Same-gender intimate friends in Chinese girls’ romantic adventures in a boarding school context

Friendship and romance are two popular topics amongst children. In children’s romantic adventures, same-gender intimate friends always play a significant role. Since romance in childhood is constructed as an inappropriate topic in the Chinese schooling context, there is a limited number of studies that can offer an in-depth understanding of Chinese children’s experiences of same-gender intimate friendships and romantic adventures at school. This article aims to close this gap by using ethnographic data collected in a rural primary boarding school in China to present how Chinese girls’ same-gender intimate friendships and their romantic adventures interweave together.


Introduction
In children's everyday personal lives, significant adults always have power over children and set up rules to regulate their behaviours in different contexts. However, with the development of New Social Studies of Childhood, children's agency to understand and give meaning to their own lives and their capabilities of negotiating with surrounding environments to gain control over their lives are increasingly appreciated (Corsaro, 2009;James, 2013). As active agents in their own lives, children are not socialised and regulated by adults passively, but tend to actively form peer cultures and challenge adults' control through practicing their agency (Corsaro, 2009). However, much work around children's agency has focused on western societies with less attention to children's vivid process of practicing agency in the Global South contexts (Punch, 2016). To contribute to closing this gap, in this article, I use Chinese girls' romantic adventures, which refer to children's practice to express and perform their interest in opposite-gender peers, in a boarding school context as an example to discuss these girls' creative ways of managing their samegender friendships and heterosexual romantic adventures in a conservative school setting in the rural area of China.
Sexuality and romance in childhood are debatable topics because they are closely linked to the existing ways of conceptualising children. As argued by Renold (2005), because of the discourse of childhood innocence, there is a debate between views of children as "sexual becomings" and as "sexual beings". However, scholars have noticed children's curiosity and excitement about sexuality (Zhu, 2021). In middle childhood, friendship and romance are two popular topics amongst children. As claimed by some scholars, same-gender intimate friendships and heterosexual romance are interwoven (e.g. Etcheverry et al., 2013;Flynn et al., 2017;Giordano et al., 2006;Mellor, 2006;Walton et al., 2002). For example, in Walton and colleagues' work with US inner-city pre-teens, they notice that same-gender intimate friends are always understood as a source of support and comfort during children's romantic adventures. However, the relationship between experiences of romance and intimate friendship is complex and could influence the quality of each other simultaneously (Flynn et al., 2017). For instance, when friends are romantically interested in the same person, their friendship might be threatened (Walton et al., 2002). Children might also feel a tension between their wish to spend time with friends and their wish to be with romantic partners (Giordano et al., 2006). As two important types of intimate relationships in children's everyday lives, experiences of friendship and romance can be emotionally charged. Some empirical studies report that children and young people can experience stress when they must cope with conflicts between friendships and romance (e.g. Giordano et al., 2006;Walton et al., 2002). In this case, an in-depth exploration of how same-gender friendships and romance are interwoven in children's peer world could contribute to the understanding of children's peer relationships in their everyday personal lives.
However, because romance in childhood is constructed as an inappropriate issue in the Chinese school setting (e.g. Farrer, 2006;Shen, 2015), there is a limited number of studies that focus on exploring the role of friendships in children's romantic adventures. Therefore, friendship and romance are rarely studied as interweaved issues in existing China-based childhood studies literature. To contribute to closing this gap, this article uses ethnographic data to indicate how Chinese girls' same-gender intimate friendships and their romantic adventures interweave together in their daily interpersonal interactions at school. Although I only draw attention to girls' experiences in this article without addressing gender differences questions, it can still open an insightful discussion about Chinese children's agentic experiences of negotiating with the surrounding environment to sometimes enforce and sometimes resist adult socialisation efforts on childhood romance.

