Sexy selves: Girls,selfies and the performance of intersectional identities

Introduction

Over the last years, girls’ ‘sexy selfies’ have become highly politicised. Both selfies that are shared in private sexting interactions and selfies that are shared in more public spaces, such as profile pictures, are often met with disapproval. Popular and academic discussions have focused on risks such as bullying, harassment, blackmailing and sexual violence (for overviews of these studies, see Karaian and Van Meyl, 2015; Salter et al., 2013); on the psychological problems these pictures might cause or indicate, such as narcissism and body dysmorphia (for overviews, see Burns, 2015; Senft and Baym, 2015); and on how the images contribute to girls’ and women’s presumed ‘sexualisation’, objectification and commodification (for an overview, see Tiidenberg, 2018). Feminist scholars have pointed out how these discourses have resulted in moralising responses, aimed at preventing especially teen girls from making and sharing sexy selfies (e.g. Burns, 2015; Hasinoff, 2015; Renold and Ringrose, 2011; Ringrose, 2016).

While moralising discourses discourage girls from performing sexiness, ‘post-feminist’ discourses (McRobbie, 2009) call upon girls to use their presumed sexual freedom to pursue sexual pleasure, making it normative for girls to produce themselves as desirable heterosexy subjects (Evans et al., 2010; Gill, 2007, 2009). This call is reflected and reinforced through (social) media, which facilitate the mass distribution of heterosexy ‘dreamgirl’ imagery (Dobson, 2011, 2015) and contribute to pressures around displaying heterosexy bodies (Burns, 2015; Ringrose and Eriksson Barajas, 2011; Ringrose and Harvey, 2015). Girls are thus confronted with contradictory norms: while they are encouraged to perform heterosexiness, they risk moral condemnation and slut shaming when they do so (Ringrose et al., 2013).

Very few studies have involved teen girls’ own reflections on sexy selfies, or sexiness in general (Lamb et al., 2016). Studies that did include teen girls’ voices revealed that girls navigate sexiness in ways that both challenge and reproduce contemporary discourses (Duits, 2008; Duits and Van Zoonen, 2011; Jackson and Vares, 2011, 2015; Lamb and Plocha, 2015; Lamb et al., 2016; Ringrose, 2008, 2011; Ringrose and Harvey, 2015; Ringrose et al., 2013). These studies demonstrate that ‘doing sexy’ is not ‘just’ a matter of sexual seduction, narcissist vanity, insecurity, or self-objectification as presumed in dominant discourses, but also part of the performance of subjectivity, or, more specifically: of gender and sexuality. Through their navigations of sexiness, girls position themselves as ‘good girls’ (Jackson and Vares, 2011) and perform ‘desirable but not too slutty’ femininity (Ringrose, 2011). Some authors pointed out that girls’ performances of gender and sexuality are also influenced by girls’ positions in terms of class (Duits and Van Zoonen, 2011; Jackson and Vares, 2011; Ringrose, 2008; Ringrose et al., 2013), ethnicity/race (Lamb and Plocha, 2015; Lamb et al., 2016; Ringrose et al., 2013) and religion (Duits and Van Zoonen, 2011).

With this article, which is based on one and a half years of ethnographic fieldwork among Dutch young people aged 12–18, I contribute to and extend this field of research by exploring how through their navigations of sexy selfies, girls perform not only gender and sexuality, but also other axes of social differentiation. I investigate the different ways in which girls navigate sexiness and sexy selfies, which axes of social difference are made (ir)relevant in these navigations, and how this contributes to the performance of intersectional identities.

Sexiness and the performance of intersectional identities

In research about young people’s online practices and the performance of identity, two approaches can be discerned. One of these regards online practices as representational acts. These studies investigate how people represent themselves online, using concepts such as self-(re)presentation and impression management (e.g. boyd, 2008; Holloway and Valentine, 2003; Papacharissi, 2009; Stern, 2008). Identity is assumed to be something a person has, develops, constructs or performs offline, and that can be represented more or less honestly online.

