“Everyone knows me …. I sort of like move about”: The friendships and encounters of young people with Special Educational Needs in different school settings

This paper examines the peer-related social experiences and friendships of young people (aged 11–17) diagnosed with Special Educational Needs in four different school settings: two mainstream schools with special units and two special schools in Southeast England, UK. Findings from qualitative research involving young people with Special Educational Needs and adults, and participant observation, are presented. The young people had one or a combination of the following diagnoses of Special Educational Need: ‘Moderate Learning Difficulties’, on the ‘Autistic Spectrum’, and ‘Social, Emotional and Mental Health Difficulties’. We use the term ‘differences’ rather than ‘difficulties’ to express the interconnected socio-spatial construction of, and corporeality of, the experiences of these differences. There has been limited scholarship about the social experiences of young people with these diagnoses. In our study, young people’s experiences of friendships, exclusion, inclusion and bullying were socio-spatially shifting. Young people had varying experiences in the different school settings. In all settings, most had friends within the school, although those in special schools and units tended to have more friends within the school. However, bullying and ‘othering’ were also experienced in all three settings based on a variety of perceived ‘differences’. All young people needed opportunities for ‘encounter’ to forge friendships. Encounters are risky and can reproduce and reinforce difference as well as generating social connections and friendships. In many spaces, young people’s opportunities for encounter were constrained by the socio-spatial organisation of schools.


Introduction
In this paper, we explore the peer social relationships of young people in four schools, two mainstream high schools and two segregated special schools in the Southeast of England. The paper contributes to emerging geographies of education (Cook and Hemming, 2011;Holloway et al., 2010) by highlighting the experiences of young people with Special Educational Needs (SEN) in different school settings. There has recently been much debate about the appropriate places to educate young disabled people, and those with SEN, as the imperative for inclusive education has been challenged in many contexts internationally (Armstrong et al., 2016). An increased diversity of school settings for young disabled people and those diagnosed with SEN has emerged. In the UK, specifically England and Wales, the Conservative-led Coalition implemented new legislation (Department for Education (DfE), 2015a, 2015b), instigated by the Conservative election pledge to: '. . . end the bias towards the inclusion of children with special needs in mainstream schools' (Conservative Manifesto, 2010: 52). Parents, young disabled people, children and youth with SEN, and their allies contest the existence of such a bias; even at the height of inclusive education in the UK, in 2007, only 57.3 % of young people with statements of SEN 1 attended mainstream school, with 37.9 % attending special schools (DfE, 2015b). The new legislation gives parents increased choice over the location and type of school that young people with SEN can attend, including the right to request a special school. 2 Indeed, the numbers of students with statements of SEN or the new Education and Health Care Plan (EHCP) attending mainstream schools had decreased to 50.9% in 2015, with 43% attending special schools (DfE, 2015a). This represents a geographical shift away from the spatial convergence of the special and mainstream education institutions witnessed in the 2000s (Holt, 2007) towards a more plural landscape of education; this paper directly speaks to these changing geographies to explore young people's experiences within segregated special and mainstream schools.
Within debates about the appropriate place for the education of young disabled people, the experiences of young people themselves are often side-lined; this paper addresses this gap. Certain groups of young people are particularly marginalised from these discussions, including those diagnosed with: 'Social, Emotional and Mental Health Difficulties' (DfE, 2015a), known as Emotional and Behavioural Disorders in the USA; on the Autistic Spectrum (AS); and, with moderate, non-specific learning difficulties. We follow Parr and Butler (1999) in using the term 'differences' rather than 'difficulties', in a political attempt to destabilise the location of the 'difficulty' exclusively within the mind/ body of the young person. Whilst acknowledging that these differences can be experienced as embodied limitations and frustrations (see also Chouinard, 2012;Parr, 2011), this reframing emphasises the socio-spatial constitution of these differences, which emerge by contrast to socio-spatially shifting norms permeating schools and broader society. Also, shifting from 'difficulties' to 'differences' brings to the fore the creative potential of difference (Cockayne et al., 2016). This positive potential is inherent in emphasising the myriad differences and connections that divide and unite people, which can challenge powerful discourses that impose hierarchies on this multiplicity (see also Shildrick, 2005). For instance, as Bondi (2014) and Philo (2014) show, a troublesome mind/body can be experienced by individuals not diagnosed or labelled as 'different'.
