Chosen and collective friendship: Negotiating contradictory social ideals and demands at an Israeli elementary school classroom

Abstract

Friendship has been predominantly conceptualised as a highly positive and voluntary relationship. This article contributes to recent sociological challenges to these notions by ethnographically examining how conflicting friendship ideals are negotiated in everyday life. It is based on a year-long study of friendship socialisation and friendship between girls in an Israeli elementary school classroom, from the perspectives of both the girls and their teachers. I argue that the teachers promoted two contradictory friendship ideals: one of ‘chosen’ friendship between specific students, which recognised the children’s agency and preferences; and the other, of ‘collective’ friendship between all the class group members, directed at engendering social cohesion and preventing loneliness. The article delineates how the girls and teachers negotiated relationships between the students under the framework of both friendship ideals – and in doing so, exposed the tensions and entanglements between the two. Moreover, the girls and teachers’ friendship discourses and practices shed light on the hefty social demands placed on children’s friendship ties in school, and how friendship can incorporate both collective and individual meanings.

Introduction

As we walked together to the physical education lesson, Koral happily told me: ‘Everyone in class is my friend: Avigail, Nofar, Dana, Hadas, Meital, Gaya. . .’. (Fieldnote, March 2016) ¹

Drawing from an ethnographic study of friendship socialisation and friendship between girls at an elementary school, this article explores how two contradictory friendship ideals were negotiated by the girls and their teachers in everyday life. The first ideal was based on a ‘collective’ understanding of friendship, as a relationship that (ideally) should exist between all the members of a group. This ideal is evident in the first part of the quote above, with Koral declaring that all the 23 children in her fourth-grade class at Herzl Elementary School were her friends. The second ideal is based on the dominant conceptions of friendship as a voluntary undertaking. This ideal of ‘chosen’ friendship appears in the latter part of the opening quote, with Koral naming only a few of her class peers as friends. Taken together, Koral’s words reflect the implicit contradiction that exists between the two ideals and her attempt to weave them together. In this article, I argue that friendship can be comprised of both individual and collective meanings, and that it carries with it weighty demands.

Sociologists of friendship, in the main undertaking their research in the Western contexts of Europe and North America, have argued that the modern ideal of friendship conceptualises friendship as a voluntary relationship, based on choice and individual agency (Blatterer, 2013; Budgeon, 2006; Pahl & Spencer, 2004; Silver, 1989). However, scholars have also pointed to the ways in which social structures and norms relating to social categories, such as gender, class, ethnicity and sexuality, constrain friendship formation (Blatterer, 2013; Cronin, 2014; Jamieson, 1998). Relatedly, critical sociological perspectives of friendship, exploring the gaps between the ideals and the experiences of friendship, have shown that supposedly voluntary friendship bonds can also be perceived as binding and difficult (Eramian & Mallory, 2020; Heaphy & Davies, 2012; Smart et al., 2012). Furthermore, Danny Kaplan (2006, 2018) has argued that contemporary Western ideologies of friendship are shaped by national and collective processes, and that friendship-making is integral to nation formation. This article advances sociological challenges posed to the dominant conceptualisation of friendship as highly positive, individualistic and chosen, by ethnographically exploring what I term collective friendship – an ideal that demands that all the ascribed members of a specific group be friends with each other – and how it was cultivated, enacted and negotiated alongside a voluntary ideal of friendship.

I begin by arguing that through their friendship socialisation, which encouraged practices of sharing, supporting and complimenting friends, teachers promoted both an ideal of chosen friendship and an ideal of collective friendship. They undertook the first ideal by nurturing friendship ties between specific students, an approach that acknowledged the students’ agency and friendship preferences. They undertook the second ideal by encouraging friendship ties between all the students in the class, so as to cultivate the social cohesion of the class group and prevent loneliness among students. The specific collective ideal of friendship in my case study reflected and reproduced particular Israeli socio-cultural understandings of sociality and friendship, partially stemming from the communal utopia of socialist Zionism (Kaplan, 2006; Katriel, 1991), and relating to the intricate entanglements of collectivist and individualistic elements in contemporary Israeli culture (Birenbaum-Carmeli, 2001). Nevertheless, I argue that this collective ideal also sheds light on collectivist framings of friendship in the West, which are generally under-researched (Kaplan, 2006).

Second, I show how the girls of the class group, and their teachers, had to carefully weave together and negotiate these two ideals of friendship when managing social interactions and relationships in everyday school life. In order to contend with the weighty demands stemming from the confluence of two conflicting friendship ideals, the girls creatively appropriated (Ahn, 2011; Corsaro, 2018) friendship practices and discourses promoted by the school to endow their friendships with both exclusive and inclusive meanings, in accordance with the two ideals. Meanwhile, the teachers sought to balance, on an ongoing basis, the two friendship ideals, deciding when and how to intervene in students’ relationships, given their wish to cultivate myriad values and achieve various pedagogic aims. I argue that the girls and teachers’ negotiations exposed the contradictions, tensions and intricate entanglements of the chosen and collective friendship ideals. Thus, a more nuanced understanding of friendship – as a relationship with both collective and individual meanings, and acknowledging the demands, labours and negotiations that this relationship entails – is made possible.