This approach has received critique for oversimplifying and underestimating online practices. Instead, scholars suggested conceptualising online practices as performative acts that are not mere ‘representations’ or ‘biographies’ of offline subjects, but an integrated part of all performative acts that produce the subject (e.g. Bailey et al., 2013; Cover, 2012; De Ridder and Van Bauwel, 2013; Ringrose, 2011; Van Doorn, 2010; Warfield, 2016). Many of these researchers are inspired by Butler’s work on performativity (1990, 1993), which argues that identities are not an expression of some stable, inner core, but instead a ‘repeated stylization of the body, a set of repeated acts within a highly rigid regulatory frame that congeal over time to produce the appearance of substance, of a natural sort of being’ (1990: 33). Identity is therefore conceptualised by Butler not in terms of ‘being’, but in terms of ‘doing’: as a continuous repetition of acts. The present article is situated in this strand of research, and conceptualises young people’s online and offline practices related to ‘sexy selfies’ as ‘performative acts of identity which constitute the user’ (Cover, 2012: 178).

Empirical studies of (adult) selfie sharing practices have demonstrated that such an approach is fruitful in exploring the dynamics of identity performance, for instance the production of gendered and heteronormative subject positions (Bailey et al., 2013; De Ridder and van Bauwel, 2013; Dobson, 2011, 2015; Ringrose, 2010, 2011; Tiidenberg, 2015, 2018; Warfield, 2016). For example, Tiidenberg (2018) found that women’s reflexive selfie practices may carve out subject positions that fit in between those of ‘self-objectification’ and ‘joyless rejection of … sexiness’: by sharing sexy selfies and reflecting on what they liked about those selfies, her research participants produced selves that felt comfortable in their skin. Studies like these demonstrate that Butler’s theory of performativity is highly relevant to the study of selfies, selves, and the ways in which these are interrelated.

Contemporary western culture compels the articulation of a self that is recognisable, and in line with available categorisations and discourses of selfhood (Butler, 1990; Cover, 2012). Contemporary discourses of sexiness are not only gendered and heteronormative, as discussed in the introduction, but also racialised and classed. Race seems to work in contradictory ways. Girls that are framed as being in need of ‘protection’ against ‘sexualisation’ are generally white girls (Egan, 2013; Mulholland, 2017; Renold and Ringrose, 2011). Girls of colour, if taken into account at all, are conceptualised differently. In some narratives, they are conceptualised as incapable of performing sexiness due to their assumed ‘oppression’ (Mulholland, 2017: 601–603). On the other hand, girls and women of colour are also presented as (hyper)sexual, and while sexy white women are presented as victims of sexualisation or as empowered, active subjects, sexy women of colour are more easily labelled as ‘out-of-control’ or as ‘sluts’ (Lamb and Plocha, 2015; Lamb et al., 2016; Wekker, 2016), or interpreted as exotic, passive objects (Gill, 2009: 150; Wekker, 2016: 32). These analyses indicate that discourses about sexiness carry specific, sometimes contradictory, racialised connotations.

Moreover, discourses about sexiness are classed, with middle-class sexiness being defined against the ‘sexual puritanism’ associated with the bourgeoisie and, simultaneously, against the ‘looseness’ or ‘sluttishness’ associated with lower class sexuality (Gill, 2009: 150). Taking this argument one step further, one can read fears about girls becoming ‘too sexy’ as fears about corruption of the middle class (Egan, 2013: 7–8), that reproduce once more the ‘othering’ of working-class sexuality (Renold and Ringrose, 2011: 391).

In this article, I analyse whether and how girls refer to gender, sexuality, class and ethnicity in their navigations of sexy selfies, but also whether there are other categorisations that matter to them. This is inspired by intersectional thinking (Crenshaw, 1993), which argues that ‘gender cannot and should not be studied in isolation from race, ethnicity, class, sexuality, religion or other structures of power because they do not exist in isolation from one another, but instead always intersect’ (Smiet, 2017: 19).

The framework of intersectionality has been criticised for its static conceptualisation of categories. According to Krebbekx et al. (2016: 3), the framework ‘assumes categories as given, knowable and stable’ and ‘claims to know which categories matter, and who belongs to them’. Instead, these authors propose to understand differences as ‘always in the making’. Indeed, this dynamic use of the intersectional framework has been advocated by multiple gender studies scholars, who demonstrated the importance of understanding ‘when, how and under which circumstances specific intersections emerge and become salient’ (Davis and Zarkov, 2017). Such an understanding of intersectionality calls for contextualised analyses of the (un)making of different intersecting identities.

In further unravelling the relation between the (un)making of different categories, Moser (2006) argues that the enactment of one difference may support and reinforce the enactment of other differences, but may also contradict, challenge or undo them. Positions, identities and differences are thus not given, stable and singular, but emerge in the coming together (‘interference’) of different ordering processes (Moser, 2006: 543–544). This dynamic and complex understanding of identity is central to the present article.