Romantic adventures in the Chinese school setting
According to the UNCRC, 'children' is used to refer to human beings below the age of 18 years. However, 'children' is not a homogeneous concept because childhood is a social construction and children's experiences in their childhoods are highly diverse according to different environments in which they live (James, 2013). When comparing constructions of children and childhoods across the world, similarities and differences coexist because of some shared constructions, such as innocent children, and unique understandings of children, such as about their agency, which vary in different sociocultural contexts. For example, because of China's sociocultural, historical, and political norms, such as Confucian-collective moral codes, a series of norms emerge about being 'normal' and 'morally good' children. To be 'normal' and 'morally good', Chinese children are expected to be 'in place' without crossing the adult-child boundaries and challenging adults' authority directly (Zhu, 2020(Zhu, , 2021. Therefore, children's agency, to certain extent, might not be appreciated by adults but be negatively constructed as disobedience. Consequently, Chinese children tend to have limited space to practice their agency, especially when about issues that were constructed as inappropriate to children, such as sexuality in childhood. Although, like other Western countries, there is an increasing recognition of the importance and benefits of talking openly about sexuality with children in China society, topics about sex and sexuality still remain 'taboo' to some extent because of the discourses of protecting children's innocence and Chinese sociocultural moral norms (Liu and Su, 2014). As summarised by Chen and colleagues (2009), culturally constructed ideas of 'expected timing' and 'appropriate characteristics' of romantic relationships cause a significant impact on Chinese people's mainstream attitudes toward childhood romance. In Chinese society, romantic relationships are viewed as an 'adult only' practice; consequently, Chinese children are viewed as the ones who are not ready to engage in such relationships (Shen, 2015;Zhu, 2021). In addition, Liu et al. (2020) and Bakken (2000) also indicate that, following Chinese traditional sexual and moral norms, there is a culturally constructed need for emotional expressiveness and self-control (zikong) in the process of dealing with self's curiosity and excitement about sexuality and oppositegender relationships. Therefore, children, who actively perform their desirability of romantic relationships, such as courtship, dating, hugging, and kissing, with less selfcontrol, normally are subject to adults' discipline, facing significant pressure from teachers and parents (Chen et al., 2009;Liu et al., 2020); their behaviours of engaging romantic relationships are viewed as inappropriate and dangerous 'early love' (zaolian), as a threat to their futures (Shen, 2015).
Apart from the influencing discourse of children's innocence and above discussed culturally constructed ideas of 'expected timing' and 'appropriate characteristics' of romantic relationships, Chinese adults' conservative and vigilant attitudes toward childhood romance can be attributed to their concerns about children's futures from both pragmatic and moral perspectives. For example, because of China's examination-oriented education and evaluation system (Liu, 2006), studying hard to achieve good academic performance is constructed as children's 'only one' mission at school and as their obligation to their families (Dello-Iacovo, 2009;Huang and Gove, 2015). Thus, 'early love' is concerned as a risky behaviour that costs children time and energy, which are supposed to be put into study and exam preparation, and consequently undermines children's opportunities to achieve a better future through performing outstandingly in exams. In addition, Chinese civilised sexual morality, such as traditional gender expectation of female chastity (Farrer, 2006;Shen, 2015), underpins the construction of 'early love' as dangerous, especially for girls, because of its potential of causing premarital sexual behaviours (Zhu, 2021). Therefore, Chinese teachers and parents always aim to reduce 'inappropriate' romantic adventures among children, which means children are normally criticised and even punished sometimes when their romantic adventures are noticed (Farrer, 2006). For example, a child who engages in 'early love' could also be asked by their teachers to have a serious meeting in the teachers' office about the negative outcomes of 'early love'; children's parents are involved in such meetings in some cases (Zhu, 2021). Children, who engage in 'early love', could also be labelled as 'bad' children, and face a risk of being excluded by peers (Liu et al., 2020); such exclusion could be caused by adults' intervention in children's choices of friends through requesting their children to befriend 'good' children and stay away from 'bad' ones (Zhu, 2020).
