Gender, class, race, ethnicity and power in an elite girls ’ state school

Selective state grammar schools are the subject of sustained political debate surrounding issues of standards, education quality and social mobility, and yet they have received little academic scrutiny in geographies of education. Increasing numbers of young people are educated in selective settings in both the UK and globally. In this paper, we argue that some selective state schools are ‘ elite ’ spaces, whose alumni hold disproportionate power and sway. This paper examines the social geographies of girls in an elite grammar school in the Southeast of England, examining how classed and ethnic/racialized femininities are performed and enacted. The data are drawn from semi-structured photo-interviews and focus groups with 23 girls aged 13 – 14. The paper examines how the girls ’ social geographies were forged by socio-psychic process of connection and di ﬀ erentiation. Class di ﬀ erences were abjected onto non-grammar school ‘ others ’ , and poverty was viewed by some girls as a moral failing. The girls were avowedly open to ethnic, racial and religious diversity, which generated a cosmopolitan sensibility as a cultural resource. Nonetheless, subtle di ﬀ erences were reproduced through friendships, which along with being emotionally nurturing, were fraught and fractured in power. These di ﬀ erences can involve subtle hierarchical performances of ethnicity/region/race, which operated beyond the immediate conscious re ﬂ ection of the girls at times, pointing to a ‘ deeper domain ’ (Philo and Parr, 2003) which can be a friction to allenging enduring relations of di ﬀ erence through the spatial contingency of encounter. Given the powerful positions these girls are likely to occupy in top professions, how they understand and perform class, gender, ethnicity/race and religion are crucial. This in-depth study has theoretical resonance to elite spaces beyond the speci ﬁ c context of the case-study school by illuminating processes through which speci ﬁ c and hierarchical subjectivities are forged in friendships and by identifying the ‘ same ’ and ‘ other ’ .


Introduction
This paper examines the social geographies of girls in an academically selective, non-fee paying, state funded, grammar school in the Southeast of England, which is given the pseudonym 'Manor School'. There has been little geographical research examining state selective grammar schools. Within the UK context, a significant minority of young people attend state grammar schools, with around 5% of young people under the age of 16 in England educated in 163 state grammar schools (Bolton, 2017;DfE, 2017). There are moves to expand grammar schools in England (DFE, 2016;George, 2018). Further, levels of selection are much higher internationally (OECD, 2016: 77), with pressures to increase selection in light of the globally competitive 'results' environment (Ball, 2015;Alderson, 2017).
This grammar school can be viewed as an 'elite' school, as entrance is gained via an examination taken at 10 or 11 years of age (the 11 plus), and it is one of the most academically successful state schools in future.
In this context, we explore the peer-group relationships of twentythree girls, aged 13-14 years, in a single-sex English state selective grammar school, to examine how they enact and perform privileged classed, gendered and ethnic/racialised femininities. The school is a site of multicultural 'encounter', with an ethnically mixed student population (Abbas, 2007;Basit, 2013;Shah et al., 2010). High expectations are placed on young people's conviviality and social mixing to transform entrenched ethnic, racial, and religious differences (Casey, 2016), in both the present and in the futuresince young people are often viewed as the future of societies (Edelman, 2004) and are embodied social beings and becomings (Holloway et al., 2018). We highlight that a 'cosmopolitan sensibility' is forged amongst the girls, who are open to ethnic, religious and racial differences, and that this is a resource or form of capital. By contrast, socio-economic difference, and particularly poverty, is externalised beyond the school and disavowed, with the girls reproducing negative ideas about poverty which define the poor as 'other' and link socio-economic hardship to moral failing. These successful girls were kind and polite, and conflict between them was usually subtly played out. Despite this, conflicts and power relations did exist, which were subtly expressed and which sometimes reinforced hierarchies in relation to popularity, or ethnic, religious and racial differences.
The argument proceeds through five principle sections. Next, we highlight the concepts of elite schools and the (re)production of privileged classed, and racialised/ethnic femininities. We emphasise that state grammar schools in England have been rather overlooked as elite spaces, as have the intersecting gendered, racialised and classed subjectivities which emerge through everyday sociality within elite school spaces. In the subsequent section, we set out the methods and data of the paper. The fourth section, drawing upon our qualitative data, explores key themes about the (re)production of intersecting gendered, classed and ethnic/racial identities in the girls' social relationships. The fifth and final section offers a discussion and conclusion.
2. Class, femininity, race/ethnicity and schooling 2.1. Selective state elite schools in England and geographies of education inequality and privilege Geographers and social scientists have sought to examine the mechanisms by which classed and ethnic advantage and disadvantage are maintained through the field of school-level education (see for instance Butler and Hamnett, 2012;Ball, 2003;Jenkins et al., 2008;Reay et al., 2011). The ways in which school-level education reproduces privilege and disadvantage stands in stark contrast to ostensible political endeavours to promote social mobility through school education, globally (OECD, 2012), and in the UK (DFE, 2016). Geographers have highlighted the spatiality of processes reproducing educational and socioeconomic inequality in the state sector, as middle class parents deploy their social, cultural and economic capitals and residential strategies to gain access to Ofsted rated 'outstanding' state comprehensive schools (Butler and Hamnett, 2012;Cullinane et al, 2017;Smith and Higley, 2012). Little attention has been paid to the investments of middle class parents in equipping and sending their children, often over relatively large geographical distances, 1 to selective state schools. There is very little geographical scholarship about the experiences of young people within these classed geographies of schooling, and the ways in which young classed subjectivities are forged within specific school spaces. This paper addresses this gap, within the context of an elite school space (Waters and Brooks, 2015;Sparks, 2016).