Research methods

This article is based on one and a half years of ethnographic fieldwork between 2013 and 2014 among Dutch young people aged 12–18, in which I combined qualitative and quantitative research methods. In this article, I only use the qualitative part of the study, consisting of one year of online and offline participant observation, focus group meetings and interviews. Most offline participant observation took place in schools (mainly in two schools: one offering secondary vocational training ¹ and one preparing for vocational college and academic learning ² ) and on public transport, for instance on the bus between the train station and the school. For the online participant observation, I established online connections with participants whom I had met offline, following them into the online spaces they used. This participant observation enabled me to take young people’s daily lives as a starting point for my analysis.

In addition to the participant observation, I conducted 28 individual and duo-interviews, two group interviews and six focus group meetings. Two focus group meetings and one group interview were conducted together with MA students. ³ These discussions and interviews allowed me to ask research participants more about their experiences and their motivations, feelings and opinions regarding sexuality and social media. While the focus group discussions and group interviews were particularly helpful in constructing an overview of which social media were used for what kinds of practices, as well as for investigating how young people discuss issues related to sexuality and social media with each other, the individual and duo-interviews allowed for a more in-depth exploration of research participants’ personal experiences. Participants who were involved in this qualitative part of the research project were diverse with regard to their identifications in terms of gender, age, educational level, ethnic background, sexual identification and religion. They have been made anonymous to protect their privacy.

Drawing together different types of data enabled me to attend to the material as well as the discursive aspects of selfie making practices, both of which play a crucial role in producing ‘the self(ie)’ (Warfield, 2016). In this article, I focus on the discursive aspect. Girls’ reflections and discussions are analysed as acts rather than facts however: as rhetorical efforts contributing to the performance of subjectivity (Duits, 2008; Roodsaz, 2015). The data were analysed using a combination of deductive and inductive coding. Themes that were derived from the available literature were for instance making/sharing/commenting on selfies, performance of identity, gender and slut-stigma. Additional themes that were identified concerned young people’s definitions of sexiness, their different ways of navigating sexiness (embracing and/or rejecting sexiness), and the multiple axes of social difference that play a role, including the undertheorised categories of smartness, maturity and popularity. The combination of deductive and inductive coding facilitated the identification of both familiar and new themes in my data, and enabled me to develop a critical, detailed and youth-centred analysis of dominant conceptions of sexiness.

Defining sexy selfies

Central to this article are ‘(sexy) selfies’. In general, the word ‘selfie’ is used to refer to a specific type of picture: a self-portrait taken by a person themselves at arm’s length or in a mirror (Karaian, 2015: 337). Among research participants, the word selfie was not very popular. Rather, participants used the broader concept of ‘pictures’, and hardly ever distinguished selfies from other portrait pictures, such as pictures taken by a friend. Therefore, this study also involves those other portrait pictures. I do refer to these pictures as selfies though, in order to capture the unique character of digitally shared pictures in terms of their networked distribution, consumption and ubiquity (Donnachie, 2015). In this section, I will outline some general trends in research participants’ definitions of ‘sexy selfies’.

Most research participants considered making, sharing and caring about selfies a ‘girl thing’. Sometimes, the topic of a conversation even ‘automatically’ changed from pictures in general towards girls’ pictures; a mechanism that also occurs in wider public and academic debates about selfies (Burns, 2015). Especially sexy selfies were often framed in this gendered way, even to the extent that some research participants found it hard to imagine what a boy could do to look sexy (see also Handyside and Ringrose, 2017).

In trying to define what makes a girl’s picture ‘sexy’, research participants typically referred to certain bodies, outfits, poses and contexts. Bodies that were usually associated with sexiness were ‘slender but curvy’ (especially breasts and bottom), healthy and able, and ‘young’, with long legs and a ‘pretty’ face that is symmetrical, without braces or glasses, and with a smooth skin, full lips and long hair (see also Duits, 2008; Gill, 2009; Naezer, 2006; Orbach, 2009). These beauty standards are reinforced and raised through social media’s features that allow for pictures to be edited.