However, children are not passive recipients of adults' demands but are social actors with the agency to creatively and actively negotiate with surrounding environments to construct everyday personal lives and relationships with others (James, 2013). Thus, agentic children do not completely follow adult-sanctioned rules about gender proximity. In school-based observations, it is commonly reported that Chinese children would challenge adults' authority and create strategies to talk about and engage in romantic adventures (Farrer, 2006). In addition, some Chinese adults' attitudes toward "liking" between boys and girls are changing in recent years because of a reconstruction of the meanings of romance in childhood and recognition of children's agency and adults' losing the power of controlling children's access to sexual knowledge via 'pornified' media in the Internet era (Liu et al., 2020;Moore and Reynolds, 2018;Mulholland, 2015). Therefore, some Chinese teachers increasingly reflect on their exclusive reliance on their traditional ways of forbidding and punishing children to control 'early love' and recognise the importance of communicating openly with children about sexuality and oppositegender relationships (Evans, 2007). However, because of the current coexistence of traditional and new thoughts about children's curiosity and excitement about sexuality and opposite-gender relationships, as discussed in-depth elsewhere (Zhu, 2021), ambiguous and complex feelings could be noticed in Chinese children and adults' narratives about sexuality and romance in childhood; for example, "liking' (xihuan, feelings of interest toward someone) between boys and girls is constructed as both "normal"/"acceptable" and "abnormal"/"unacceptable". As a result of these coexisting thoughts, children might notice teachers' complex and ambiguous attitudes toward "liking" between boys and girls and different reactions toward their practice of 'early love'. Such complexity and ambiguity could make children feel more curious, excited, confused, and anxious about childhood romance, and might blur the 'forbidden' lines drawn around issues about 'early love' then create some 'grey area', in which children might feel more exciting to engage romantic adventure to test their agency and challenge adults' authority.
As suggested by Western-based empirical studies, in children's romantic adventures, peer friends, especially intimate friends, could play significant roles (Walton et al., 2002). Because of a high level of intimacy, the exchange of personal confidence, emotional support, and common interests and companionship are often involved in intimate friendships (Jamieson, 2005;Pahl and Spencer, 2004). Then, between intimate friends, they can 'feel free to be sincere, spontaneous, and open about themselves' (…) feelings, preferences, and life facts' (Sharabany et al., 1981: p. 800). Romance is one highly private topic that children often discuss and explore with intimate same-gender friends. For example, as Mellor (2006) noticed in his ethnographic fieldwork, romance, including practices and gossip, is one of the most existing topics between children at school. Walton and colleagues (2002) also suggest that same-gender friendships can significantly shape children's romantic adventures. Unfortunately, as an underdeveloped topic in the Chinese context, there is limited information to show the vivid picture of same-gender friends' influence on Chinese children's romantic adventures. However, the role of peer friendship in Chinese children's romantic adventures might be more important compared with some other societies because of Chinese adults' generally conservative attitudes toward children's romantic adventures (Liu et al., 2020). Therefore, peers could be the best choice to share their curiosity, excitement, and confusion about romance. The significant role of peers, in this case, can be even more prominent in boarding school settings. Being a residential child means that most of the time is spent with peers far away from families; therefore, in a boarding school setting, peers are the ones, who are always being there, companying children to explore their everyday lives. Therefore, in this article, I introduce ethnographic data to draw a picture of same-gender intimate friends' roles in Chinese girls' romantic adventures in a rural primary boarding school setting to unpack the vivid process through which these girls negotiate with the surrounding school environment to enforce some adult-sanctioned values and resist some adult restrictions on romantic adventures.

Methods
Data used in this article was collected in a 5-months intensive ethnographic fieldwork with 49 Primary Year five children (aged from 11 to 13 years old, 25 boys and 24 girls) in a rural primary boarding school (given the pseudonym "Central Primary School") in Hubei Province, China. Unlike boarding schools in western countries, boarding schools in rural areas of China are the solution for children without sufficient daily family support (e.g. children with migrant parents) and children who need to commute from surrounding villages. Since Central Primary School serviced surrounding nine villages, over 80 percent of its students were residential students. These residential students spent all their time at school from Sunday night to Friday lunchtime, and were picked up by their guardians on Friday afternoons for weekends. This research setting was chosen after considering my capacity of speaking the local Chinese dialect, relationships with local gatekeepers, and ethical agreements with the school authority team and teachers. This project gained ethical approval from the University of Edinburgh and used Gallagher's (2009) ethics toolkits to adopt ethical principles in fieldwork.
Although this broader project focused on exploring children's peer friendships at school, "liking" between boys and girls and peers' romantic adventures were very popular topics amongst children emerged in both conversations and observations. Being inspired by these popular topics, I was interested in collecting more information about children's understandings of liking opposite-gender peers and experiences of romantic adventures. However, like in other Chinese schools, "liking" between boys and girls was treated as a highly sensitive topic among children (Zhu, 2021). Therefore, various strategies have been applied to overcome dilemmas experienced in both ethics and data collection as will be discussed.