There are clear shared characteristics of 'elite' schools (cf. Ball, 2013). Elite schools are vehicles for the reproduction of the cultural privilege of the upper and upper middle classes, including the 'state nobility' (Bourdieu with de Saint Martin, 1996); who Kenway and Koh (2013: 274) identify as: "the dominant or ruling class formed in significant part through elite education rather than through direct reproduction via economic wealth or family power". Importantly, as opposed to the direct handing down of political power or economic wealth through families, the reproduction of privilege through cultural capital and education is hidden (Twine and Gardener, 2013) via: 'an operation of a social alchemy by which a social hierarchy dissimulates itself…. a historically arbitrary social order rooted in the materiality of economic and political power transmutes itself into what displays every outward appearance of an aristocracy of intelligence' (Wacquant, 1996: x).
There are arguably two archetypal 'elite' educational spaces, which have come to forge current understandings of what elite school spaces are, and around which elite schools often model themselves. The first are 'elite' by 'ability'; schools (and universities) which are highly ability selective, but which are also socially, culturally and economically exclusive, and serve to reproduce the advantages of specific factions of the upper and upper middle classes (see Bourdieu with de Saint Martin, 1996;Bourdieu and Passeron, 1990;van Zanten and Maxwell, 2015, in France;Deppe and Krüger, 2016 in Germany;and, Kenway and Koh, 2013 in Singapore). The other model of elite schooling is fee-paying, academically selective, English 'public' schools, such as Eton and Harrow Aggleton, 2016a, 2016c;Waters and Brooks, 2015; see also Ayling, 2016), which have used this specific 'national' characterization to attract an increasingly international student population (Waters and Brooks, 2015).
English state grammar schools are particularly interesting, as they represent a type of elite schooling which is prevalent, and increasing, in much of the globalizing world, within the national setting which provides a model and reference-point for fee-paying elite 'public' schools (Maxwell and Aggleton, 2016a; see also Ayling, 2016). They are second only to high-ranking fee-paying schools in the proportion of students to graduate to the most prestigious universities (Sutton Trust, 2008). State funded selective schools do not charge fees; however, they are highly selective in terms of socio-economic background of their student population (Jenkins et al., 2008). Middle class parents often deploy their cultural and economic capital by paying for private primary education, private tuition, or moving to locations within grammar school catchments to strategically gain access to grammar school education (Andrews et al., 2016).
There is a particular spatiality to selective state grammar schools, which are present only in specific Local Authorities in England and Northern Ireland. There are initiatives to increase the number of grammar schools (DFE, 2016;George, 2018) and selection is becoming more widespread as academies (exactly the kinds of 'outstanding comprehensive' schools discussed above) are now allowed to select up to 10% of their students, under the Schools Admission Code (DFE, 2014). Although relatively limited in geographic scope in the UK, schools with such selective admissions criteria are more pervasive across the globalising world; indeed: 'across the OECD countries, up to 43% of students are in academically selective schools' (OECD, 2016: 77). Globally, school education is increasingly governed by neoliberal agendas and rationales. Schools compete to achieve high grades, be effective, and attractive to the highest achieving students (Ball, 2015). A perception can pervade that selection in education will enable the most 'able' students to achieve (DFE, 2016), following 'effective' selective systems, such as Singapore, which regularly tops the OECD Programme for International Student Assessment rankings (PISA, OECD, 2018); however, Gorard and Siddiqui's (2018) detailed empirical research challenges this assumption. In this broader context, the ways in which girls in a state non-fee-paying selective school (re)produce and perform privileged, and subtly hierarchical gendered and 1 Some of the girls in this school travelled an hour each way to get to school (Andrews et al., 2016). racial/ethnic, identities is intriguing.

(Re)producing gendered, classed and ethnic/racial subjectivities in elite school spaces
Privileged subjectivities are intersectional, and studies have highlighted the racial, ethnic and gendered aspects of young people's emerging identities in elite schools (e.g. Gaztambide-Fernández et al., 2013;Chase, 2008;Maxwell and Aggleton, 2014). Focusing upon the: 'centrality of the girl as a figure in the production of normalised subjectivities', Allan and Charles (2014: 335), explore the 'classed femininities of private girls' schooling in Australia and the UK, wherein girls perform the subjectivities of the 'successful girl' (Ringrose, 2007) motif of privileged middle class feminine subjectivity, and denigrate its abjected working class 'other'. As Baker (2010: 2) highlights there is a 'ubiquitous media representation of an unbalanced educational arena in which girls are outperforming boys (and muscling in on their territory of educational success)'. The pervasive discourse of the successful girl is somewhat at odds with young women and girls' lived experience, and Walkerdine (2003) and others have pointed out that the 'successful girl' discourse conceals continuities with the past in terms of gendered inequalities in paid and reproductive work, life trajectories and aspirations. Similarly, Strand (2016) among others questions the empirical basis for these media discourses, and argues that educational success is complex and multifaceted, but that success is still more closely tied socio-economic background than gender.
Allan and Charles (ibid.) draw upon Althusseur's concept of interpellation, especially as evoked by Butler (1997) and McRobbie (2009). Although the authors do not fully explain what these interpellations are, from Butler (1997) we understand this to be a process of bringing a subject into being in a social sense through language, as people are 'hailed' (called and named) and respond or take on that name. These interpellations: ''hail[s]' and construct[s] the normative, middle class girl subject …. At the same time, silently infer[ing] the failed 'other' who is required for the production of the norm' (Allan and Charles, 2014.342). Few studies have examined the intersection between gender and race/ethnicity, and this paper addresses this gap.