This article demonstrates the importance of examining friendship demands in a group setting, the role of friendships in forming collectivities, and the negotiation of friendship expectations in relation to the ideal of friendship as chosen. It thus opens up fruitful areas of enquiry for scholars studying organisations and groups where friendship ties and discourses play a central role, such as feminist movements (Friedman, 1995; Lugones, 1995; Roseneil, 2006) and workplaces (Cederholm & Åkerström, 2016; Pettinger, 2005; Wilkinson, 2019).

Theorising friendship: Chosen, constrained and collective

The body of sociological research on friendship, which has gradually evolved over the last three decades, generally explores the social and cultural understandings of this relationship form in the Western contexts of Europe and North America. The modern ideal of friendship is of a voluntary, private and enjoyable relationship, based on choice, individual agency, intimacy, support, mutuality and self-disclosure (Blatterer, 2013; Budgeon, 2006; Jamieson, 1998; Pahl & Spencer, 2004; Silver, 1989). Friendship has therefore been celebrated as a model for the ‘pure relationship’: it is symmetrical, based only on the rewards it provides, and (even though it requires commitment) can be ended at any time (Giddens, 1991).

While many sociologists continue to place emphasis on the chosen and relatively uninstitutionalised nature of friendship, they have also shown that friendship is shaped by cultural, structural and local contexts, and that its formation is constrained by structures and norms relating to various social categories, including gender, class, ethnicity and sexuality (Adams & Allan, 1991; Blatterer, 2013; Cronin, 2014; Jamieson, 1998). Recently, sociologists of friendship have critiqued the idealisation and overly positive framing of friendship in sociological research, calling for closer attention to be paid to the gaps between the ideals and expectations of friendship on the one hand, and their lived realities in the West on the other (Heaphy & Davies, 2012; Jamieson, 1998; Pahl & Spencer, 2010). Consequentially, research studies have begun to explore difficult aspects of friendship ties, which may involve power relations and tensions as well as painful and ambivalent experiences. These studies have shown that despite the voluntary ideal of friendship, in practice friendships are not always easy to end: the relationship is perceived, at times, as binding, its termination potentially damaging to the sense of self (Eramian & Mallory, 2020; Heaphy & Davies, 2012; Smart et al., 2012). This article attends to the gaps between friendship ideals and realities and augments scholarly attempts to rethink the supposedly chosen nature of friendship by examining a case study where a collective ideal of friendship and a voluntary ideal of friendship were cultivated, enacted and negotiated in everyday life.

Collective conceptions of friendship, challenging the individualistic conception of the modern friendship ideal, have so far remained under-researched. Danny Kaplan (2006) argued that contemporary ideologies of friendship are simultaneously shaped by processes of individualism and nationalism. Kaplan delineated how formations of nationalism have been associated with organisations like fraternities and secret societies, which promote notions of civil solidarity and brotherhood. He thus called for consideration of collective framings of friendship, as informed by national ideology. Kaplan (2018) further argued that friendship is ‘the social glue of state and nation’ (p. 7), as citizens are expected to cooperate and become friends with the strangers whom they encounter in the different social institutions they belong to, such as school. In turn, these experiences engender the imagination of society as a nation community bound by solidarity.

A collectivist conception of friendship is also implicit in Junehui Ahn’s (2010, 2011) ethnographic study of a middle-class American pre-school, in which teachers laboured to convince the children that everyone in the class was their friend, socialising them to the importance of group life. Ahn (2010) argued that this socialisation conceptualised the self as relational rather than individualistic. Notions and ties of friendship also facilitated the formation of collective bonds in college sororities (Handler, 1995) and the Freemasons (Kaplan, 2018). Furthermore, anthropological studies of friendship undertaken outside of a Western context point to understandings and forms of friendship that incorporate both voluntary and obligatory aspects, simultaneously endowed with both collective and individual significance (Desai, 2010; Obeid, 2010; Santos, 2010). Since its formation, Israeli culture has possessed strong collectivist elements, which in the early and mid-twentieth century were influenced by the Jewish tradition, Zionism and socialism. Over time, however, Israeli culture has also been shaped more by individualism, resulting in an amalgamation of these two ideologies (Birenbaum-Carmeli, 2001). This article brings to the fore a collective understanding of friendship, one often overlooked in sociology in favour of placing emphasis on choice and individualistic framing. I will show how a collective ideal of friendship was employed by the teachers to encourage the social cohesion (gibush) of the class group – a principal pedagogic goal of the Israeli educational system (Erdreich & Golden, 2017; Katriel, 1991). The collective friendship ideal reflected and reproduced Zionist nationalistic understandings of sociality (Kaplan, 2006; Katriel, 1991) and potentially contributed to militarised socialisation. It was therefore comprised of both inclusionary and exclusionary dimensions.