To immerse myself in children's school life, I lived in the teacher's accommodation during the entire fieldwork to participate and observe their full-day school routines. During everyday observation, moments when teachers were absent, such as class breaks, meal breaks, evening self-study time, and the period between the end of self-study time and sleep time, were the 'golden' time to observe vivid peer interactions. To triangulate and add additional information to data collected through participant observation, 35 of my 49 child participants were interviewed privately by me in school's cosy consulting room (20 girls and 15 boys, 30-40 min semi-structured interviews, recorded by an audiorecorder) about their understandings and experiences of peer relationships at school, 36 of 49 child participants also joined a diary exchange programme (23 girls and 13 boys). The diary exchange programme was conducted for around 3 months by providing each child participant with a notebook to record any thoughts and questions about peer relationships at school and hand them in to me whenever they wanted; then, I returned notebooks to them with comments after their records to build up private dialogues. This alternative approach offered children, especially those who didn't feel comfortable having face-toface conversations with me, opportunities to use a writing-based approach to communicate with me at their own pace.
The combination of observation, interview, and diary programme was very helpful in terms of collecting and triangulating data. For example, through hearing a small number of boys' and many girls' stories about romantic experiences in private conversations, I found out about "where" and "when" to direct my focus to observing children's romantic adventures. Based on an educational ethnography of children in primary school, Renold (2005: 33) highlights time and space as important elements when locating sexuality in school settings. Similarly, children in Central Primary School were keen to locate their romantic adventures in inconspicuous times and spaces. Thus, at the beginning of the fieldwork, in public spaces, I rarely observed intimate interactions between boys and girls. In most cases, I learned about such interaction (e.g. sending gifts and love letters) from the children in our private conversations. From children, I noticed that although different children chose different means of expressing their affection to the ones they liked, these actions always took place during the same time and in a similar place. The specific time was after evening self-study time and before the children arrived in their dormitory buildings; the specific space was the path from the classrooms to the dormitory buildings, including the hallways, the entrance to the teaching building, and the playground. As explained by children, they chose these particular times and spaces because of two reasons. Firstly, children felt less stressed away from the gaze of watchful adults during these times because, in the evening, there were only four on-duty teachers and two wardens on the entire campus to supervise around two hundred children. Secondly, the path from the classrooms to the dormitory building is dimly lit, which made them feel that their interactions could go unwitnessed in the relative darkness. After gaining this information from the children, I updated my observation plan and successfully observed several P5 children's romantic involvements by myself.
In this project, my identities have significantly shaped the data collection. During the data collection process, my "unusual adult" (Christensen, 2004) role significantly contributed to my rapport with children then supported me to gain access to sensitive topics, such as "liking" between boys and girls and romantic adventures. To distinguish myself from teachers to play "unusual adult" role, I introduced myself as an "older sister", who loved to play, chat, and befriend them to learn from them about children's friendships. Also, through ongoing interactions, I was able to pass some trust tests, which proved that my presence would not put them under surveillance, and I did not 'betray' them by disclosing our secret chats to their teachers and parents. Apart from the trust test, the 'openness' test was another common test that I experienced in the field, especially when topics between children and me were constructed as 'sensitive' in the schooling context. To pass the 'openness' test to open free discussions with children about their romantic adventures, I particularly used the self-disclosure strategy of "exchanging secrets" as mentioned by Mellor (2006) in his ethnographic studies about children's playground romance at school. Like Mellor's experiences, my child participants also showed a high level of curiosity about my romance, and I was open to answering the enormous number of children's questions about my romantic adventures at school. Although sometimes I felt a bit stressed, annoyed, and uncomfortable when children jeered at my stories, this experience of "exchanging secrets" contributed to the richness of the research data (e.g. more in-depth conversations with children about childhood romance and even closer observations of their relevant practices, such as cross-gender interactions) by building up rapport and emotional resonance between us.