Taking a different 'line of flight' from Allan and Charles (2014) and also drawing upon Butler (1997), we examine processes of 'recognition' and 'abjection', which are open to more embodied and emotional figurings of the subject/agent, in the forging of gendered, classed and racial/ethnic subjectivities. The concept of 'recognition' draws upon Butler (2004), particularly her reworking of Jessica Benjamin's (1988) account of socio-psychic relations. Butler's notion of recognition emphasises the importance of a psychic and emotional need of people to have relationships with others in explaining the performance of 'appropriate' subject positionings (see Bondi, 2005;Holt et al., 2013). Rather than theorising socio-psychic relations dualistically, Butler (2004) suggests that psychic relations to others are in a constant tension between competing desires for mutual recognition, and to conceive the other as outside and distinctive to the self. Benjamin views psychic life as ''vacillat[ing] between 'relating to the object and recognizing the outside [O]ther'' (Benjamin, 1998, cited in Butler, 2004. As Butler (2004: 132) states: 'communication becomes both the vehicle and example of recognition'arguably, as in 'geographies of encounter' (Wilson, 2017;Valentine, 2008;Valentine and Waite, 2012) by communicating with the 'other', we can then 'recognise' them as a social subject (we can of course abject themcommunication does not necessarily lead to recognition).
These socio-psychic-spatial interplays of power resonate with psychoanalytic geographies (Kingsbury and Pile, 2014), which unsettle self-conscious and rational notions of human agency (Davidson and Parr, 2014). In this field, Bondi's (2002) illumination of empathy and identification similarly focuses upon relatively positive socio-psychicspatial relationships, as opposed to 'othering' and 'abjection'. Bondi's (2002) paper emphasises the role of empathy and identification, along with projection, in understanding relationships between the researcher and the researched; these: 'unconscious processes of introjection and projection, which operate as dynamic exchanges within all interpersonal relationships' (Bondi, 2002: 64) are, we argue, of equal import to the socio-psychic-spatial/interpersonal relationships between the girls that we are examining in this paper, and have broader theoretical resonance. Drawing upon Melanie Klein (see Klein, 1997) and Donald Winnicott (1965Winnicott ( , 1971), Bondi (ibid.) emphasises the permeability of the relationship between a person and 'the other'. She also focuses upon the role of empathy, as an: 'act of imagination required to recognise the other person's feelings' (Bondi, 2002: 71).
In this paper, we focus on processes of identification and the emotional imaginative sharing that is empathy, along with exploring process of 'othering'. We examine how complex socio-psychic-spatial relationships are typified by elements of connection and disconnection, identification and 'othering', forged around emotional needs to be recognised in a social context. This critiques celebratory notions of encounter, which emphasises the role of encounters in breaking down enduring socio-cultural differences (see also Holloway et al., 2019). At the same time it also cautions against the polar-opposite, by emphasising that encounters between any individuals or groups are forged by interconnected processes of both identification and othering, underpinned by an emotional need for recognition.

Methods and data analyses
This paper draws upon qualitative research with twenty-three girls in a selective girls' grammar school in the Southeast of England, conducted in the early weeks of the summer term 2012. The school is given the pseudonym 'Manor School' and the girls' names are also pseudonyms. The local authority and town in which the school is located are not named, as this would identify the school and potentially the girls. The school is in a large town, with levels of ethnic and socio-economic diversity broadly reflective of England as a whole. According to its website, 97% of students gained 5 GCSEs graded A * -A including Maths and English in the most recent 16+ exams. The school is identified as one of the best performing state schools in the country, with girls generally transitioning to elite Universities. At the time we selected the case-study schools, only 0.7% of the students at Manor School were eligible for Free School Meals, an accepted, although imperfect measure of poverty (Ilie et al., 2017), whereas the local mean of eligibility for Free School Meals was 14% (School Census, 2008), and most of our participants were from established and technical middle class backgrounds (Savage et al., 2015), although there were differences between them (see Appendices A and B). The student population was ethnically mixed, and more diverse than the Local Authority mean, with 63.2% of the population from white backgrounds, compared to a local mean of 67% from white backgrounds, and a national mean of 82% white ethnic backgrounds (School Census data, 2008, Table 3.1). The school has parallels to high-status-fee-paying schools (Maxwell and Aggleton, 2016a) in both is physical characteristics, with a quad and tennis courts, and its curricula, with subjects such as Latin and classical history being offered. The research consisted of three key stages: first a small (maximum of four) self-selected, friendship based, focus group; second, a period of self-directed photography, for which girls were offered digital disposable cameras; and third, a self-selected paired or individual interview to discuss the photographs. Visual methods are commonly used with young people in conjunction with more traditional methods (Grant, 2016), as they are seen to facilitate engaging with different ways of communication, to enhance the inclusion of young people in research. Visual methods go beyond representational frames of language (Pink, 2013;Rose, 2016). There are limitations to the approach as the photographs encapsulate a specific moment in space and time, and can heighten the importance of this moment above other times, spaces and events (Barker and Smith, 2012). We used a photo-elicitation approach, whereby the photos were discussed with young people in the interviews (Grant, 2016). Although clearly limiting the analysis of the visual within the frame of the young people's language, this enabled us to ensure that the meanings interpreted from the photographs were the young people's own interpretations. The photo elicitation stimulated discussion about different aspects of social life and gave a deeper insight into certain elements of the girls' friendships. In the interests of preserving the anonymity of the girls, any faces would need to be blurred out, consequently making the photos of poor visual quality and they are not therefore directly presented in the paper.