In this paper, we examine whether the particular school contexts (special, mainstream, and segregated units in mainstream schools) are more or less socially inclusive for young people. The answer is complex. Young people need opportunities for encounter (Wilson, 2016) to forge friendships, but opportunities for encounters and how these unfold are influenced by the socio-spatial contexts in which they take place. Young people in our study who were educated in segregated special spaces -units or special schools, often had more friends than those in mainstream schools; however, bullying, exclusions and otherings also persisted in these spaces. In the paper, we contribute to the existing research about encounter, which has tended to focus on people who are ostensibly different to each other, particularly in relation to ethnicity, religion and/or sexuality. We examine the experiences of young people diagnosed with SEN, with others both with and without such a diagnosis. The young people were variously differentiated and connected in ways that both reinforce or challenge a variety of intersecting 'axes of power relations' (Butler, 1990) including (dis)ability.
The paper presents findings from qualitative research carried out with 46 young people who were diagnosed with SEN and aged between 11 and 16 years, participant observation in schools, and semi-structured interviews with 45 adults. The majority of the young people in this paper were labelled with, sometimes overlapping, diagnoses of: Social, Emotional and Mental Health Difficulties (DfE, 2015a), on the Autistic Spectrum (AS), and/or with mild or moderate, non-specific learning differences. In addition to being marginalised in research about education, disability and SEN, these groups are often excluded from, and within, schools (Broomhead, 2013;Jull, 2008;Youdell, 2010). These young people are also often the focus of negative discourses about the failure of inclusion (Warnock et al., 2010). A small minority of the young people had impairments similar to Down's Syndrome, sensory or physical impairments, and/or specific learning differences, such as dyslexia or dyspraxia.
The paper proceeds through four further sections. Next, we situate the paper in the context of the wider literature about geographies of education, young people's sociality and micro-spatialities of encounter. We then move on to discuss the methods. The fourth section details the findings of the research and is divided into four sub-sections focusing on: the variable importance of sociality and friendships to young people; multiple space/times of encounter; the limits to inclusion in mainstream spaces; and, finally questioning whether segregated spaces are more socially inclusive for young people. We end in the conclusion by reflecting on the contribution of the paper to debates about the appropriate 'space' of the education of young people and whether particular settings are more socially inclusive. We highlight the importance of opportunities for encounter, and emphasise that these are sociospatially situated, and constrained and enabled by the material spatiality of and everyday practices in schools.
School geographies of young people, mind-body-emotional difference and micro-geographies of 'encounter' Much policy and academic debate has focused on the most appropriate spatial location of education of young people with SEN, which ignites vitriolic debate (Armstrong et al., 2016). Geographers have made little contribution to this inherently geographical question. Recent research has begun to explore young disabled people's experiences (for reviews, see Curran and Runswick-Cole, 2014;Holt, 2016); however, certain groups of young people have not been fully included in this research, such as those on the Autistic Spectrum or with socioemotional and mental health differences. These critical studies have had little influence on educational policy and practice, and geographers have tended to shy away from explicitly addressing the question of the most appropriate 'spaces' of education. In this paper, we seek to address the question of whether one setting promotes more social inclusion than the other. The answer which emerges is inevitably complex, as much depends on micro-spaces of 'encounter' (Wilson, 2016) in different contexts.
Young people's peer relationships are important for two interconnected reasons. First, whereas young people often prioritise their friendships in school (Morrow, 2001), the importance of friendships is often understated in education literature (McCoy and Banks, 2012). Second, social relationships (re)produce, and have the potential to challenge, specific subjectivities which are differentially located on a variety of axes of power-relations (Ringrose and Renold, 2010;Thomson, 2005;Valentine, 2004). Young people's social relationships are therefore of crucial importance to their experiences of school and the ways in which they learn to become gendered/sexed, classed, racialized, (dis)abled, and so on (Thomas, 2011), possibly in new and creative ways, which challenge hegemonic identity categories and their associated inequalities.
Indeed, co-educating 'different' groups of young people is premised upon the idea that the 'othering' of difference can be transcended through young people's sociality, as young people are viewed as less entrenched in their negative views of each other as 'different' on racial, religious or ethnic, grounds. For instance, the recently published Casey Review into Opportunity and Integration (2016) states: 'If our children grow up playing and learning with people from different backgrounds, they will be less prejudiced, more understanding of difference, more confident and more resilient living in a globalised and connected society ' (p. 4).
Similarly, in relation to young people and mind-body-emotional difference, inclusive education has been underpinned by a belief that co-educating disabled and non-disabled young people will reduce dis/ableist attitudes when young people become adults (Centre for Studies on Inclusive Education, 2016).