However, my identity as an "older sister" also caused certain limitations to this research. Being a female researcher with memories of my girlhood, my relationships with girls were closer than that with boys (see also Thorne, 1993). Consequently, compared with boys, girls were not only more prominent in my observation notes but also showed a stronger interest in participating in other research activities. Moreover, since discussing self and peers' romantic adventures was risky and "inappropriate" in public places at school (Zhu, 2021), dormitory rooms were a significant space for children's conversations about romance. As a female researcher, it was inappropriate to access to boys' dormitory rooms; but on a few Sunday afternoons, I was able to visit the girls' dormitory rooms with their invitation for around 30 min to 1 h each time to chat and play with them, with the result that girls are even more prominent than boys in my data about "liking" and children's romantic adventures. Therefore, in this article, I use data collected both in private conversations with girls and in observations, to explain how the girls negotiate space, time, and same-gender intimate friendships to increase interactions with the opposite-gender peers whom they liked and to reject unwanted pursuers. This project's ethnographic data was analysed by using thematic analysis. Through close reading, coding, and analytic writing (Emerson et al., 2011), I was able to generate themes and topics to organise the findings displayed in this article. Before moving on to a detailed discussion, it is necessary to clarify the 'missing' data about lesbian, gay, bisexual, transgender, and queer (LGBTQ) groups in this article. In this article, I use a binary girl-boy division and discuss heterosexual romance in childhood. LGBTQ seemed to be a difficultly-accessed topic at school, which might be because the gender binary of male-female and heterosexual gender relations are still dominant social discourses in school settings and the attitude toward LGBTQ in the Chinese society are generally conservative (UNFPA, 2018). Furthermore, I did not collect much information about LGBTQ groups in my fieldwork in both observations and conversations with participants. In conversations with children and significant adults (e.g. teachers) about children's romantic adventures at school, such romantic adventures were tacitly referred to as heterosexual courtship, which might be an example to suggest that children's certain understandings of romance are aligned with the heterosexual norms established by the adults.

Involvement of same-gender intimate friends in romantic adventures
In everyday lives, children's behaviours suggest ambivalent attitudes toward adult socialisation efforts. On the one hand, children resist adult education and restrictions on their behaviours, such as rules of befriending peers (Walton and Davidson, 2017), especially choosing 'good' friends (Zhu, 2020). On the other hand, children enforce some adultsanctioned values, such as drawing a fine line to distinguish "acceptable liking" and "unacceptable liking" between boys and girls (Zhu, 2021). Such ambivalent attitudes were also noticed in girls' narratives about their suitors. For example, on 15 April 2016 in a P5 girls' dormitory room, I observed an episode that showed how girls resisted adult restrictions on romantic adventures while enforcing some adult-sanctioned values when judging a boy's popularity among girls: When Cai showed the girls a necklace that was left on her desk by Ouyang as a gift, the other girls laughed and passed it around to have a look. Jieyu held up the necklace to me and complained about Ouyang's pursuit of Cai: 'I am confused. How does Ouyang have the confidence to pursue Cai? Cai is a student leader and has a good academic performance, but Ouyang, he is very bad, he is a rogue. He does not match Cai. Cai deserves a better boy. We [Cai's intimate female friends] all tell Cai to stay away from him.' When Jieyu said this, the other girls all nodded their heads and added words and phrases, such as 'yes' and 'such a daydream', to support Jieyu's argument. Cai listened and laughed with the girls. Then she said: 'Yes, you [Cai looked at Jieyu] help me to give this back to the rogue. Who wants to have his gift?' (Field note, 15 April 2016) As Walton and colleagues (2002) highlight, same-gender intimate friends play an important role in listening, supporting, and comforting children in their romantic adventures. These friends' opinions (e.g. approving or against) are always taken seriously (Etcheverry et al., 2013). In this example, same-gender intimate friends were the ones who served as the "candidate selection panel", with whom Cai chose to share romantic experiences and seek suggestions. As suggested in this example, when judging romantic objects, "matching" in heterosexual romantic relationships was commonly discussed amongst girls. In this example, Ouyang's failure to "match" Cai encouraged the girls' agreement on rejection. Among girls, there were many different criteria referred to when evaluating whether a boy and a girl are "matching". Most of the criteria commonly used by most girls were in line with the adult-sanctioned pragmatic characteristics of a "good" child in the context of China, such as psychologically and mentally healthy, academically competent, prosocial, and moral (Xu, 2017). Among these criteria, academic performance was a prominent one, which was commonly mentioned by both boys and girls when describing desirable opposite-gender peers. This strong relationship between a child's academic performance and their popularity amongst peers might be different from some western-based findings (Allen, 2013;Houtte, 2004) because of China's examinationoriented educational culture, which is more significant in the rural area of China (Dello-Iacovo, 2009).