The study

This article is based on an ethnographic study of friendship socialisation and friendship between girls in an Israeli public elementary school classroom, conducted primarily through participant observation. The ethnographic approach can facilitate a nuanced exploration of children’s understandings and constructions of their peer relations, by striving to suspend adult conceptions of childhood and gaining access to children’s own cultures (Corsaro, 2018; Duits, 2008; Emond, 2005; Fine & Sandstrom, 1988). My fieldwork consisted of observations of the school life of the same class group, for between one and one-and-a-half days each week between March 2016 and February 2017. When I commenced my fieldwork, the students, aged 9–10, had passed the mid-point of the fourth grade; by its conclusion, they has passed the mid-point of the fifth grade. Most of the students were Mizrachi or Russian Jews of low socio-economic status. As I spent almost my entire time in school observing, interacting and speaking with the students (in groups and with many girls also on a one-on-one basis), I conducted in-depth semi-structured interviews with the fourth-grade and fifth-grade class teachers, and with three other teachers who taught the class in fifth grade.

While there were both girls and boys in the class group, my decision to study the friendship between girls alone stemmed from the theoretical framework of girlhood studies. This interdisciplinary field emerged in the 1990s, following critique by feminist scholars of the marginalisation of girls in society, as well as in youth studies and women’s studies. Focusing solely on girls, they argued, would shed light on issues and experiences unique to them (Kearney, 2009; Lipkin, 2009). This focus proved apposite during my fieldwork, as most of the close friendships in the class did not cross gender boundaries. Even so, this study is limited in that it does not explore the friendships of the boys, and how these were negotiated in relation to the friendship discourse and demands of their teachers – which, as far as I could observe, were directed at all students.

Informed consent was obtained from the school’s principal and the fourth-grade class teacher before commencement of the study; consent was obtained from the other teachers when I first met them. Because research studies were frequently conducted at the school, the class teacher decided that the children’s parents only needed to be notified that a research study was in progress, and to be given the choice of opting-out. (None did.) The consent of the children, as participants, was negotiated as an ongoing process (Vincent et al., 2018; Zhu, 2022). Throughout the fieldwork, I explained to the children what I was doing and I wrote short fieldnotes in their presence, to keep the research visible and give them opportunities to refuse to participate (Duits, 2008; Thorne, 1993). There were only a few instances in which the children purposefully kept their distance from me, or asked me to withdraw so that I would not be privy to private conversations. This was an encouraging indication that they felt comfortable enough to withhold their consent; and that in general, their participation in the study was active and consensual.

In my fieldwork, I adopted the role of an ‘adult friend’, maintaining non-authoritative and positive relationships with the children, distancing myself from the typical child–adult school dynamic and utilising the difference between adults and children to enquire about their everyday life and perspectives (Emond, 2005; Fine & Sandstrom, 1988). Embodying the role of ‘adult friend’ enabled me to experience first-hand some of the practices and negotiations of friendship between the girls: for example, in choosing a play partner, and exchanging small items. However, at times I also had to take on the role of ‘responsible adult’, which was expected of me not just by the school principal and teachers, as gatekeepers of the field (Wyness, 2012), but also by the girls at times; when they were distressed, for example, or when serious altercations occurred between the children (see also Emond, 2005; Thorne, 1993). I thus had to navigate my two roles in the field carefully and continuously.

Whilst I did record the interviews conducted with the teachers, I did not record my observations and conversations in school, instead documenting them in my fieldnotes at the time they occurred or shortly afterwards. Thus while the quotes from my fieldnotes presented in this article are not always word-precise, they do reflect the style and content of what was said accurately. After concluding fieldwork, I manually analysed the data by reading and re-reading my fieldnotes and interview transcripts; identifying key themes and concepts arising both from the field and the theoretical framework; coding and categorising my data accordingly; and interpreting the data (Rossman & Rallis, 2017).

Friendship socialisation at school

Tamar, the class teacher, explained the upcoming maths test to her class. She then added that the students were being tested ‘all the time: in your behaviour, in helping and saying nice things to friends (haverim), in making an effort’. Tamar proudly noted that ‘some of the children here always help their friends and are being transformed’, and reminded the students that this was their test. Later, a line began to form next to Tamar’s desk as several of the students approached her for help with the maths assignment they were working on individually. To save all of them time, the class teacher suggested that the students turn to friends for help, reminding them that ‘this is the test’ that she had mentioned earlier. A few students proudly declared who they were helping. (Fieldnote, October 2016)

Apart from times when friendship problems negatively affect the school environment, children’s friendship ties and skills receive relatively little pedagogic attention from teachers (Carter & Nutbrown, 2016; Healy, 2011). In contrast, friendship socialisation was a key feature of the pedagogy at Herzl School. The school’s friendship discourse permeated everyday school life, teachers constantly encouraging ‘friendly behaviour’ (hitnahagut haverit) between the students: specifically, sharing school supplies, snacks, emotions, food and personal experiences; helping each other with schoolwork; offering emotional support in times of need; and saying nice things to each another. The teachers promoted friendship ties and practices between the students through academic means. For instance, throughout the fifth grade, the class teacher reminded the students that she would award additional grade points to children who displayed friendly behaviour, and to those who fulfilled the various tasks she assigned, such as sending her pictures of themselves meeting up with friends from the class group during the school holidays. Furthermore, at several points during the school year, the students were asked to fill out a self-assessment and goal-setting form. The form had an entire section dedicated to friendship, which included prompts like ‘the qualities I have that help me get along with friends’, and ‘things I need to improve in myself to deal with friends’.