By contrast, a recent secondary quantitative analysis of the Millennium Cohort Study has suggested that, in all contexts, young people with SENs experience relatively high levels of bullying (Chatzitheochari et al., 2015). This concurs with the previous geographical and social research with young disabled people which has critiqued assumptions that the colocation of 'disabled' and 'non-disabled' children will transform negative representations of difference. This literature demonstrates that young people are disabled by abelism -an insidious set of exclusions/marginalisations, based on the assumption of, and in contrast to, normative expectations of able bodies/minds (e.g. Campbell, 2009;Chouinard et al., 2010;Hodge and Runswick-Cole, 2013;Stephens et al., 2014), and we would add emotional states (Holt, 2010). These are (re)produced by, among other things, the interconnected everyday practices and material spatialities of schools, which are part of broader societies where dis/ableist attitudes abound.
To date, these discussions have tended to focus on young people with impairments, or, more recently learning differences (De Vet et al., 2012;Hodge and Runswick-Cole, 2013;Shah, 2012). However, including the experiences of young people with learning, socioemotional/mental health differences, and those on the Autistic Spectrum, highlights how these young people face the particular challenge that their 'differences' are in contrast to pervasive, insidious norms of behaviour, sociality and/or learning competence (although Gagen, 2015 emphasises how these are becoming increasingly formalised). We have argued elsewhere that it is helpful to consider a set of socio-spatially shifting ableisms, rather than a single, monolithic, ableism. This emphasises that despite their power, these ableisms are unstable and can change (Holt et al., 2012).
In this context, schools can provide 'spaces of encounter' (Wilson, 2014(Wilson, , 2016 or 'meaningful contact' (Valentine, 2008), which facilitate forging social relationships: '. . . that actually change values and translate beyond the specifics of the individual moment into a more general positive respect for -rather than merely tolerance of -others' (p. 325).
As Valentine et al. (2015) found in families, these encounters, sometimes simultaneously, challenge and engender respect for differences, and/or reinforce negative perceptions of 'others'. Wilson (2016) suggests that most studies about 'encounter': 'focus on the perspective of the majority or the powerful' (p.11). In this paper, we prioritise the perspectives of a relatively marginalised group, young people who are diagnosed with SEN. Drawing upon Cockayne et al.'s (2016) vision of a Deleuzian appreciation of 'difference-in-itself', we suggest that young people can make connections: 'through a multiplicity of microconnections that do not require or are not premised upon identification or opposition (relating difference to difference)' (p. 4).
In the following sections, after discussing the methods, we reflect upon young people's sociality and encounters in special and mainstream high (secondary) schools.

The study -Methodology and background
The data presented here come from a broader ESRC funded project (RES-062-23-1073-A). In the paper, we examine the experiences of young people defined as having SEN in four schools: two special and two mainstream high schools in two different Local Authorities (LAs) in the Southeast of England. 3 All the schools discussed here had low ethnic diversity, with most students being of white, British ethnic origin. The rural mainstream school had few students eligible for free-school-meals (an accepted, if limited, measure of relative poverty, partially replaced by the new pupil premium), and a nationally average, and locally above-average, proportion of students with SEN (both on the register and with statements). It had two special units, one for young people on the AS and one for young people with socio-emotional and mental health differences. The coastal mainstream school had a slightly higher than national average intake of students eligible for free-school-meals, and higher than average proportion of students on the register of SEN, with a slightly lower than average proportion of students with statements of SEN. It had primary-school style classroom for young people with all kinds of SEN, who were deemed to require this support. In both special schools, 100% of students had statements of SEN, with almost twice the national average proportion of students eligible for free-school-meals in the rural special school and a higher than average proportion in the coastal special school. More boys than girls experience socio-emotional and mental health differences and/or are on the Autistic Spectrum, and this is reflected in the gender balance of the schools and special units which were the focus of this research. All the young people discussed in this paper were aged between 11 and 16 years. Appendix 1 presents some demographic characteristics of the young people discussed in the paper. The young people have been given pseudonyms and some characteristics have been generalised to avoid potential identification. All schools and the LAs are also not named to preserve the anonymity and confidentiality of participants.
The research discussed here primarily consisted of in-depth participant observation (around 30 days in each school) and qualitative research with 46 young people. Semistructured interviews with 28 staff in the schools and LAs, 10 national level actors, and seven parents/carers are alluded to where relevant, although the focus is on young people's perspectives. The research with young people consisted of photo-interviews (Barker and Smith, 2012). The photography was undertaken individually, and interviews were either individual or paired with a friend depending on the young person's preference. The young people were shown how to use the camera and asked to take photos about their friendships and leisure activities. Photography and other visual methods are widely used in research with young people, as they allow young people to express themselves without relying on the vocal (Grant, 2016). Visual methods facilitate going beyond the immediately representational in language (Pink, 2013;Rose, 2016). There are limitations to the approach; photos are a static snapshot in time, which frame and constrain the visual field (Barker and Smith, 2012). As in much research which uses photos with children and young people, we used a photoelicitation approach, whereby the photos were discussed with young people in semistructured paired or individual interviews (Grant, 2016). Clearly, this brings the visual back within the frame of the representational in language. The photos did, however, prompt discussions which otherwise were unlikely to have taken place.