However, this does not mean girls only value such pragmatic characteristics educated by adults when they are judging peers' attraction. Other factors, such as good appearance and temperament were also frequently mentioned by girls in their narratives of desirable boys. However, compared with some studies with girls from western societies (e.g. Mulholland, 2015), Chinese girls seemed uncomfortable about making sexual assessments (e.g. 'the guys are hot') and rarely used words with sexual connotations to describe desirable opposite-gender peers. These Chinese girls commonly claimed the behaviours of making sexual assessments and using words with sexual connotations as inappropriate, which could risk their reputation as "good" girls. In contrast with girls, as noticed by Renold (2005) as well, it was not uncommon to notice that some boys used "sexy" and "hot" to describe the girls who attracted them and engaged in "sex talk" with their male friends. Such gender differences might provide further evidence of these Chinese girls' internalisation of female-focused moral expectations in China (Evans, 2007).
In the above example, by calling Ouyang a "rogue" and asking Jieyu to help her to return the gift to him, Cai was able to show her alignment with her girl friends' attitude toward Ouyang. Once a child and her same-gender intimate friends concluded that a particular suitor was not the "matching" one, these friends played the role of gatekeepers to keep the unwanted suitor away. For example, after the previous conversation, I kept following these girls' interactions with Ouyang. According to the girls, they had tried several approaches to keep Ouyang away from Cai. They threatened Ouyang, saying that if he did not stop pursuing Cai, they would report him to the teachers. However, as observed by Walton and colleagues (2002: 685), many girls report that some boys will persist in their pursuit even when they have been warned not to. Similarly, in the Ouyang-Cai case, the girls' warning did not work. Therefore, as I noticed in observations, these girls then tried an alternative approach of using the power of 'the public (peer) gaze' (Renold, 2005: 33) to 'protect' Cai from Ouyang in the way described below. On a couple of occasions, in observations after evening self-study time, Ouyang was seen waiting in the hallway outside Cai's classroom, seeking opportunities to talk with Cai or to give her some gifts. Interestingly, in such cases, there were always a few Cai's classmates who saw Ouyang outside and returned to the classroom to alert Cai. That done, Cai always left the classroom with her intimate female friends surrounding her. Whenever Ouyang approached Cai, her female friends would try to stop him and would start to shout warnings such as 'Ouyang is being a rogue!' or 'Help!' Their loud and sharp voices always attracted many children's attention and, very quickly, other children came to witness the disturbance. Ouyang would normally give up when the surrounding peers started to tease and jeer at him. Then, these girls would quickly help Cai to get away from him.
This strategy was successful because it exposed Ouyang's action to a wide audience and thwarted the inconspicuousness of his pursuit. In follow-up chats with Ouyang, he complained that because Cai's friends clustered around her, he had no chance to develop a one-to-one private connection with her. Moreover, he was annoyed by Cai's friends' action of screaming to attract other children as witnesses, because he not only felt embarrassed when surrounding peers teased and jeered at him but also feared that the loud teasing noises might attract teachers' attention. Such strategy provides clear evidence of these girls' capabilities of understanding and using the peer culture of 'teasing' crossgender interactions heterosexually (Renold, 2005) and adults' restrictions on children's romantic adventures (Shen, 2015). Although these P5 children have always been troubled by the annoying heterosexual teasing produced by witnessing peers (Zhu, 2021), they simultaneously learned the "power" of such teasing and witnessing from such experiences. In this case, these girls creatively and effectively used this "power" to increase the social distance between girls and boys, thus marking and policing gender boundaries to protect their friends from unwanted suitors.