The researcher attended classes to introduce the research, with girls invited to a further briefing meeting if they wanted to participate. Here, they were given written information materials targeted at both parents and the young people, and invited to ask questions at any time. Written consent was gained from both parents and the girls. Girls selected their own friendship-based focus. Despite promoting a relaxed and less daunting atmosphere, this strategy had potential ethical issues; the focus groups became a semi-ethnographic space in which relationships between the girls were played out in ways that might have influenced their on-going social relationships (see Valentine, 1999, in relation to couples). On balance, we would argue that although the focus groups occasionally allowed tensions to be expressed (see for instance Section 4.4), they did not cause those tensions which were already present among the girls.
Given the qualitative nature of the research, we do not seek to make statistically generalisable claims about social relationships in girls' grammar schools. Rather, we suggest that the detailed in-depth accounts of the girls' social relationships and the ways in which their subjectivities are played out in these relationships in the school spaces, has potential theoretical resonance beyond this particular case-study (Alderson, 2017). The interview and focus group data were analysed via a 'critical' thematic analysis. All interview and focus group transcripts were closely coded, and then key themes were identified across the dataset (Gibbs, 2007). The analysis was abductive, involving both 'theoretical/conceptual' themes emerging from the literature review and research questions and 'invivo' themes emerging from the voices of the participants (Mason, 2017). The visual data was similarly analysed via a critical approach (Rose, 2016), although the key analysis of the visual data was by the young people themselves, in relation to how they discussed the photos they had taken. When analysing the data, we reflected upon how people are always subjected in power, that agency emerges within the context of this subjection, and that inequalities in relation to economic, social and cultural capital are also constituted within subject positioning in beyond conscious ways which might not be evident to the researched or indeed the researcher. We have been honest, critical and reflective in our analyses, and share our interpretations of the data presented. We realise these data can be interpreted in multiple ways, and that the reader may not concur with our own analyses. The full transcripts are available via the UK data archive.
The key themes discussed in this paper are: the (re)production of a 'successful girl' femininity in an 'elite meritocracy' and the simultaneous 'abjection' of the 'poor' non-grammar school young person; diversity as ethnic diversity forging cosmopolitanism as a resource; enacting privileged, regulated and 'nice' femininities in the hallowed spaces of an elite school via processes of mutual identification and empathy; and yet (re)producing power inequalities through subtly hierarchical acts, coalescing around ethnicity.
4. (Re)producing gendered and ethnic/racial/religious identities in an elite school space "My brother's been my closest friend, that I've kind of become more boyish than girlish, so I find it hard to sometimes make friends because I find like I don't want to talk about all these girly conversations and petty things, and I just want to have a normal conversation with someone. So I guess I become more boisterous, if that makes any sense at all" (Sada, focus group).
In this high-performing and selective school, the girls discussed how there was a pervasive atmosphere of being 'the top' with high academic expectations, redolent of a 'successful girl' subjectivity (McRobbie, 2009;Ringrose, 2007): "… in our school there's this whole big, I feel like there's a big atmosphere where the teachers are telling us we're really, really, really clever and we're like top…" (Sabelle, focus group).
Other girls talked about being 'driven'. All the girls had high academic and career aspirations, aiming for elite universities and professions, including law and engineering. In the grammar school, the nature of the elite space is specific, given entrance is predicated on a measure of academic ability, rather than fees. Some girls expressed a sense of achievement in gaining entry into the school: "I mean it's a grammar school so about, only 96 students are selected and out of, loads of people take the exam for this school because it's such a good school, and they take the exam because they want to be here, to do well in their studies, and yeah I got chosen! And I was really happy because I studied really hard for it. Because people, like they study for this exam from year 4, year 3, just for the exam in year 6" (Paavai).
This sense of meritocracy, as the girls had worked hard to gain entrance to the school, conceals the limited socio-economic diversity of such grammar school intakes, giving legitimacy to an 'aristocracy of intelligence' (Wacquant, 1996: x). This academic potential is underpinned by a scaffolding of supporting mechanisms to which families with higher levels of cultural, social, and economic capital have more access; ranging from private tutors (Andrews et al., 2016) and private primary schooling, to understanding that passing the 11 plus requires dedication and studying from an early age.
Being 'in place' in the academically elite space of the grammar school did not involve a one-off passing of a test; rather being a 'successful girl', needed to be continually worked and reworked; high academic achievement was both facilitated and expected. These high expectations lead to intense academic pressure. It is intriguing that a B mark is seen as a failure, moving them out of the 'A1 girl ' Strand (McRobbie, 2009;Allan and Charles, 2014). This was, however, seen to be relational, depending on the social and spatial context within the school: "…But I think it depends, it depends who you're sitting next to or like who you're working with when you get the mark back because some girls will be like, oh you got a B, that's really good (hesitatingly), and some will be like yeah, no, that's awesome, well done you, good for you, well done!" (Emma, focus group).
Importantly, however, any mark lower than a B was disavowed and externalised beyond the space of Manor School; as Lucy states: "…but I've got some friends from year 10 [in other schools] but they think that an E's good and then I feel really awkward because I think that's not that good". (Lucy, focus group).
This discourse of success and the importance of academic achievement is akin to that reproduced in other elite school spaces, with failure disavowed and externalised from the school (Allan and Charles, 2014). Marks lower than a B belonged to a disavowed 'other' (Skeggs, 2005;Tyler, 2013). Lucy and other girls clearly differentiated between their expectations and those of young people in other schools.