Full informed consent was gained from both young people and parents/carers prior to the qualitative research. All young people and parents/carers were also provided with details of the study and the opportunity to opt-out prior to ethnographic observation. We were acutely aware of ethical concerns and endeavoured not to cause emotional distress to young people. For instance, we asked who young people did things with, sat next to and so on rather than asking directly about friendships. Sometimes we followed this by questions about who friends were, if the young person seemed happy to discuss this. If young people started to talk about things that seemed to upset them, discussion was subtly directed to less sensitive topics. We were aware of our responsibilities in relation to child protection, and had prepared our response to disclosure of abuse/harm (Alderson and Morrow, 2011). Although young people sometimes told us troubling things (such as health harming behaviours like drinking, smoking or smoking cannabis), these were issues well known by the adults around them, and we did not feel that they were disclosures to us per se. Clearly, researching in-depth with young people about their social relationships engenders many ethical considerations that warrant in-depth further discussion (see Holt et al., 2017).
The interviews were recorded and transcribed in full. The research diaries were fully written up. Both were analysed via a broadly narrative approach, in which we sought not to find an objective reality but to understand (young) people's interpretations (Ansell et al., 2012). The research therefore gives insight into the lived experiences of young people with SEN in different spaces in different types of school. The key themes which emerged of relevance to this paper were: the variable importance of sociality and friendships; the importance of encounters, both face to face and 'virtual', which are often interconnected; limits to inclusion in mainstream spaces; relative social inclusion in special segregated spaces; and, social exclusion and bullying in all spaces.

Young people's peer relationships in segregated and mainstream spaces
The variable importance of sociality and friendships to young people Peer relationships are generally regarded as crucial to young people's experiences of school (McCoy and Banks, 2012), including young disabled people (Worth, 2014). Most young people cited friends as important and took photos of friends. For instance, friends from school featured highly on Kyle and Adam's (coastal special school) photographs, and they also talked about their friends; Adam states: 'Oh these are my friends at this school, I took pictures of them, that's Ozzie. Yeah, oh this is Bernard. That's Callum in a Michael Jackson pose'.
Although peer relationships were important to many young people, other aspects also mattered, such as subjects they enjoyed and relationships with key adults: 'The teachers are very nice <. . .> 4 because we're right next to a field and that's very nice . . .. 5 ' (Barnaby, coastal special school).
Friendships at school did not seem to be essential for some young people, however. Those who attended the AS unit in the rural mainstream school took relatively few photos of friends, and lots of things, pets and sometimes families. Most discussion of friends in the interviews with young people in the AS unit were prompted by ourselves. On reflection, we may have inadvertently reproduced normative ideas of the importance of friendships, and imposed these upon young people, by asking about friendships -albeit carefully and tentatively. We are also wary of reproducing dis/ableist assumptions about young people on the AS, which naturalise any social problems they may experience as 'part of their condition'. Indeed, although friends did not feature as highly on the photos of many of young people on the AS, as in those of some other young people, most claimed to have friends, best friends and good friends, presenting a more positive experience than Chatzitheochari et al. (2015) suggest. For instance, along with being mentioned as a friend by Adam, Ozzie (coastal special school) states: 'I do like my friends in year 10, they're probably the best friends I've ever had, and I do like people in other years as well'.
A small number of young people stated that they had no friends at school, although most suggested they had friends outside of school; Conrad (coastal mainstream school) states: 'Not, no, [I don't have any friends at school] except for my two cousins. Yeah, they're in year 9 and 10'.
Some young people's ideas of friendship were expansive and did not always tie to adult (and our own) preconceptions; however, the friendships which the young people had were of value to them. For instance, many young people, particularly, although not exclusively, those on the AS, enjoyed communicating with friends in cyberspace, and found this facilitating compared to face-to-face contact (Bertilsdotter Rosqvist et al., 2013;Davidson, 2008;Davidson and Parr, 2010;Henderson et al., 2014), as further detailed in the next section.