The centrality of friendship socialisation at Herzl School also came up in my interviews with the class’s teachers, who described how teaching the students social and friendship skills counted amongst the most important aspects of their jobs. The two class teachers, who were intimately aware of the intricate social dynamics between the students in the class group, encouraged the formation of friendship ties between specific students, and often intervened to settle conflicts and arguments. They did so both publicly during class time and in private conversations with the students.

On the one hand, the understandings and practices of friendship promoted by the teachers corresponded with some of the defining characteristics of the modern ideal of friendship: support, intimacy and self-disclosure (Jamieson, 1998; Silver, 1989). On the other hand, other important characteristics of this ideal, specifically choice and individual agency (Budgeon, 2006; Silver, 1989), were constrained due to the teachers’ management of and intervention in their students’ friendships. In their study of English primary schools, Carol Vincent and colleagues (2018) similarly found that children’s friendships were somewhat constrained due to the teachers’ friendship cultivation and regulation.

The teachers anchored their friendship socialisation in not only the dominant modern ideal of chosen friendship, but also in an ideal of collective friendship. They attempted to cultivate friendship between all the members of the class group, and encouraged them to extend voluntary friendship practices of sharing, supporting and complimenting to all the students in the class. This collective ideal of friendship is evident, for example, in the following excerpt from my fieldnotes:

Tamar, the class teacher, asked the students to open their language workbooks. On realising that Keren did not have a workbook, she instructed her to work with Sigal, sitting next to her. As the teacher began to read aloud from the workbook, an argument developed between Sigal and Keren. On noticing the conflict between the two girls, the teacher paused her reading and turned to the two girls to enquire what the issue was. Sigal complained that ‘it’s uncomfortable’ to share her workbook with Keren, because she had to place it between the two of them. Attempting to placate her, Tamar explained to Sigal: ‘It’s friendly (haveri) to share, I give some of what I have, like with a sandwich that I would give a piece of to someone who didn’t have any.’ (November 2016)

Here, Tamar, the class teacher, expected all her students to share with one another willingly: a key characteristic of the school’s friendship discourse, which she and the other teachers underscored every day. In effect, Tamar was asking Sigal and Keren to consider each other as friends, even though such a relationship did not necessarily exist between them. Similarly, when teachers wanted students to pay attention to their peers while speaking during a lesson, they implored them to listen and show respect to ‘your friends’ (haverim shelachem). This signalled to the students that they should see each other as friends. During her research in an American kindergarten, Ahn (2010, 2011) observed a similar process: teachers labouring to convince the children that all the other children in the class were their friends by referring to them as ‘friends’ and asking them to call each other thus – emphasising relatedness and consideration of the group.

As friendship is shaped by specific socio-cultural contexts (Adams & Allan, 1991; Blatterer, 2013; Cronin, 2014), I propose that the ideal of collective friendship at Herzl School was informed by, and in turn reproduced, Israeli collectivist understandings of sociality and friendship (see also Kaplan, 2006). As is common in the Israeli education system, teachers at Herzl used the term ‘class friends’ (haverim lakita) to describe the relationship between students in the class group. The Hebrew-Israeli term ‘friend’ (haver) captures a spectrum of interpersonal and supportive relationships, together with membership of the same organisation or collective domain (Kaplan, 2006). This multiplicity of meaning encapsulates the Jewish-Israeli ethos of togetherness (Rozental, 2014). The togetherness ethos is, in turn, a key characteristic of the prevailing Israeli ideal of social cohesion (gibush), along with cooperation, solidarity, joint activities and shared sentiments (Katriel, 1991).

While the term and the aims of gibush can be applied to many social and institutional settings in Israel such as the workplace and the army, the social cohesion of the school class group serves as a model of group life (Katriel, 1991). Israel’s education system has long treated the creation of a socially cohesive class group as a primary pedagogic goal (Erdreich & Golden, 2017; Katriel, 1991); both teachers and students at Herzl School expressed a desire for the class to be socially cohesive. In Israel, a school class group is perceived as a meaningful social unit with collective characteristics: its makeup remains the same across each educational stage (elementary, middle and high school), and it usually has the same class teacher for a few years in a row (Katriel, 1991). In elementary school, the class teacher is highly involved in managing the relationships between the students of their class, both in school and in some cases outside of it; for instance, some class teachers decree that all the students in a class should be invited to each other’s birthday parties (Yariv, 2018). A central way for teachers to manage relationships, and to ensure harmony and order in the class setting, is to encourage social cohesion within the class group by organising enjoyable group learning and social activities for the students (Yariv, 2018).