The difference between Manor School girls and other young people was a common theme in many discussions, and the school was set apart as distinctive from other schools in the area: "Like some schools are really different to our school … because some schools are, they do different things and have different, and think different things … they're like so different schools." (Saabira, focus group).
It is interesting to note that it is the other schools which are seen as different in this quotation, rather than the grammar school, although of course it is the grammar school which is distinctive to the other, and more numerous, state high schools, in the area. Other girls started to unpick the nature of this difference, suggesting that other schools had easier, less academic, curricula, as opposed to being "all nerds" (Karolina) like the girls in Manor School: Yaso: I mean like school wise, like their subjects are really from ours, like they get more creative subjects -Jaya: like 'doss' subjects! Carolyn: They find different things important.
In the following excerpt the idea of being academically different is replaced by the notion of young people in other schools being morally different: Jameela: They don't have the same … … morals. Yana: Yeah. Jameela: Or like … Anita: They don't have the same morals? Carolyn: They're quite different, and just generally the things … Anita: Yeah the way they're brought up (focus group).
As the conversation continued, it became evident that these different morals were perceived to be tied to coming from specific areas, which were 'dodgy' or dangerous, with later discussion emphasising that this was tied to having less money.
Yana: Sometimes it's not always parents, it's like where they've been brought up and the people they're around, like constantly, like their friendship group might be different in their own school. So then when we like meet up with them, there will be like different, you can see the difference". Researcher… OK, so when you talk about where, is it in terms of the different place or …? All talk over each other: It might be the area! Jameela: Yeah the area. Clearly, some girls (re)produced negative stereotypes about poverty, conflating a lack of money with a moral failure; a dominant discourse about poverty, which has intensified in recently in line with entrenched neoliberalism and Austerity policies (Nayak and Kehily, 2014;Skeggs, 2005;Tyler, 2013).
Rather than setting other young people up as different from themselves, some girls expressed how young people from other schools identified them as different, and conflated attendance at the school with being wealthy, as the extract from a focus group below demonstrates: Anya: I remember when we were walking one day, and there was a bunch of girls and they were all coming up to us asking do you have a pony, do you have a pony, just because we were like from Manor School, and they obviously really like labelled us as these posh people, but it's like … Researcher: Sort of countryside with ponies and … Sabelle: I think it's more like rich … We're not like that at all! Patty: We're just normal.
The girls resisted this representation; however, this conflation of the girls with relative wealth demonstrates a broader awareness from outside the school that those with higher levels of economic and cultural capital are more likely to attend grammar schools. Unsurprisingly, the girls considered their class backgrounds 'normal', and this sense of normality can conceal privilege from the relatively advantaged (Bourdieu with de Saint Martin, 1996).
Nonetheless some girls presented the grammar school space as relatively wealthy and with few students who come from poorer backgrounds. As the following excerpt from the focus group with Jameela, Yana and Anita emphasises: Researcher: Ah ah, but are there any people from council houses in this school? Jameela: Probably a few but I guess …compared to other schools there's hardly any. Yana: But also the fact that they're in this school shows that they are still motivated to work hard and stuff. Anita: But where they come from doesn't affect who they are … Jameela: Much, yeah.
Anita: Yeah, it's not a big factor. Jameela: It's quite small, that percentage. (focus group).
Class difference was externalised beyond the boundary of the school and denigrated (Francombe-Webb and Silk, 2016).

Diversity as ethnic diversity, and a source of 'cosmopolitan capital'
In contrast to the repudiation of socio-economic diversity, there was a celebration of ethnic, religious and racial diversity as a social and cultural resource (Mitchell, 2003;Maxwell and Aggleton, 2013). The girls assumed that questions about diversity was about ethnic, religious or racial diversity: "Yeah, there's quite a lot of diversity, even in our friend group there is a lot of like different backgrounds and stuff like that… you feel that when you come to this school you can relate to people and like you have someone you can like, who is sort of the same as you, and people who are different, and you can really get along with those kind of people. I mean like if, because I'm Hindu and I believe in some things which other people may not believe in, but I still get along with other people" (Paavai, focus group).
This ability to get along with people who are in some ways different forged a 'cosmopolitan sensibility'an 'openness, interest and ease of engagement with the Other' (Maxwell and Aggleton, 2016b: 782). This cosmopolitan sensibility is a resource, arguably a form of cultural capital (Waters and Brooks, 2015;Weenink, 2008) for the girls; Emma discusses how she has become more open and learned about the world through meeting people from different cultural backgrounds at school: "It broadens your mind really … I went to a private school where most people there were Christian and I hadn't really had much experience with other religions and things like that. But when I camehere, because there are so many different types ofpeople, we found out about so many different hobbiespeoplehave. And it's really quite interesting….". (Emma, focus group).
These positive discussions of diversity stand in stark contrast to the perceived challenges of diversity in resource limited state schools. It would seem that middle class ethnic, religious and racial diversity is open to cosmopolitanism, whereas working class ethnic, racial and religious diversity is a multicultural challenge.
The openness to ethnic and religious diversity was tied to the lived experiences of the girls, who had ethnically and religiously diverse friendships; here 'encounters' between young people from different ethnic backgrounds led to broad acceptance of difference (Wilson, 2014;Casey, 2016). This also meant that girls from ethnic minority backgrounds felt included in the school. Here, Paavai contrasts feeling included and accepted in Manor School with the bullying she experienced at primary school: Paavai: in [my primary school] … it didn't really work out and I was bullied and … basically, I don't exactly remember but I think it was about like the way, things I eat, ate and … Researcher: Oh, about your culture? Paavai: My culture. … Yeah, and then that's why I needed a big change, it was, I was really like a bit scared going from that school to this school because I was a bit scared about …, how people were going to be, but everyone was just like me here and …Yeah and it was really easy to just make friends".