Multiple space/times of encounter
In all spaces, young people needed the opportunity for meaningful encounter (Valentine and Waite, 2012) to meet and socialise with peers to discover if they had shared interests and characteristics. Our participant observations provide insight into the micro-time-space geographies that can lead to such an encounter; Shelly, who had recently joined the coastal special school and Beatrice, started to become friends after sitting together in the classroom: Shelly and Beatrice came out and looked a bit unsure, then walked up and sat on the bench next to me and chatted . . . Beatrice asked Shelly who her friends are and Shelly said ''Polly, Ozzie, Owen -everyone really.'' Beatrice told Shelly the names of some of the other people in the school -pointing them out. Beatrice asked Shelly if she likes films and she said ''yes''. Shelly said ''do you like horror?'' Beatrice said ''no''. Beatrice said ''do you like funny films?'' And Shelly said ''yes''. . . Beatrice asked Shelly if is she allergic to anything and explains that she is/was allergic to the sunshine. Shelly says she is allergic to school. . . Shelly says ''are you going in?'' and follows her. (Research diary, coastal special school, playground) Shelly and Beatrice were demonstrably different from one another. Shelly had only been at the school for half a term. She was diagnosed with socio-emotional and mental health differences, was 15, seemed mature for her age and was looked after by the LA (living with a foster family with whom she reported a close relationship). Shelly wore make-up and in her interview she talked about boys, smoking and engaging in anti-social behaviour. By contrast, Beatrice was 13. She was diagnosed with a genetic impairment similar to Down's syndrome, and had learning differences. Beatrice was more interested in Disney films and cuddly toys than boys and antisocial behaviour. She lived with her mum and dad, who were involved in school and keen to participate in our research. In this excerpt, Beatrice and Shelly made connections through their myriad differences (Cockayne et al., 2016). It seems that they became closer friends during the period observing the school, as Shelly and Beatrice were often seen talking and sitting together in lessons.
Clearly, time and repeated encounters helped forge meaningful friendships, as suggested by Wilson (2016: 12): 'it is possible that encounters accumulate, to gradually shift relations and behaviour over time . . .'. A shared history and an awareness of connections across differentiations emerged through repeated encounters, whereas brief and superficial encounters might not have presented the opportunity for such in-depth knowledge and understanding. The time-space patterning of everyday behaviours within a school affects the opportunities for encounter between young people in the classroom, playground and school buildings, as can the presence of classroom assistants (see excerpt with Ali on page 9). We have suggested elsewhere ) that a range of 'minor' aspects of the organisation of the space/time of the school, ranging from transport to seating arrangements in classrooms, can hinder or promote opportunities for encounter. Often, given competing institutional pressures, teachers and other adults prioritised pedagogic concerns above young people's social encounters.
Longer-term and larger-scale movements also affected opportunities for encounter. Young people who had experienced multiple movements between schools had fewer opportunities for multiple encounters to forge friendships across their myriad of differences (Cockayne et al., 2016). Almost every young person who participated in the research had moved schools outside of key transition periods (i.e. the shift from primary to high school or high school to college), and sometimes multiple times (see also Broomhead, 2013). These moves were precipitated by formal exclusions or parents seeking to find an appropriate setting for young people, along with other reasons that affect all young people, such as family moves. Moving school outside key transition times can disrupt young people's education (Rumberger, 2014) and constrain their social relationships. Brandon (rural mainstream school) had only been attending the school for a matter of months, and discussed how many times he had moved school: . . . first school I believe was [Primary School 1], then it was I can't remember any of the other schools I went to temporarily, because I only went there for a short time, I wouldn't remember their name <. . .> I was there for a shorter amount of time than I've already been here, it was like a couple of weeks before I had to leave again <. . .> Then I went to some more temporary schools, like pupil referral units. <. . .> And then I eventually went to [Primary School 2] where I stayed till I was 11, then I went to [High School 1] until I was about, well a few months ago obviously, and then I had to come here.
This high mobility of students helps to explain the problematic social relationships they encountered. For instance, Brandon emphasised that he had not made any friends (yet): 'Well usually [I sit with] the teacher that's come in with me, and if I'm sat next to a student, it's just a random place that's, because I haven't really made any in-school friends' (Brandon, rural mainstream school).
Although Brandon had few friends in school, he had many online friends which he had forged through social media. Cyberspace provides scope for particular types of encounter, which do not rely on face-to-face communicative competence and is thus enabling for some young people, particularly, those on the AS (Henderson et al., 2014). Cyberspace connects people across space/time. For young people, this often facilitated maintaining friendships forged in previous school spaces or through leisure activities, as discussed by Lucy, a girl on the AS who was a national-level sports person: 'Yeah [I communicate on Facebook] mainly school and sailing, I've got a friend from Germany, I met her at the Worlds [championships]. . .' There are specific risks to young people's use of cyberspace, which are well documented (Ortega et al., 2013), although Holloway and Valentine (2003) emphasise that cyberspace is a medium for young people's variously supportive, powerful and conflictual social relationships. Some young people discussed conflictual relationships in cyberspace, which were often connected to their relationships in school and outside-school spaces. For instance, Ava was having a dispute with another girl on Facebook and Twitter, which culminated in a physical fight (discussed in the next section): There had been an altercation between Ava and another girl on Facebook and lots of tweetsalthough Ava's response had been that she wasn't going to be stupid enough to get drawn in because she'd get excluded . . . (Research Diary, rural mainstream school)

Mainstream spaces and the limits to inclusion
Since the opportunities that young people had for repeated meaningful encounters (Valentine, 2008) were crucial to their ability to make friends, different material spatialities influenced young people's peer relationships, by limiting their opportunities for these encounters. Many of the young people in mainstream schools spent some times in a special unit and some in mainstream classes. A small number had friends from mainstream spaces of the school; however, the majority of friends of the young people in our study were within their unit (see also Ytterhus, 2012): 'Yeah, mostly [my friends are] from down here [in the unit] <. . .> They're all from down here really' (Leroy, coastal mainstream school).