The Israeli ideal of social cohesion has historical and ideological context which should be taken into consideration. The notion of gibush stems from the communal utopia of socialist Zionism (Katriel, 1991), an ideology supported by Jewish people who came to Palestine from Central and Eastern Europe in the second and third wave of immigration (occurring between 1904 and 1914 and between 1918 and 1923, respectively). Many of these immigrants arrived without their parents and constructed meaningful peer relationships as an alternative to family life (Kaplan, 2006). The communal ideology and notion of male comradeship intensified over time with the growth of a national ethos of military power (Kaplan, 2006). Indeed, the high levels of social cohesion in units of the Israeli Defence Forces (IDF) have been credited to the Israeli school system: the class group typically remains together throughout school, and students are socialised towards developing strong friendships and a sense of commitment to each other (Gal, 1986, in Katriel, 1991). While these specific ideological meanings and aims were not made explicit in Herzl School, the students and staff did often use the social cohesion discourse, therefore potentially contributing to the militarised socialisation and culture that are prevalent in Israeli schools both formally and, as in this case, informally (Hoffman, 2016; Levy & Sasson-Levy, 2008). ²

Social cohesion cannot be forced, only encouraged indirectly (Katriel, 1991). As personal relationships constitute a ‘key form of social cohesion’ (Jamieson, 1998, p. 2), I argue that one of the main ways in which teachers at Herzl sought to foster the social cohesion of the class group was through the ideal of collective friendship. Teachers at Herzl perceived both friendships and gibush as means of fortifying their students’ social and emotional wellbeing; the friendship practices they encouraged, such as sharing and supporting one another, corresponded with or generated the characteristics of social cohesion: cooperation, solidarity and a sense of togetherness (Katriel, 1991). In our interview, I asked Tamar what the expression ‘class friends’ means to her. She replied:

For example, if we go on a trip and I tell my class group that we’re splitting them up it moves me that they don’t want to. [They say] ‘We want to be together. No, Tamar, leave us, we don’t want to be with [the other class groups].’ This place of, yes, we’re socially cohesive, we’re together, we’re class friends. . . The togetherness is power, once you split us then we become weak. That is the gibush, the togetherness, it’s people who know me. (August 2017)

Tamar’s story also reveals that the more inclusive ideal of collective friendship still entailed exclusionary aspects. Gibush is characterised by homogeneity of group members and clear boundaries between them and group outsiders (Katriel, 1991) and the Jewish ethos of togetherness has, at times, lead to separatism and xenophobia (Rozental, 2014). Students of the class group resisted teachers’ attempts to cultivate the social cohesiveness of the entire year group and often forcefully banned other students from entering their classroom.

Because membership of the class group was an ascribed status, and because the notion of gibush emphasises solidarity and sameness over acknowledgement of individuality and diversity (Erdreich & Golden, 2017), the collective ideal of friendship stood in stark contrast to the voluntary ideal of friendship – the latter as an achieved relationship underlining choice, individuality and agency (Blatterer, 2013; Budgeon, 2006; Pahl & Spencer, 2004; Silver, 1989). Despite the contradiction between the two friendship ideals, both were present in the school’s friendship discourse, and were thus entangled within everyday life at school. This was exemplified through the teachers’ use of the term ‘class friends’. For instance, when teachers encouraged students to share with their ‘class friends’, it was often unclear whether the teachers meant that the students should share with those whom they considered their friends, or whether they should share with everyone in the class, as all the students were habitually defined as ‘friends’. Similarly, the boundary between being ‘friendly’ (haveri) towards a class peer and being their ‘friend’ (haver) was unclear. I argue that these ambiguities were purposeful, as they enabled teachers at Herzl to cultivate both specific friendships and collective friendship at the same time. In the next two sections, I discuss how the girls and their teachers experienced the entanglements of the chosen and collective ideals of friendship, and how they negotiated the conflicting demands stemming from these.

Girls’ negotiations of friendship ideals, demands and preferences

In their everyday school life, the girls had to negotiate and balance the weighty and competing demands created by the two contradictory friendship ideals that the teachers promoted. In doing so, they exposed the tensions between these ideals, and the gaps between the ideals and the lived realities of friendship (Heaphy & Davies, 2012; Jamieson, 1998; Pahl & Spencer, 2010). Many of the girls fulfilled the ideal of collective friendship by having friendship expectations of many of their peers, for instance that they share their snacks and toys with them, complimenting them and expressing a desire for their class to be socially cohesive. At the same time, all the girls fulfilled the ideal of chosen friendship by having their own friendship preferences, and by demarcating boundaries, which they used to signal who they were close to over others. The tension between the girls’ attempts to perform both ideals was exemplified when one of the girls, Gaya, described the annual class trip to me as one where ‘we are socially cohesive (megubashim) and make up lots of songs’, before quickly adding that ‘there are also arguments, the girls fight about who they will spend time with during the trip, [because] everyone wants to be with me and with Nofar’ (Fieldnote, December 2016). This quote also reveals the gendered limitations of the collective friendship ideal, as most of the girls were only interested in being friends with other girls, and not with the boys of their class. I will now examine some of the ways in which the girls negotiated the conflicting friendship ideals and the hefty expectations that grew out of these.

