

Interviewer

So did that just come up in conversation?

Female student

There was something on the TV at that time, and they asked us about it … We just said the Queen was just like their Pope. Like, just who they follow kind of thing. (Protestant school, Whitecliff)

At the same time, where some pupils were willing to challenge existing boundaries within the context of shared education, they found that their actions could be met with resistance outside the school. In these community settings, different norms governed both the potential for interaction and the forms of contact that were sanctioned. In the example below, a pupil at the Protestant school reported that she had been encouraged by teachers and pupils from her shared classes to attend a Gaelic football match. However, in discussions with people outside the school, she found that her actions were challenged for violating the mores of group separation and distinctiveness.

Female student 1

There’s some people who, like, I have said that I have went to the Gaelic football match to support them and they’d be like, ‘oh, why did you go to that?’

Female student 2

But that’s your own choice.

Female student 1

Yeah, that’s my own choice, but it wasn’t just my, like, my religion that people said that, but it was like other religions as well. As if to say, why did you go to a sport like mine? I’m just like, why can I not just go to support? (Protestant school, Bellevue)

This incident illustrates the context-specific nature of boundary change. While the boundaries within the school were sufficiently relaxed to permit attendance at the Gaelic football match, the response of the student's acquaintances and peers re-inscribed these boundaries in the community. Like the research of Mayblin, Valentine, and Andersson (2016), which found little indication that ‘meaningful encounters’ occurred across different social settings, this reveals how positive experiences of encounter in one place may be challenged or resisted in other settings and circumstances.

Conclusion

The purpose of this article has been to examine the influence of shared education on the boundaries governing Catholic-Protestant relations in Northern Ireland. While a psychological perspective has dominated previous research on shared education, the current study focuses on the social and spatial processes involved in sustaining division and how these are affected by the programme. In this respect, the research considers participation in shared education as an emplaced experience, influenced by the environment and the meanings that this holds for participants. To address the paper’s aims, the author conducted small-group interviews with 60 participating pupils in two school partnerships and employed Tilly’s (2004) social boundary mechanisms as a framework to examine the extent and nature of boundary change through shared education. This analysis identified multiple examples of boundary change, with more than half of Tilly’s mechanisms evident in the data. Importantly, these included mechanisms that created or intensified boundaries (‘encounter’, ‘activation’ and ‘inscription’) as well as those that relaxed and reduced them (‘conversation’, ‘deactivation’). Consequently, while shared education’s stated aim is to create ‘porous boundaries at school level' (Gallagher 2017, 44), the nature of boundary change in these partnerships was more mixed and could vary between situations and settings within the school.

Focusing first on the reduction or relaxation of boundaries, this study found that shared education provided pupils with new motives for boundary-crossing in the form of access to courses and qualifications that were otherwise unavailable to them – consistent with what Tilly terms ‘incentive shift’. Moreover, shared classes provided opportunities for personalised interaction and, for a minority of pupils, the formation of cross-group friendships, which in turn encouraged the relaxation and relocation of intergroup boundaries. These newly mixed spaces and nascent friendships, albeit within an otherwise ‘strongly classified’ space, appear practically and symbolically important in a context where only 1 in 14 children attends a formally integrated school and almost two-thirds of young people report that ‘all’ or ‘most’ of their friends share the same background as them (ARK 2016; DENI 2019). Finally, although most pupils were reluctant to approach issues of difference, there were indications that sensitive conversations on such topics could increase mutual understanding among pupils, helping to ‘deactivate’ the boundary and reduce the anxiety that accompanied it.

Examples of boundaries being created and reinforced through shared education were particularly apparent in the transitional spaces of the school, such as entrances and corridors. These offered opportunities for only fleeting contact that seemed to encourage group-based responses and intensified intergroup boundaries. This was also exacerbated by the requirement that pupils wear school uniform and, in Bellevue, by actions of the school that reinforced separation, such as requiring visiting pupils to use rear entrances that were locked on their arrival. Even in the classroom, where encounters were reportedly more positive, regular interaction was not the norm and a minority of pupils reported feelings of awkwardness. For this group, the experience in the classroom could lead to the inscription of boundaries that had already been activated in the other spaces of the school.

One of the aims of this paper has been to explore the spatial dimension of shared education, examining how the physical, material and social environment shapes the encounters within the programme and its contribution to boundary change. In line with Gieryn’s (2000) observations, the findings demonstrate that schools are not ‘just a setting, backdrop, stage, or context’ for pupils’ meetings ‘but an agentic player in the game’. First, following authors such as Dixon and Durrheim (2000) and Askins (2016), this study demonstrates how place, identity and emotion are intertwined in these schools and shape the interaction that occurs therein. Most participants had attended their school for three years or more and it was often a place of comfortable familiarity – with peer groups, institutional geography, and a shared religious and cultural identity. To create opportunities for interaction across schools, shared education disrupts these associations between place and identity, arousing feelings of apprehension within newly-mixed spaces.