Further, some young people did not seem to have any good friends. Whereas some young people were isolated, and seldom seen with peers, others had friendships which were ambivalent, transient and particularly conflictual. As Simon (rural mainstream school) states: Everyone knows me. So I don't have like a friendship group in school, I don't hang around with like that group just every day. I sort of like move about <. . .>I don't sort of, because I get bored of people, I know it sounds weird but <. . .>.
At times, encounters between young people reproduced and reinforced difference negatively and resulted in bullying; encounters are risky (Wilson, 2014(Wilson, , 2016. These risks were arguably greatest for those whose subjectivities were relatively marginalised. For instance, in the rural mainstream school, mainstream classroom encounters for Ali (a boy diagnosed with ADHD and on the AS), were often constrained by, and served to reinforce, the negative expectations of difference from his peers (and his teachers): Ali started making tapping noises on the table, his legs were going and he had his shoes off and he also made whistling noises. These were made more obvious by the silence in the classroom (a point made by Jan [the teaching assistant] and the teacher as they chatted on the way out of the lesson). He banged his pen on the table and Jan kept on telling him not to . . . The noises and so on didn't cause reaction from the other children until about half way through the lesson. One or two boys hissed ''stop it Ali!'' at various points -mostly taking advantage of when the teacher was otherwise engaged . . . Once, Ali replied ''stop it Danny''. Jan throughout tried to stop Ali making noises -she was sitting on the edge of the desk, hovering over him as there weren't any spare seats. The teacher finally said ''we've all had enough Ali'' . . . Ali didn't speak to anyone else throughout the lesson -not even at the end when all the rest of the class were chatting. For example, the girls next to me were chatting about X Factor. (Research Diary, rural mainstream school) Occasionally, conflictual encounters between young people can even result in physical violence. For instance, the virtual altercation between Ava and another girl, discussed above, culminated into a physical fight: The Teaching Assistant (TA) and Unit Manager had been walking round school and they had seen a fight -there had been Ava and another girl at the centre of about 180-200 students. (Research diary, rural mainstream school) Most of the young participants with a diagnosis of SEN had experienced being bullied at some point in their school career, reinforcing the findings of Chatzitheochari et al. (2015). Generally, when young people discussed experiences of being bullied, these were in relation to schools they had left, often for this reason. For instance: 'Yeah, because all my old friends at my old school kept picking on me . . . Because I was the only one [they] didn't play with and everyone kept calling me names, I just ignored them <. . .> all swear words <. . .>. Interviewer: So is that why you changed school? Yeah . . .'. (James, coastal special school) Few of the young people reported currently being bullied in the mainstream school that they were now attending. An exception was Carl (rural mainstream school), who discussed negative experiences in both the unit and mainstream school spaces: 'Yeah, because most of the time usually they've been winding me up <. . .> Constantly. I'm starting to think that it's going to become a form of bullying or something <. . .> soon, because it's just been going on . . . Well they like say that I farted in dance when I didn't.<. . .> . . . one of them, it was like yesterday, and one of the people who's been winding me up since primary school in Year 2 said something quite racist actually to me, I don't know why.' This quote demonstrates that enduring encounters, repeated over time can also lead to entrenched and enduring differences rather than being more transformative than fleeting encounters (cf. Wilson, 2016). Indeed, a fleeing negative encounter might be less psychologically harmful than repeated problematic encounters, where [young] people might have to endure repeated exposure to being 'othered'. Interestingly, the 'racist' (xenophobic) comment made was about Carl (who is white British) having been born in a different European country, highlighting that the grounds of othering, as well as connections (Cockayne et al., 2016), forged by young people, can be around a myriad of subtle differences which escape the observation of adults, and therefore go 'under the radar'. The young person who had called Carl the negative name did not perceive this to be bullying, which emphasises the difficulties of defining bullying (Ortega et al., 2013). All the social relationships, across the study, were simultaneously inclusive/exclusive to varying degrees (Thomas, 2011). Most young people would be left out in some timespaces and included in others.