In some of their playtime activities, the girls attempted to weave both friendship ideals together. During play, children negotiate their friendships through processes of inclusion, exclusion and boundary-making (Ahn, 2011; Corsaro, 2018; Thorne, 1993). In Katarina Gustafson’s (2009) study of school playground spaces, most of the children from a single class group constructed an identity of ‘we-in-our-class’ and a notion of solidarity through a shared game while two girls who only spent time with each other constructed an identity of ‘the-two-of-us-as-best-friends’. In contrast, the girls at Herzl School attempted to simultaneously construct both friendship ideals. For example, in fourth grade, the entire class often enjoyed playing hide-and-seek together during playtime, thus fulfilling the collective friendship ideal. Nevertheless, the girls usually chose to hide together with their close friends, so as to enjoy each other’s company more privately and thus fulfil the chosen friendship ideal. The challenge of weaving the two ideals together in practice before various audiences, and specifically of enacting the ideal of collective friendship, was also apparent during these games. At times, the seeker declared the round to be over because they believed that everyone had been found – whilst in fact, some children were still hiding. A few girls confided that they felt hurt when this happened to them, as they had effectively been forgotten by the other students.

Furthermore, the girls laboured both to fulfil the inclusive meaning of collective friendship and to creatively appropriate the school’s friendship discourse (Ahn, 2011; Corsaro, 2018) in demarcating more exclusive chosen friendship boundaries. As mentioned earlier, the teachers constantly encouraged ‘class friends’ to share with one another – not just material objects, but also emotions and personal experiences. A staple of the social and emotional education at Herzl School was a weekly ‘listening circle’ in each class group, directed by the class teacher and centred on a different question or prompt each time, such as ‘What scares me?’ and ‘Pick a friend from the class (haver mehakita) and say a quality they have that you would also like to have’. The listening circle was often emotionally charged, with many children voluntarily sharing intimate details about their lives or crying openly in front of their peers. Collective friendship was thus enacted through engendering intimacy between all students, intimacy being a key characteristic of the modern ideal of friendship (Jamieson, 1998; Silver, 1989). In contrast, friendly relationships that are formed in order to get along with one another, for instance at work and the neighbourhood, are conceptualised as rather superficial, lacking in intimacy and self-disclosure (Kurth, 1970; Wilkinson, 2019). Nevertheless, although the girls openly shared intimate experiences in the listening circle, some information – for example, who they had a crush on – was only divulged to their close friends. Importantly, friends often performed intimacy with one another in front of other students, by whispering to each other or making up secret languages. This served to both affirm their chosen friendships and exclude others (Corsaro, 2018; Zhu, 2022). In doing so, the girls endowed emotional sharing with a more exclusionary meaning.

Although the girls demarcated their close friendships from their relationships with other students in the class group, at times they also perceived these latter relationships as collective friendship. After Avigail got into a heated argument with her close friend Koral during a shared game, she told me that she does not want to be her friend anymore. A week later, when I enquired about their relationship, she informed me: ‘I’ve decided not to be angry (brogez) with anyone. I will be Koral’s friend (havera) at school but not after school’ (Fieldnotes, March and April 2016). Koral’s statement distinguishes between collective and chosen friendship ideals, which the teachers tended to blur together, yet also demonstrates that she still conceptualised her relationship with Avigail as friendship. As I have shown, cultivating collective friendship with others in their class group provided the girls with many play partners; enabled them to make sharing demands (though they were not always keen to share when asked themselves); made them feel comfortable to disclose intimate details about themselves; and generated a sense of social cohesiveness. Additionally, fulfilling the ideal of collective friendship helped the girls get along with others in the intense environment of school life and to be perceived by both teachers and students as a good friend and student. The various meanings that the girls attached to the collective friendship ideal meant that they mostly oscillated between considering and treating those who weren’t their close friends in the class group in a shallower friendly manner (Kurth, 1970; Wilkinson, 2019) and more significantly as (collective) friends. Some girls seemed more prone to one approach than another, yet it was not always easy to discern between the two in practice.

During fieldwork, I myself experienced some of the tensions born out of the two contradictory friendship ideals. In my role as ‘adult friend’, I sought to create friend-like relationships with all the girls. However, I soon discovered that I had more in common with some of the girls, and tended to prefer their company to that of others. Nevertheless, I did not feel that I could display my preferences – on ethical grounds, and because I was expected to fulfil the role of ‘responsible adult’, one who embodied the school’s ethos, which included the collective friendship ideal. Like the girls, I had to carefully manage my social interactions and ties under the framework of the school’s friendship discourse.

Teachers’ negotiations of friendship ideals, aims and values

While the teachers laboured to promote both collective and chosen ideals of friendship in their pedagogy, they too, like the girls in the class, had to carefully weave the two ideals together and negotiate the tensions that arose from their contradictions. As with the girls’ negotiations, I propose that in doing so, the teachers exposed the gaps between the ideals and realities of friendship (Heaphy & Davies, 2012; Jamieson, 1998; Pahl & Spencer, 2010).