In Paavai's narrative, she emphasises that the other girls are 'just like' her, demonstrating how socio-psychic processes of recognition (Butler, 2004) or identification (Bondi, 2002) forged a sense of commonality across ethnic and religious differences (Cockayne et al., 2017). In these discussions the cosmopolitan sensibility of connections across diversity does not belong exclusively to white girls, but belongs to all the girls in the school. The sense of connection around being a bright, and nice, grammar school girl was powerful in diminishing other lines of difference between the girls.

Enacting privileged, regulated and 'nice' femininities in the elite schoolmutual identification, recognition and empathy
Overall, the girls embodied, performed and (re)produced privileged subject positions, being 'nice', 'polite', 'kind' 'civilised' and able to regulate their emotions (Gagen, 2015). Their femininities tended to be regulated into good behaviour (or carefully choreographed and limited 'naughty' behaviour) and avoided the shame of stepping outside of these well-regulated femininities (Wolfe, 2017). The girls claimed to get along with most of their peers, and behave kindly and politely, even to those whom they liked less, demonstrating empathy towards others to avoid leaving anyone out: "We always want to include everyone because like sometimes we ourselves have been in that situation and it's just like you know how it feels and we always include everyone. And like, I mean in our class everyone's happy to go and help other people…" (Padma, focus group).
Clearly nice, middle class girls make an effort to get on with others, and have empathy for other people's feelings (Bondi, 2002). Whilst reproducing a 'perfect girl' subjectivity, is experienced positively by girls who have previously experienced bullying, as Padma emphasises, being nice involves skilled emotional labour (Hoschild, 2012), the cost of which is internalised; as Aashna (interview) stated: "It's quite hard, it can be quite an emotional strain, yeah, to be nice to someone that you don't like, yeah".
Although open conflict was repudiated, there was discussion of conflictual social relationships, which were projected (Bondi, 2002) onto others. For instance, as Emma (interview) states: "Occasionally the whole class will be quiet because there are two people yelling across the classroom at each other, like really mas-sive…" These conflicts were rare, and girls never admitted to having been involved in such open conflict themselves. Nonetheless social relationships are effuse with subtle interplays of power, and young people are hierarchically positioned and position themselves and others in their social groups (Vanderbeck and Dunkley, 2004;Thomas, 2011;Kustatscher, 2017). Among these polite, civilised and empathetic girls, these power relationships were generally insidious and played out in relation to different degrees of closeness of friendships or through other subtle means.

(Re)producing power inequalities through subtly hierarchical acts
Girls identified a hierarchy of friendships from people you dislike (but for the most part are polite to) to very close and 'best' friends. Most girls identified one or two 'best' friends. Specific spaces, such as classrooms and lunchrooms provided opportunities for encounter for friendships to be forged; as Jenny exemplifies: "Well I met these two, well I started talking to these two because we were in the canteen, and so we just started talking and we've just recently got quite close because they're in some of my lessons and I just go to see them at lunchtime and break times" (Jenny, interview) Friendships tended to be forged with a subsection of girls who regularly encountered one another, often in formal spaces of classrooms (particularly 'forms'groups for pastoral care, registration and mixed ability lessons). As the girls' homes were dispersed, the principle space of the performance of friendships was the school, transport to school, or when waiting for trains. Social relationships were maintained outside school, with much effort, by meeting up for lunch, shopping, going to the cinema or each-others' homes, and via social networking sites. The time for this was constrained by the high levels of homework required to maintain academic success.
Some girls emphasised two different key types of friendships; those generated among peers with similar interests, and 'close' friends based on empathy, trust and understanding. Aashna states that she has a friendship based specifically upon a shared interest in the elite curricula of Latin and Classics, whereas Harriet discusses friendship groups based around shared tastes in music: "…because like we all have the same kind of taste in music, so like that's something we have in common and that's how like most friendships have the same taste in music" (Harriet, in a joint interview with Saabira).
There were subtle, hierarchical distinctions between girls, as Aashna (interview) sums up neatly: "Yeah, like if you're, if, for example, like in a class there's always like the joker or the popular person, so if you're like for example the rock band person…" Harriet describes herself as a joker, and is therefore probably one of the popular people in the class, compared to the quieter group of Aashna's friends who coalesce around a love of the classics, suggesting subtle hierarchies between the girls and their social groups (Kustatscher, 2017).
Another subtle hierarchy is between 'close' friendships and groups predicated on looser connections forged by shared interests, although Harriet suggests that these two types of friendships can merge: "I have two different groups, and Mel's in the one where I go and talk to them about like different things to this group, and this is like mostly the people that are in the group I talk to about like comic books and stuff! … Well I don't know, we kind of, we went to watch like The Avengers, like the Marvel movie and we've kind of gotten into it a bit more …" (Emma, focus group) Paavai, goes on to emphasise: "Yeah, a friend is someone who you can trust. It's a two-way relationship, so you do things for them and they do things for you. And yeah, as Emma said, it's someone who you can turn to when you're in need".