Relative social inclusion in special segregated schools?
Special schools and units were often identified by young people as 'spaces of acceptance' (Hall, 2005), although these are also spaces of containment of deviant others (Barker et al., 2010;Lea et al., 2017). Young people attended segregated special schools for many reasons, from formal exclusions, to parental choice, or expert advice. Parents of young people with SEN often select a special school for their child believing they will have more friends, be happier and appear less 'different' in special rather than mainstream schools (Bajwa-Patel and Devecchi, 2014); this was found in our study: '. . . I don't agree with sending Autistic children to mainstream schools because [in] the mainstream schools, they bully them' (Connor's mum, rural special school).
Our observations suggested that the special schools in our study did have more expansive norms of practice than the mainstream spaces: Owen and Barnaby are both independently running up and down the length of the field in bursts of speed. Barnaby grins while he runs. Owen makes noises. There seems to be more freedom than in the secondary school to do what you like. (Research diary, coastal special school, playground) However, as Runswick-Cole (2008) highlights, parental choice is constrained by the pervasiveness of personal tragedy models of disability and the limited inclusiveness of mainstream schools. In our study, the sense of being 'the same' in segregated spaces was counterpoised against being different in mainstream spaces: 'No, because I used to have fits and that and I was, I wasn't like anyone else, so it made me feel a little weird, but now I'm at this school I have friends who are like me' (Lenny, rural special school).
Despite these narratives of sameness in special schools, the only shared characteristic of many of the young people in a special school is an SEN diagnosis, and this is often not the same diagnosis (Farrell, 2006). Clearly, even young people with shared diagnoses have differences between them in terms of level and type of 'difference', academic ability, class, gender, race/ethnicity, religion, and personality. As in the mainstream schools, young people in the special schools often forged friendships around shared characteristics, empathy, personal similarities, shared interests and hobbies, rather than a shared diagnosissuggesting that they were not bounded by the strictures of representation and were open to the creative multiplicity of difference (Cockayne et al., 2016). For instance: 'And plus Mason is good drawing and a good, and he wants to be a game designer. <. . .> And plus he's good at making up stories and games and characters. <. . .> And that's what I do also' (Sam, rural special school) These friendships provided emotional capital -a resource constituted from the emotional interconnections of peer and other relationships, and social capital (Holt et al., 2013;Reay, 2004). Emotional capital is integral to social capital, since many social relationships are affective, non-purposeful and engaged in for emotional support and reciprocity rather than for rational and strategic gains (Holt et al., 2013). Most of the young people who attended special schools, had friends and good friends in this context; as suggested by David (rural special school): '[At lunchtimes] I just tell my best friend some jokes <. . .> And that's it <. . .> Yeah, at playtime we <. . .> Yeah, we just made some new games, fairies . . .'.
Nevertheless, as in the mainstream schools, social relationships in the special schools simultaneously included and excluded young people. For instance, in the rural special school, many of the young people played football; inclusion was based on competence, and was gender-inclusive: When the ball is kicked towards Andy (boy), he cowers away from it. Asha (girl) is obviously skilled and dribbles the ball, scores and celebrates vocally. She put her hands in the air, and is screaming ''yes''. (Research diary, rural special school, playtime/recess, on the field) Some young people were, however, excluded or marginalised due to their relative physical inability, as Andy in the quote above. Andy emphasised that he had few friends and disliked most of the other young people in his class: Our observations and interview with Andy's mother and other young people also suggest that Andy was excluded and marginalised in the school. Other young people mentioned not liking aspects of Andy s behaviour, such as his use of bad language. Although special schools might be more expansive in their norms of acceptable practices, limits of (un)acceptable behaviour, and an idea of the 'normal' and who is outside of the norm (Foucault, 2003;Philo, 2007) also pervade these spaces. It is also possible that Andy's body-shape influenced this bullying, as he was overweight, which his mother discussed in the interview (Colls and Evans, 2014;Hopkins, 2012). Clearly, in special schools young people include, marginalise and excluded each other on a variety of grounds which reflect those in mainstream school spaces. This might be focused on 'unacceptable' behaviour; however, other differences are also reproduced negatively, such as body morphology.