In one of her science lessons, Tamar, the class teacher, explained to the students that they would be undertaking, in groups, a creative activity. She asked them to choose partners who they ‘enjoy being with’. When asked by a student whether they could work by themselves, the class teacher replied: ‘I’d rather not, I would be sad if a child is by themselves.’ As Tamar provided detailed instructions for the activity, the students talked among themselves hectically, securing their work partners. Olga turned to Marina and asked: ‘Wanna be with me?’ Marina agreed. When two other girls asked if they could join them, Olga replied: ‘No, Marina and I are working alone.’

After concluding her explanation, Tamar announced: ‘I want to see that everyone has someone to work with.’ She then turned to each student and asked who they would like to work with, and encouraged other students to join the groups that had already formed. When the teacher asked Marina who she would like to work with, Marina named Olga. Tamar asked if someone else would like to join them; Olga quickly protested, because she did not want others working with them. Tamar asked Sigal if she would like to join them, but both Sigal and Olga objected to the teacher’s suggestion. Tamar tried to convince them, but was unsuccessful. Later, Gaya suggested that Sigal join her group, to which Sigal agreed. Tamar praised Gaya for her suggestion. (Fieldnote, January 2017)

This scenario played out, with minor variations, throughout my fieldwork. It demonstrates how the teachers attempted to promote both friendship ideals and weave them together. Letting students choose their work partners corresponded with the ideal of chosen friendship, while the ideal of collective friendship encouraged students to agree to work with many of their peers. However, the teachers’ interventions in the makeup of the workgroups, to ensure that no one was left to work alone, placed constraints on the ideal of chosen friendship, and implied that collective friendship did not necessarily exist between all the students in the class. Furthermore, while the teachers did not direct their friendship discourse differently at girls and boys and referred to all students as ‘class friends’, they did recognise that the girls were mostly friends with each other, as was the case with the boys (a fact they attributed to the children’s age). Consequently, they generally did not attempt to create mixed-gender working pairs and small groups. In the excerpt above from my fieldnotes, Tamar’s attempt to encourage Olga and Marina (very close friends) to work together with Sigal was an endeavour to frame all three girls as friends who enjoyed each other’s company according to the collective friendship ideal. At the same time, her eventual concession recognised the girls’ own friendship preferences according to the chosen friendship ideal. She then praised Gaya, a popular girl with two other close friends in the class, for inviting Sigal to join her group, as Gaya was seen as fulfilling the collective ideal of friendship.

The class teachers were committed in their pedagogy to cultivating collective friendship, a relationship that went beyond friendliness, because they saw it as promoting social cohesiveness, relatedness and the importance of group life (Ahn, 2010). Hagit, the fourth-grade class teacher, encouraged a group of students who spent time together outside of school to invite everyone from the class. In our interview, she told me how pleased she was to hear about occasions when the entire class group met up after school. Nevertheless, the teachers also acknowledged privately that collective friendship and social cohesiveness did not always occur, and were not always possible in the class group. This was because the collective friendship ideal stood in contrast to the significant pedagogic emphasis placed by teachers on recognising the children’s agency, and on addressing their specific social and emotional needs and interests. These pedagogic goals, in turn, were reflected and reproduced through the promotion of the chosen friendship ideal, which is anchored in individual agency (Silver, 1989).

The tensions between the underlying values and aims of the conflicting friendship ideals came up in my interview with Tamar, when she reflected on the relationships in the class group:

No one is a complete outsider, which is very good in my opinion. That was really important to me. I think the negative is that the friendships (havruyot), the friendship groups (havurot) are more formulated. It’s good and it’s not good. You can’t separate Nofar, Gaya and Hadas, it’s forever. . . It’s very, very clear who are the good-good friends.

. . . So when I split them into workgroups, it’s obvious to me that I won’t separate them, because it will cause a mess. And also, it comes from [a place of] – yes, it’s fun to be with your friends. I don’t always need to fight to make, I don’t know, Keren and Nofar socially cohesive (legabesh). There is no connection between them. I don’t need to force it now. Keren has it good with Reut, and Nofar is more comfortable [with Gaya and Hadas]. But it is nice, for example, when we play a game and I sometimes ask them – try to talk to someone else [in the class group]. (August 2017)

Here, Tamar acknowledges the existence of a gap between the ideal of collective friendship that she promoted and the reality of social relations in the class. She displayed ambivalence about enforcing this ideal, as she recognised the girls’ agency in creating and maintaining their friendship ties (Ahn, 2011; Corsaro, 2018; Wyness, 2012). But she also acknowledged the benefits of encouraging the girls to define many or all their class peers as their (collective) friends. Similarly, in Vincent and colleagues’ (2018) study of English primary schools, teachers recognised the importance of children’s friendships for their wellbeing, and did not significantly limit their freedom to form friendships; nevertheless, they did encourage connections between multiple children, through practices such as seating arrangements and constantly switching learning partners.