Outfits and poses that were typically regarded as contributing to sexiness were those that were evaluated as ‘emphasising’ (certain parts of) the body. Outfits that were generally considered as (potentially) sexy were outfits that left ‘some’ skin uncovered, especially of breasts, belly, legs and bottom, and outfits that ‘revealed’ bodily shapes, such as tight shirts and trousers, shirts revealing cleavage (inkijk), crop tops and short skirts and shorts. Research participants explained that certain outfits were sexier in some contexts than in others. For instance, bikini pictures taken in a bedroom were regarded as more sexual than bikini pictures taken on the beach. Poses that were often described as (potentially) sexy were poses evaluated as ‘emphasising’ buttocks or breasts (e.g. flexed hip, leaning forward) or lips (‘duck face’). The characteristics that are regarded as markers of sexiness are thus multiple and to some extent subjective, and research participants engaged in lively debates about whether specific pictures were sexy or not. These debates became especially vigorous when they concerned girls’ own pictures.

In these debates, it became clear that rather than demonstrating a unified girl culture, girls negotiated sexiness in different ways. In many instances, girls rejected sexiness and resisted a labelling of their own selfies as sexy (see also Ringrose, 2011). One important reason for this became clear during a focus group meeting, where I asked participants whether sexiness is different for boys and girls:

A lot can be said about this interaction and its reproduction of the gender binary, double sexual standards and heteronormativity, but what I want to emphasise here is these girls’ interpretation of sexiness as a marker of sluttishness. This association between sexiness and sluttishness, which was common among research participants, makes it complicated for girls to share selfies that might be evaluated as sexy, or even to claim a positive attitude towards sexiness and sexy selfies.

Other girls did share selfies that could be labelled as sexy, and some girls even enthusiastically embraced sexy selfies, made and shared them, and defended them against negative comments. These pictures sometimes brought them considerable advantages. For instance, the pictures played a role in attracting the attention from potential partners, they contributed to intimate or erotic conversations, they yielded positive feedback resulting in feelings of self-esteem and connectedness, and they helped to increase popularity (see also Lamb et al., 2016; Ringrose, 2011; Ringrose et al., 2013). Indeed, I observed both young people and adults rewarding ‘sexy’ pictures with compliments. Sexiness is thus not only rejected, but also embraced. In the following paragraphs, I will further explore girls’ different positionings towards sexiness and sexy selfies, and analyse how these contribute to the performance of intersectional identities.

Embracing sexiness

In some instances, research participants embraced sexiness. This happened for instance during a focus group meeting with five girls of colour who were selected based on their identification as (Dutch-)Antillean. ⁴ Participants were asked to comment on five Facebook profile pictures of girls and women unknown to them. Some of the pictures might be evaluated as sexy or even ‘slutty’, based on the outfits and poses (e.g. ‘bikini pictures’). The focus group participants were aware that some people might consider the pictures to be ‘slutty’, but explicitly refused to categorise them as such. One of the girls remembers a similar case:

Through her comment, Cynthia produces two groups: ‘Dutch kids’, meaning white peers without a migration background, and ‘others’, whom she constructs as non-Dutch, thus reproducing a common Dutch discourse that works to exclude certain racial and ethnic groups from ‘Dutchness’ (see also Mepschen, 2016; Wekker, 2016). According to Cynthia and Joella, ‘Dutch kids’ have a relatively negative attitude towards sexy pictures. Later, during a discussion about one of the ‘bikini pictures’, the girls extend this argument:

Here, another racial/ethnic category is introduced by the girls: that of ‘Moroccans’, referring to non-white Dutch citizens with an immigration background. Not only ‘Dutch kids’, but also ‘Moroccans’ have a negative attitude towards sexy selfies, according to the girls. They contrast this to their own attitudes:

Carmen and Cynthia claim to be unaffected by other people’s comments on their pictures: if they want to share a sexy selfie, they will do so. Their attitude is ‘typically Antillean’, they explain while discussing a picture of a woman whom they consider to be overweight: ‘This could be a typical Antillean woman. This is an Antillean woman, because fat Antillean women don’t give a damn. … They just think: I will wear this, I like it, and I will go out in it.’ According to the girls, Antillean women express pride in their body by showing off that body, regardless of possible negative comments. This contradicts Lamb and Plocha’s findings (2015), which describe how girls of colour in a US context refer to ‘respectability’ rather than sexiness in their performance of black femininity. For my research participants, their positive attitude towards sexiness functioned as a marker of racial and ethnic difference between them and other girls. Their emphasis on embracing sexiness can be interpreted as a form of boundary work, that reproduces racial and ethnic boundaries and contributes to the performance of Antillean (and sexy) femininity. This was supported by the context of this particular focus group meeting, which was explicitly aimed at girls identifying as ‘(Dutch-)Antillean’ – a framing that strongly highlighted ethnicity.