Apart from protecting friends from unwanted suitors, in the field, it was also common to observe girls' interactions with boys whom they liked taking place simultaneously with other female intimate friends' companionship. Although the path between the classrooms and the dormitory building after the evening self-study time could provide the children with a relatively inconspicuous environment for interactions with desirable oppositegender peers, it was not enough for some of them. Therefore, in the daytime and in public areas, when some girls wanted to create opportunities to spend time with the boys whom they liked, joining group games that involved these boys was a frequently observed strategy. When using this strategy, same-gender intimate friends' companionship was an indispensable condition to not only protect these girls from heterosexual teasing but also improve the chance of joining in games successfully. The most popular games played by most P5 boys during my fieldwork were "Run for Time" and "Nametag Ripping Battle", learned from the popular variety TV shows "Run for Time" and "Running Man". Both were competitive games that required participants to hide and seek, and run and tussle, until the targeted participants were caught, or their nametags were ripped from their upper backs. Because these games included chasing and frequent body contact, girls commonly complained that asking and being accepted to play these games with boys was not easy because of the risk of heterosexual teasing. To cope with this risk, girls came up with the strategy of asking to join the boys' games with a group of other girls rather than as individuals. Jieyu, a P5 girl, explained their rationale as below: If you go alone, other people will be gossipy. They will say you came for someone. It will also be difficult for them to decide, because they might want to allow you to join, but they are also afraid of the gossip. But if you go with other girls, the more the better, people will say nothing, because there are so many people (Field note, 17 May 2016). Thorne (1993: 69) points out that, in some cases, gender terms could override individual identities. According to Jieyu, when interacted with boys while in the company of other female friends, their gender identities would override their individual identities. Hence, when an individual girl went alone to ask for the boys' permission to join in the boys' games, she only represented herself; while if she went to ask for permission along with her female friends, she represented not only herself but also her gender group. Thus, when an individual girl asked to join in boys' games, the boys' decisions were made based on their attitude towards this individual girl. But when the girls went as a group, the question faced by the boys became whether to allow girls, rather than an individual girl, to join their game. In this case, the company of other girls could increase an individual girl's chances of being accepted to join in boys' games. Moreover, joining boys' games in companionship with other girls could conceal an individual girl's desire for a certain boy. Since interactions between boys and girls are frequent and messy in group games ("Run for Time" and "Nametag Ripping Battle"), interactions between an individual girl and the boy she liked would be less prominent and were equipped with a "good reason" (e.g. 'we are just playing the group game like everyone else', as defended by some girls in observations). As a result, girls would worry less about being teased and criticised for their cross-gender interactions. In sum, although this strategy might not succeed every time according to my observations and girls' narratives (e.g. failed when some boys engaged in the games "really hated" someone in the girl group), it does suggest another creative strategy by these girls, namely using the idea of "gender group" to camouflage their individual identities and motivations in cross-gender interactions.
Although same-gender friendships often play positive roles in girls' romantic adventures, it was not rare to hear girls' stressful complaints about the conflict between their friendship and romantic adventures. In their complaints, three commonly shared conflicts troubled most of these children: friends also liked the ones they liked; friends hated the ones they liked; friends cared more about romantic relationships than about friendship. These conflicts could also be noticed in other Western scholars' studies about children's friendship and romance (e.g. Giordano et al., 2006;Walton et al., 2002). However, in Central Primary School, such conflicts between friendship and romance might be even more emotionally charged because of long school time away from family support and "unapproachable" support from adults, who have a conservative attitude toward children's romantic adventures.
Because same-gender intimate friends tend to be deeply involved in children's romantic experiences, the situation might sometimes end up as a romantic contest between same-gender intimate friends. For example, there were a couple of cases of girls and their intimate female friends simultaneously or successively liking the same boys. Because romantic contests between friends could be a threat to friendship (Walton et al., 2002: 679), two girls emotionally disclosed how upset and disappointed they were when they recognised that their friends were interested in the same boys as they were. However, among these P5 girls, it was very rare to observe or hear them fighting over boys. The most common result when two friends, especially intimate friends, liked the same boy, was that one girl "gave up" the romance to maintain the friendship. Conversations with these girls suggested a common belief among girls that, as stated by Duan, a P5 girl, 'it is shameful to have conflicts with close friends because of a boy.
[…] I am afraid to break the friendship' (Interview, 18 May 2016). These girls' feeling that conflict with close friends over boys was shameful might suggest a gendered subjectivity in children's attitudes to romance. Such gendered subjectivity was also noticed in some western-based studies, for example, girls in Walton et al.'s study claimed that 'boys, but not girls, could fight for a desirable romantic object ' (2002: 684). The fear of breaking the friendship not only shows these girls' strong determination to maintain friendships when dealing with conflicts within friendship (MacEvoy and Asher, 2012) but also indicates their choice of placing friendship with same-gender intimate friends above heterosexual romance.