Since the playing out of power in friendship groups is subtle, powerrelations between the girls were difficult to trace, but ranged from conflict within close friendships, not having close friendships, exclusions and being left outparticularly as dynamic friendship groups shifted, to performing ambivalent roles in games as a precursor to recognition. In general, girls reciprocated close friendships, which were reflected in the self-selected focus groups. Even close friendships are simultaneously emotionally nurturing and conflictual, and typified by relationships which oscillate between and projection and identification (Bondi, 2002). In the intense spaces of academic high achievement, emotions and tensions can run high, even among good friends, as Tina (focus group) notes, close friendships can leave you "emotionally dragged down". Sabelle emphasises the importance of having a break from your best friends: "…[It is] nice to have breaks from them [best friends] because otherwise when you're with them too much you become agitated by them. …[when spending lots of time together] even the small stuff that they do that you normally like don't care about, suddenly becomes like World War II …" (Sabelle, interview). Some girls suggested that they did not have close friends in the school, and this was an experience discussed by some young people with Special Educational Needs and Disability in an earlier paper, showing a commonality of experience in different spaces among young people with very different characteristics (Holt et al., 2017). For instance, Laura, interview, claimed that most of her close friends were outside school: "Yeah, I don't like having, I don't like being fixed to one set of friends….I don't really have a friendship group, it's more me being able to talk to anyone I want to [and later] I don't really have a best friend but I have close friends, yeah, Jane is one of mine and I've got a lot of close friends outside of school as well, so … Because I've known them for much, much longer…" Friendships and social relationships are dynamic and shifting, and it was often in these processes of change that subtle exclusions and otherings were enacted as girls' positioned others as 'closer' or 'less close' friends, as Sabelle relates.
"At one point I have like Charlie and I have a friend called Nat, and they were both … it would be like two twos and they would both call me up and tell me what each other said about me, and it was basically girls getting really like, you know, mean and stuff … but we're friends now…" (Sabelle, interview) Mostly, the girls did not talk about their own experiences of being left out or isolated, this tended to be projected onto others, or discussed in relation to other space/times. Here, Sabelle reflected on a past experience of being left out, but emphasised that the issue was now resolved.
The research captured a specific moment of change, as the form groups were being divided up to make smaller classes for next academic year, and the girls had recently been asked which friends they wanted to be with. This had the potential to exclude those who were not chosen. In the interchange below, Karolina contests the version of the story given by Cora about her sensitivities in relation to this change: Although many friendships groups were ethnically and religiously diverse, subtle differences were played out around ethnicity and religion. For instance, Aashna and Paavai mentioned not being allowed by their parents to sleep at other people's houses, with Aashna emphasising that this is 'an Indian tradition'; although this was negotiated by going along to the sleepover but going home at bedtime. Similarly, although Paavai emphasises how included she feels in Manor School compared to her previous experience, her difference is subtly expressed as she discusses being a vegetarian in Nandos, a popular chicken restaurant chain, and identifies as preferring Pizza Hut.
Power-relations were subtly reproduced by the different roles girls took in games; this produced different subject positionings and hierarchies in the friendship groups, as Aashna discusses: "It's like this comic book thing where it's like superheroes and the bad guy is Loki… So like whenever we go and see these kind of movies we always give a character to each of our friends, so I'm Loki, apparently, for no reason, just! And one of my friends is Thor, so we keep like teasing each other, being like oh I'm better than you … Loki's better than Thor" (Aashna, interview).
The Loki referred to is a character in the Marvel Comics and in the 2012 film The Avengers, who is reimagined as a complicated 'bad' character from the ancient Norse god of mischief. There are clear racial and ethnic undertones, as Loki is the god Odin's adopted son, but a biological member of the race of enemy giants. By contrast, the unequivocal 'good' Thor is the god Odin's biological son and rightful heir (Arnold, 2011). Although subtle, we argue that it is telling that Thor is acted by Emma, a white British girl. By ascribing the role of Loki to Aashna, 'recognition' in this social group required adopting an equivocal role, which had racialised and ethnic undertones of which Aashna and the other girls do not seem to have been reflectively aware. Through these subtle performances, insidious ethnic differences were, arguably, reproduced (see also Gaztambide-Fernández et al., 2013). These differences were so subtle that they were perhaps operating beyond the direct consciousness of the girls, who were certainly not deliberately reinforcing racial and ethnic stereotypes amongst their friends. These games might give insight into a ''a domain, full of "deeper" drives…'' (Philo andParr, 2003:285, cited in Davidson andParr, 2014: 121).

Discussion and conclusion
This paper has explored the intersecting performances of classed, gendered and racialised/ethnicised, privileged subjectivities in girls' peer social relationships in a multi-ethnic, elite, selective state grammar school. Relationships between girls in the school and young people outside the school, oscillated between socio-psychic, spatialised processes of 'identification', 'empathy' (Bondi, 2002) 'othering' or 'abjection'. These connections and differentiations are underpinned by an emotional need for 'recognition' (Butler, 2004). Social relationships of 'identification' were important to a sense of belonging to the school for girls and had two key consequences; first, those outside the school spaces were 'othered' -there was an 'abjection' of the poor nongrammar school young person; second, belonging to Manor School required the performance of an appropriate, successful hardworking, polite and kind, well-regulated subjectivity, resonant of an A1 girl (McRobbie, 2009), but one in which different and hierarchically placed subjectivities were subtly wrought.
Those who did not attend the grammar school were seen to have lower academic and moral standards, and on close examination were more likely to be 'poor' than girls who attended the school. Some girls reproduced common negative stereotypes of poverty (Nayak and Kehily, 2014;Skeggs, 2005;Tyler, 2013). Class difference within the school was disavowed, and somewhat conversely, attendance at the school was seen to overcome class difference; any girl from a workingclass background who attended the school would have exceeded the perceived moral failing of her background.