One of the remarks quoted above reveals that there is more to say about this meeting however. By imitating a posh voice that Dutch peers presumably use when regarding a bikini picture as ‘trashy’, Enith brings class into the discussion. She categorises Dutch people’s presumed negative evaluations of sexy pictures as a ‘higher class reaction’, and distances herself from such an identity by mocking it – a strategy called ‘anti-pretentiousness’ by Skeggs (2004: 114). Skeggs explains that, on the one hand, this strategy functions to critique people from higher social classes, but on the other hand, it also operates as a mechanism that keeps people in their classed place. Enith categorises her own positive attitude towards sexy pictures as ‘lower’ class, and reclaims this as a positive identification. Gender, sexuality, ethnicity and class thus intersect in the girls’ comments on sexy selfies, through which these girls perform lower class, Antillean, sexy femininity. Moreover, through this performance they redefine sexiness as a form of ‘erotic capital’ (Hakim, 2010) that can work to their advantage, especially when other forms of capital (economic, social, cultural) are less accessible to them.

Rejecting sexiness

Whereas in some instances girls embraced sexiness, in many others they (partly) rejected it. This happened for example during an interview with one of the ‘older’ girls, Femke (18), who had just started her bachelor studies at the university of applied sciences. We met at her school, and talked about her study before we started the actual interview. At one point during the interview, Femke explains that she does have some experience with sharing sexy selfies, but she immediately nuances this: ‘But of course, I’m more careful about what kind of pictures I send to whom, and what I post online, than girls who are still in secondary education. Because sharing such pictures is risky.’ According to Femke, the difference between secondary school pupils and university students is not just about age, but also about maturity: ‘Here [at the university of applied sciences], people are more mature.’ Maturity is an important dimension for young people, which is associated with positive characteristics (see also Duits, 2008). For Femke, reproducing the ‘danger discourse’ about sexy selfies and partly rejecting these pictures worked to distance herself from girls in secondary education, and to perform a self that is not only ‘older’, but also more ‘mature’; a performance that was supported by the place where the interview took place and my questions about her studies prior to the interview.

The (partial) rejection of sexy selfies was also related to the construction of other differences. For instance, while interviewing ‘higher’ educated girls about sexy selfies, these girls repeatedly advised me to also interview ‘lower’ educated girls. According to Erica (14), I had a higher chance of finding sexy selfies among lower educated girls, because: ‘vocational students are worse. They curse more, are trashier, with short skirts that show their ass and shirts that are too small, with leopard print.’ During a focus group meeting, Lea (15) provides a slightly different explanation: ‘I think that people with a lower [educational] level are ready for certain things earlier and think less about the consequences of their actions. The consequences of being with different boys, or of posting pictures of yourself online.’ Both Erica and Lea thus associate sexy/slutty pictures with a ‘lower’ educational level, which especially Erica associates with classed characteristics such as bad taste, lack of reflexivity and responsibility, and hypersexuality/sluttishness. This fits in the longer history of the slut category carrying particular ‘lower class’ connotations such as ‘tastelessness’ (Attwood, 2007; Skeggs, 2004). Distancing themselves from sexy selfies and saying that I should interview ‘lower’ educated girls worked for Erica and Lea to construct educational level as a relevant category, and to perform a higher class, higher educated femininity; a performance that was afforded by the school context in which the interviews took place.

Girls who are ‘lower’ educated do not have such an obvious lower educated group to contrast themselves with. Nevertheless, they too make claims about smartness, like Kyra (15) did in an interview:

Contrary to the Antillean girls described in the previous section, Kyra distanced herself from sexy selfies, especially (semi-)nude selfies, claiming that she did not make such pictures because she ‘knew better’ (see for similar findings Jackson and Vares, 2011: 139; Lamb et al., 2016: 537). She explicitly linked this to her identity as ‘not a fool’, thereby constructing smartness as a relevant category and performing a smart, knowing self that carries higher educated/higher class connotations, while at the same time making her ‘lower’ educational level less relevant. This performance of ‘smart girl’ femininity, which is part of present day post-feminist discourse (Pomerantz and Raby, 2017), intersects with ethnicity, as Kyra’s whiteness probably contributes to her being able to do this: in Dutch culture and abroad, whiteness usually connotes higher class status (which is in turn associated with smartness), while blackness is more easily associated with lower class status (Wekker, 2016: 47).