Conclusion
Debate continues as to which are the most appropriate spaces for educating young people with disabilities and SEN, and there have been recent shifts towards a more diversified landscape of special education, allowing parents to select a segregated special school setting for their child. We have found that most spaces both constrain and enable young people's sociality and that experiences differ in mainstream spaces, special units and segregated special schools. Often, young people had more and better friends in segregated special spaces, both units and special schools, but particularly in special schools. However, of course, in special schools the friends made were exclusively other young people with diagnoses of SEN, with implications for social and cultural capital. This was also largely true of the mainstream settings in which the research was conducted, where most of the young people with SEN attended a special unit for at least some of the time, and were mostly friends with other young people from this unit, who also had SEN. (This was not always the case across the study -see Bowlby et al., 2014.) Special schools are not of course a social utopia, and exclusions, marginalisations, and stigmatisations were seen to persist in the segregated special schools and units. The apparent 'inclusiveness' of segregated schools was, moreover, emphasised by contrast to exclusionary processes linked to SEN at play in mainstream schools.
In all contexts, young people needed space to 'encounter' other young people, in order to forge friendships, although these encounters were risky and might lead to negative 'otherings' as well as friendships. Wilson (2016: 15) claims that: 'encounters are mediated, affective, emotive and sensuous, they are about animation, joy and fear, and both the opening up and closing down of affective capacity'. They are also mundane, and happenstance, habitual and not particularly remarkable moments. The importance of the encounter might not be obvious to the people involved. Certainly, affective but not necessarily animated -just two young people happening to share a bench, starting a conversation, and finding they like each other. Importantly, most scholarly discussions of 'meaningful encounters' focus on meetings between groups who are ostensibly and hierarchically different, often on ethnic, racial or religious grounds; however, some of the encounters presented above are between young people who are not evidently different. Nonetheless, the encounters are meaningful, forging friendships and associated social and emotional capital across the myriad of (sometimes subtle) connections and differentiations that separate and unite people (Cockayne et al., 2016).
It is therefore important that teachers and educational policy-makers recognise the potential significance of encounter and how the material spatiality and mundane both practices of adults in different schools facilitate and constrain encounters between young people. Individual schools are tied to broader institutional constraints, policy measures, statutory requirements, and influenced by the backgrounds, experiences, dispositions, capitals, and so on, of the young people and adults in them (De Vet et al., 2012). Schools are increasingly subject to neo-liberalisation -implementing efficiencies, reductions in (SEN) funding, competition, and (inter)national league tables. In England and Wales, new curricula have been introduced in primary and secondary schools, which emphasise key skills, testing, a reduction in coursework, and inculcation of 'important facts' -which has been likened to the pedagogy of the School Master Gradgrind, in Dickens' Hard Times (Garner, 2013). In these circumstances, it will take creative and dedicated professionals to support the inclusion of young people with SEN in schools and to prioritise young people's social relationships in school design and everyday practices (see also Galton and MacBeath, 2015).
More hopefully, we have found that when young people have opportunities for meaningful encounters, they can sometimes forge social relationships that connect them across a myriad of differences, and which challenge the more general pattern of association between a diagnosis of SEN and negative peer relationships (Chatzitheochari et al., 2015). Paying attention to how meaningful encounters are enabled could allow an understanding of how contexts in all school spaces might be able to facilitate meaningful encounters. Here the broader spatiality and temporality of schooling of young people with SEN is significant, as whether young people attend segregated special schools, mainstream schools or special units in mainstream settings and/or frequently move school influences the encounters which can take place. Young people have little control over these broader spatialities.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Economic and Social Research Council [grant number RES-062-23-1073-A]. Notes 1. At the time of the research, there was a graduated, step approach to identifying and intervening in SEN, from School Action, and being placed on the register of SEN to 'School Action Plus' support, which involved agencies outside the school. The highest level of intervention was a 'statement of SEN' -a statutory document setting out the support the young person must be provided with as a legal right, which often included financial support. Under the new Code of Practice (DfE, 2015a) School Action and School Action Plus have been merged into SEN Support and Statements are being phased out to be replaced by Education and Health Care Plans (EHCP), which are meant to provide a more holistic package including education, social and health care. 2. The Manifesto states the aim to: 'Give every parent access to a good school'. This demonstrates how parents are often viewed as proxies for their children's needs and desires in education policy.
Presumably it is not the parents who will be accessing schools. Norwich and Eaton (2015) argue that the new policy framework demonstrates more continuity than change, with many of the initiatives building upon previous policy, and missing the opportunity to fundamentally challenge underpinning, conceptualisations of SEN and disability. 3. Local Authorities (LAs) are the local level organisational bodies of under 18 education in the UK.
Although much education policy has been centralised to national levels (England, Northern Ireland, Wales and Scotland), LAs maintain control over some aspects of education, such as the delivery and management of SEN policy and admissions policy for schools still within LA control. 4. <. . .>: This symbol indicates cut researcher prompting, which has been cut in the interests of brevity for the paper. Full transcripts are available on the UK Data Archive. 5. '. . .' indicates cut text from the interviewee.