Moreover, the importance Tamar placed on having no outsiders among her students alludes to the sometimes difficult realities of friendship, which can include experiences of ambivalence and pain, power struggles and ontological damage to the self (Eramian & Mallory, 2020; Heaphy & Davies, 2012; Smart et al., 2012). The teachers at Herzl, the class teachers especially, were deeply concerned about children who were lonely and struggled to make friends and fit in socially. Thus, the teachers promoted the collective ideal of friendship as a way of protecting their students from the sometimes painful realities of the chosen friendship ideal. While the teachers had to contend with the tensions resulting from promoting contradictory friendship ideals, at times they may have also perceived them as complementary.

Conclusion

In this article, I argued that the friendship socialisation and discourse at Herzl School, which encouraged friendship practices of sharing, supporting and complimenting, promoted two contradictory friendship ideals which students and teachers had to negotiate in everyday life at school. The first ideal is chosen friendship, used to cultivate friendship ties between specific students; the second is an ideal of collective friendship, demanding all class group members become friends. By interrogating a collective conception of friendship, this article helps challenge the dominant conception of friendship as solely individualistic (Kaplan, 2006). The collective friendship ideal in this case study was based on specific Israeli understandings of friendship and social cohesion, which reflected and reproduced Jewish-Zionist ideologies, thus potentially contributing to Israel’s militarised socialisation and culture.

Collective understandings of friendship also manifest in Western, supposedly more individualistic, societies (Ahn, 2010; Kaplan, 2006). Specifically, the proposed concept of collective friendship is relevant to organisations such as the Freemasons, where the goal is to cultivate friendship among members (Kaplan, 2018), and college sororities, where friendship between women is institutionalised, made obligatory and framed as sisterhood (Handler, 1995). As in the case presented in this article, notions of close friendship inform understandings of collective ties in the above organisations and close friendships develop alongside collective ones (Handler, 1995; Kaplan, 2018). In contrast, chosen friendships are forbidden in Cistercian monasteries as they are perceived to threaten the existence of the collective ties of fraternity, which are characterised by love yet are impersonal and distant. Nevertheless, personal relationships still develop between some members (Sundberg, 2019). In school, unlike the above organisations, membership in the class group is not chosen and students must spend a large portion of their day together, which possibly means that these tensions are experienced more starkly and require more careful and continuous negotiation.

The contribution of this article is in ethnographically examining how conflicting friendship ideals are cultivated, enacted and negotiated in practice. I have explored how the girls of the class group saw one another as friends in some instances yet also creatively appropriated the inclusive meanings of ‘friendly behaviour’ to demarcate exclusive friendship boundaries. The teachers tried to balance between encouraging preferred friendships to cultivate students’ agency and promoting friendship among all students to shield them from difficult experiences of friendship and loneliness. Through their negotiations, the girls and teachers exposed the tensions between these ideals, and the gaps between the ideals and realities of friendship, yet also wove the two ideals together in everyday life at school. This article thus contributes to a more critical and complex understanding of friendship, by foregrounding the demands, constraints and negotiations that it entails. This stands in contrast to the dominant narrative that conceptualises friendship as a purely voluntary and overly positive relationship (Eramian & Mallory, 2020; Heaphy & Davies, 2012; Smart et al., 2012).

This article invites further research into how friendship ideals and ties can play a role in the formation and dynamics of groups and collectives, including processes of inclusion and exclusion. This can be particularly useful in research into political and national groups. For instance, feminist scholarship in the 20th century argued that friendship between women is vital to the success of the feminist movement (Friedman, 1995; Lugones, 1995; Roseneil, 2006). Relatedly, scholars have begun to explore the theoretical potential of political friendship for engendering solidarity among strangers and bringing citizens together (Devere & Smith, 2010; Kaplan, 2018; Mallory & Carlson, 2014). However, this article has also pointed to exclusions that can occur through the promotion of collective friendship, though it is beyond its scope to delineate the related tensions that arose in Herzl School between the notion of a homogeneous and socially cohesive class group and the culturally diverse background of its members. Moreover, the Jewish-Zionist ideologies embedded in the ideal of collective friendship mean that it will likely exclude non-Jewish people, especially Palestinians. Sociological research on the relational aspects of work, which has pointed to the importance of friendship ties and friendly relations, and the blurred boundaries between the public and private spheres (Cederholm & Åkerström, 2016; Pettinger, 2005; Wilkinson, 2019), can also benefit from an exploration of how discourses and experiences of friendship are endowed with collective meanings and aims and where and how the lines between being ‘friendly’ and being friends with colleagues are drawn and blurred.

Finally, these findings showcase the significant insights that sociological research into children’s friendships can offer to the sociology of friendship, which is mostly based on research undertaken with adults (Vincent et al., 2018). In particular, examining children’s friendship socialisation in settings such as the school and the family can reveal dominant socio-cultural understandings of friendship and their lived realities. This is vitally important, because while friendship is a significant form of relationship in many – if not most – people’s lives, there is still much to unpack about it sociologically.