The social relationships of girls represented an important way that privilege was reproduced as the girls forged a sense of belonging to the elite space of the grammar school, which in turn provided them with specific social and cultural capital. These included access to an elite curriculum and the outstanding teaching and high academic expectations which allow them to become 'A1 girls' (McRobbie, 2009). Most girls transition to top universities. In addition, a 'habitus' of 'assuredness' was reproduced for most girls though this sense of belonging and recognition to this elite space Lingard, 2013, 2015). Importantly, the high individual and group academic expectations of the girls is necessary to the continued outstanding school results, which needs to be constantly re-enacted with every cohort of girls, among whom the lowest grade that can be countenanced is a B. The school was a space of multicultural encounter, with girls from a variety of ethnic, religious and racial backgrounds (Wilson, 2014). Diversity was always assumed to be ethnic diversity and any question about diversity automatically turned to ethnic or religious questions; even with prompting, class difference was externalised or subsumed within a Manor School subjectivity. For the most part, friendships were forged across ethnic, racial and religious differentiations. This provides the potential for girls in this space to adopt progressive (Maxwell and Aggleton, 2013) and transformative performances of ethnic and racial subjectivities. Indeed, this acceptance of difference and diversity forged a cosmopolitan sensibility, which provided an embodied form of capital for the girls (Weenink, 2008).
In the school space, friendships were forged around emotional needs for recognition, trust and reciprocity, which could involve performing subjectivities which are ambivalent or problematic (Butler, 2004). All the relationships were forged by psychosocial and spatial connections and differentiations, 'identification' (you are like me) and 'empathy' (I can emotionally understand your position and relate to you) and projections or differentiationsyou or that aspect of you is not like me (Bondi, 2002). Along with providing emotional support and nurturing, friendships are fraught with power relations, given the oscillation between 'identification' and 'projection' inherent in socio-psychic-spatial relations. These power relations were subtle, since girls embodied a kind and caring feminine subjectivity, and disavowed open conflict. Girls articulated hierarchies of friendships, with the most important relationships based on 'closeness and trust', which is based around empathy and deep emotional connection, as opposed to shared interests and hobbies, tied to identification, which are seen as relatively superficial. Hierarchies were also (re)produced via exclusions and inclusions (Thomas, 2011;Kustatscher, 2017) with girls projecting exclusions onto other people or other times and spaces and disavowing such experiences for their current selves.
Friendships were dynamic, with time an important factor in both forging friendships and in the fracturing and changing of friendships. The balance between identification and projection can shift over time, as friendships become more or less close or break up. Girls talked evocatively about experiences and anxieties over these changing friendships, and expressed empathy for others who were being left out. Formal aspects of the school, and transport to and from school, were critical spaces, providing opportunities for encounter in which friendships could be forged. These formal arrangements were changing in the new academic year, leading to breaking up of larger groups and questions around allegiances, as the girls were asked to select particular friends to be kept with. This provided a critical moment (Holland and Thompson, 2009) in which friendships were brought to the fore.
Subtle framing of subjectivities occurred via the roles that girls performed in playing games, which had classed, racial/ethnic and gendered undertones, and which demonstrate that emotional recognition can involve adopting and performing equivocal and problematic subjectivities (Butler, 1997). The girls certainly forged friendships across ethnic, racial and religious differences, but sometimes ethnic or racial differentiation was subtly reproduced through these role-playing games. Previous research has demonstrated differences can be reaffirmed and more entrenched through co-educating people with ethnic and racial differences if the stereotypes which circulate in broader society are not critiqued (Thomas;. Schools are not bounded sites cut off from the rest of society and discourses that circulate elsewhere clearly pervade the school space. Subtle enactments of power arguably point to deep seated prejudices and stereotypes that might be consciously challenged by the girls, but surface in subtle ways which are hidden, potentially even from the girls themselves, pointing to a 'psychic life of power' (Butler, 1997), or "a domain, full of 'deeper' drives, passions, and repressed psychic materials returning in 'distorted' form" (Philo andParr, 2003:285, cited in Davidson andParr, 2014: 121). This suggests that socio-psychic relationships to 'difference' are complicated by broader social and cultural contexts, and by deep seated psychic processes, which might not be universal or pre-social (Bondi, 2014), but are certainly experienced as visceral, beyond conscious, and often evoke embodied responses which are difficult to straightforwardly consciously challenge. Nonetheless, attention to these deeper drives and processes emphasises that challenging preconceived ideas of 'others' is complicated.
To conclude, this paper has examined the gendered, classed and ethicised/racial and religious subjectivities forged by girls in an academically elite girls' school, which was ethnically and religiously diverse but had limited socio-economic diversity. The small-scale indepth study of the processes by which girls forge connections and differentiations between themselves and others in subtle and more stark ways, performing a distinctive classed and cosmopolitan sensibility, whilst sometimes subtly reaffirming differences between girls, provides insight into the reproduction of diverse privileged identities. This study is pertinent to the ways in which privileged and gendered subjectivities might be forged in a type of setting which is currently relatively limited within the UK context, but set to expand, and more prevalent, and increasing, throughout the globalising world. It is evident that these engaging, hard-working, lively, reflective and kind girls were, for the most part, oblivious to both the scaffolding which had supported their entry into this privileged space and the structural disadvantages of 'poor' groups who attended 'other' schools. Whereas ethnic and religious diversity was seen as a resource for the educated classes with high levels of cultural capital, class difference was equated with lack of success, low moral fibre and individual failings. This is problematic as these A1 girls, who are likely to gain future positions of power, are likely to have little motivation to challenge prevailing neoliberal attitudes towards class difference or poverty.