For some younger pupils, this unease prompted the selection of subjects that required them to mix only at their own school, affording them the comfort of a familiar setting. Those pupils who attended the partner school spoke of their sense of being ‘out of place’, the impact of which was evident in the feelings of vulnerability they described, their sensitivity to others’ embodied behaviours (notably staring), and feelings of isolation in the shared class, each of which sharpened perceptions of the intergroup boundary. Such feelings of displacement were heightened by instances where group boundaries were reasserted, however unintentionally, through spatial practices such as locating shared classes in outbuildings. These findings demonstrate the dynamic ‘emotional geographies’ that participants negotiate during these encounters (Askins 2016), with different spaces evoking different affective responses that, in turn, intensify or relax social boundaries (Bondi, Davidson, and Smith 2016).

Second, this research continues in the vein of scholars of diversity and intercultural encounter who have examined how the spatial organisation and the norms of a setting affect the nature of encounters and thus their potential to influence social change (Askins and Pain 2011; Mayblin et al. 2015; Neal and Vincent 2013). In practical terms, interaction in these schools was facilitated or inhibited by the space in which it occurred (classroom or corridor, for example) and the material arrangement of the setting. The dominant norms of different spaces were also influential. While the classroom is posited as the site for interaction in shared education, conflicting norms that discouraged (particularly non-task-related) conversation could impede this, at least without intervention from the teacher. Where encounters were moved to other spaces with less formal norms, pupils reported finding interaction easier. Norms concerning behaviour-in-place also conditioned the conduct of visiting pupils, who were reluctant to initiate interaction or to express an opinion in a space that was not ‘theirs’. Even as shared education progressed and the norms governing cross-group interactions became more relaxed at school, corresponding norms in the community appeared to resist similar change (see also Hughes 2014), and the boundaries between these spaces appeared more defined.

Based on these findings, this article affirms the argument of Askins (2016) and Valentine (2008), that favourable intergroup encounters – those capable of altering boundaries between groups – do not emerge simply from spatial and temporal co-presence, but are subject to a range of social and spatial influences. Through the examples in this paper, we see how these different influences combine to challenge or sustain difference, facilitate or inhibit interaction, and thus inscribe or erase boundaries. With this in mind, those seeking to promote more favourable relations might consider how settings of encounter could be refashioned to minimise physical and symbolic barriers to intercultural contact. Following Mayblin et al. (2015), one valuable approach might be to involve participants in the creation of spaces for interaction and dialogue. Particularly in schools, where pupils have limited influence over the arrangement and use of classrooms and buildings, ‘the sense of ownership that comes with creating your own space’ (Mayblin et al. 2015, 71) may increase engagement among young people and foster more equal and collaborative encounters.

The challenge of creating environments conducive to contact also speaks to the wider tension facing shared education as it aims to break down social boundaries through contact in schools that have themselves been implicated in social division. Historically (and arguably still) part of a ‘place-grounded [social] order’ (Wallwork and Dixon 2004, 24), denominational schools are now positioned in shared education policy as sites of integration. Significantly, this new role requires no change to schools’ buildings, ethos or policies, other than to welcome children from the other religious group at particular times of the week. While Hughes (2014) noted a dissonance between the pro-contact values of schools involved in shared education and the separation norms of the wider community, this study also suggests a dissonance in the current arrangements where denominational schools are simultaneously symbols of separation and agents for integration. In light of this, and in view of the mixed findings of this study, one might question whether this approach could have more than a modest impact on intergroup boundaries. However, in view of the stalling of growth in integrated education and the lack of neutral settings for pupils to meet in, the more pertinent question is, perhaps, whether there is any feasible alternative to this approach. Even if shared education effects only modest change in boundaries, the symbolic value of shared spaces and mixed groups in settings associated, historically, with division should not be underestimated.

Disclosure of potential conflicts of interest

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Atlantic Philanthropies.

Notes on contributors

Rebecca Loader

Rebecca Loader is a Research Fellow within the Centre for Shared Education at Queen’s University Belfast. Her research interests include education in conflict-affected settings, educational inequalities and intergroup relations.

Notes

1. ‘Post-primary schools’ is the term used in Northern Ireland to describe all institutions of secondary level education. As the region continues to operate academic selection, ‘post-primary’ is preferred as a collective term to avoid confusion between secondary-level education as a whole and the specific sector of secondary (modern) – i.e. non-grammar – schools.

2. Until Key Stage 4, participation in shared education is not optional for most pupils unless their parents choose to withdraw them from shared classes and activities.