In keeping with Duan's remark, it seems that, for both girls and boys, there is a friendshipromance "rule": when friendship and romance conflict, they are expected to place friendship first (or at least achieve a reasonable balance); otherwise, there might be a risk of being criticised by friends. In private conversations, both boys and girls complained emotionally that they needed to make a stressful choice between friendship and romance when the same-gender intimate friends "hated" the ones they liked. They claimed that they did not want to be criticised as "zhongse qingyou" by their friends. "Zhongse qingyou" means putting romantic relationships above friendship. This phrase is traditionally used as a stigmatised-manhood phrase to shame men who were controlled by their desire for women and thus were 'unreliable and untrustworthy' in male friendships (Zheng, 2008: 454). Although it is no longer a male-focused phrase, but a phrase used by both boys and girls in observations and narratives to complain about their friends' failure to guarantee them a similar level of attention and company after they had developed a crush, it is still a criticism that stigmatises. When using "zhongse qingyou" to criticise friends, both boys and girls described their feelings of jealousy, upset, and neglect when their friends paid more attention to romance than to friendship. For example, since bringing along same-gender intimate friends as a companionship group was a creative strategy used by some girls to increase the opportunities to play games with boys they liked, a couple of girls, like Xiaoyue, a P5 girl, complained that they had less time to play girls' games as they had to accompany their friends to play with the boys (Field note, 22 April 2016). They added that when they were fed up with these boys' games and refused their friends' requests for companionship, their friends' complaints made them feel that they were being used as 'tools for boys rather than close friends' (as Xiaoyue put it). Xiaoyue's complaint suggests that since most intimate friends always behaved as the expected source of support, there might be a risk of some children increasingly taking for granted friends' support in their romantic adventures, ignoring their friends' sacrifice of things like playtime. Therefore, a failure to balance the attention paid to friendship and romance respectively could be a source of tension between friends.

Conclusion
This article uses Chinese girls' romantic adventures at school to provide evidence that children's friendship and romance are interwoven and simultaneously shape each other. In the school's teasing and discouraging context, girls creatively developed many strategies to engage in romantic adventures. In these P5 girls' romantic adventures, support and companionship provided by their same-gender intimate friends played significant roles. Female intimate friends not only were involved in the process of judging suitors but also worked as gatekeepers to protect girls from unwanted suitors. In addition, when girls wanted to participate in boys' games that involved the ones they liked, companionship offered by the female friends could contribute to their romantic adventures by decreasing the risk of being teased heterosexually and increasing the possibility of being accepted by boys to join the game. However, the close connection between friendship and romance could also present certain challenges, which could bring children stress when they had to choose sides. When handling such conflicts, a fear of being culturally criticised as zhongse qingyou, maintaining the friendship, and restoring harmony between friends was prioritised by many children.
Through this article, children's agency, and their sophisticated approaches to negotiating with the surrounding school environment to engage in romantic adventures are prominent. Children clearly understand surrounding sociocultural norms and others' attitudes toward childhood romantic adventures, such as the peer culture of teasing crossgender interactions heterosexually and adults' restrictions on 'early love'. Then, to live their school life, children sometimes enforce and sometimes resist cultural values and norms endorsed by adults. Such coexistence of enforcement and resistance adds to the understanding of children's agency in an adult-dominated society. This article argues that children's agency not only can be suggested by their resistance to the rules, which are set up by adults to regulate children's behaviours, but also emerge in the processes of enforcing some adult-sanctioned rules and 'borrowing' adults' power to police interpersonal boundaries to manage interpersonal relationships sophistically to live their own life, such as policing gender boundaries for their own good at school.
Besides adding insight into agentic Chinese girls' interweaved friendship and romantic adventure at school, this article further endorses the methodological implication that rapport and the researcher's "unusual adult" role matter for studies of some socially constructed sensitive topics, such as gender and sexuality, in childhood. In the field, thanks to my "unusual adult" role as an "older sister", I gained a glimpse of Chinese girls' creative strategies in their romantic adventures at school. However, my gender identity also caused limitations to data collection. Therefore, further studies with boys are needed to contribute to a more comprehensive understanding of Chinese children's understandings and practices of romantic adventures at school.

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

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the University of Edinburgh and Chinese Scholarship Council Scholarship (CSC No. 201508060129).