Nevertheless, also girls of colour used a rejection of sexiness as a way of claiming smart femininity. Although it does not directly address sexy selfies, this observation of a sex education class for girls (both white and of colour) is a telling illustration:

Class is about to start and most pupils have already arrived when Rita comes in. She wears an orange, tight, short dress with black tights, slippers and a short black jacket. The dress only just covers her buttocks and as she sits down, her underpants become visible. Upon her arrival in class, the atmosphere changes. The girls sitting across Rita look at her and start giggling. Rita notices and pulls her skirt down a little. Two girls repeatedly ask if Rita went to see somebody in that outfit, emphasising the word ‘that’ with a disdainful tone of voice. During the entire class, the atmosphere remains noisy. Rita herself is quiet, and has only little interaction with the other girls.

By condemning and correcting Rita’s sexiness, the other girls not only distance themselves from her, but also construct themselves as being in a position where they can advise her about where to go in such an outfit, thereby communicating that they ‘know better’. This specific performance of smart femininity was facilitated by Rita’s background as a girl who was formerly in a school for special education; a characteristic that is associated in dominant discourse with limited cognitive abilities, dependence and helplessness (Benjamin, 2002), which put the other girls in a position of power from where they could openly criticise Rita’s sexy appearance and thereby position themselves as smarter, and by implication as higher class.

Yet another dimension seems to be at stake here. One week after this particular observation, I participated in the same group. Rita was not present; she had been transferred back to the school for special education. This time, another girl, Sydney, wears a short skirt, probably even shorter than Rita’s orange skirt, with black tights similar to Rita’s. Not a single negative comment is made about Sydney’s skirt though, and during class, Sydney and the other girls share stories about adventures they experienced together. Sydney is much more popular than Rita, which seems to provide her with extra space for performing sexiness (for a more elaborate discussion on the relation between heterosexiness and popularity, see Duncan, 2004). This role of popularity was voiced explicitly by another research participant, Erica (14). Erica was very much aware of the slut stereotype, and complained to me about Amy, whom she considered slutty because she ‘always’ posted selfies showing cleavage or with an ‘accidental’ bra somewhere in the background. However, when I asked Erica about the profile picture of her classmate Tess showing a bare shoulder and a small piece of bra, she replied: ‘But she is sweet so then it is allowed.’ Here too the popular girl gets more space for performing sexiness than the less popular girl. Vice versa, this also means that performing sexiness or demonstrating a positive stance vis-a-vis sexiness can in some cases contribute to the performance of popularity.

Contradictory positionings

Girls’ positionings towards sexiness can be ambiguous and inconsistent. One case that demonstrates this is the case of a Christian girl (Judith, 15), who voiced contradictory opinions about sexiness during a focus group meeting with girls identifying as Christian. ⁵ Religion was often described as a reason to reject sexiness, and also during this focus group meeting, research participants indicated that looking sexy was ‘un-Christian’. At several points, Judith (15) agreed with this and demonstrated a negative attitude towards sexiness. For instance, she explained that in order to be recognised as a Christian girl, she did not share sexy selfies in online spaces such as Facebook, and she made several negative comments about sexy outfits. She even showed us a ‘joke’ on her phone of a girl revealing cleavage, accompanied by the text: ‘Only God can judge me’, followed by an image of God, answering: ‘You’re a whore.’ Judith couldn’t stop laughing about this slut-shaming ‘joke’ and showed it to the other girls, who agreed with her that this was funny. By sharing this image, and laughing about it, Judith constructs religion as a relevant category, distances herself from the sexy girl on the picture, and performs a ‘pious’ Christian femininity that was recognised and rewarded by her peers.

At other moments, however, Judith took a much more ‘un-Christian’ position. For instance, while the other participants agreed that French kissing should happen only after a girl has flirted with somebody for a while, Judith said: ‘You can also kiss … for example at a party, when you see somebody whom you don’t know, and you just kiss that person. You don’t have to flirt first.’ She also openly discussed her romantic and sexual experiences. About monogamy, a norm that was regarded as important by the other focus group participants, Judith said: ‘If I see a cute guy, even when I’m in a relationship, I’ll just flirt with him. That’s just who I am.’ About ‘sexy outfits’ she commented: ‘To God, it doesn’t matter. He created us and always sees us, also when we’re in the shower, so … Why would you be ashamed?’ Judith even explained that she liked to dress sexy in spaces where (adult) church members were not part of the audience. These (partly) positive positionings towards sexuality and sexiness seem contradictory to her identification as Christian, and the rejection of sexiness that is associated with that identification.

According to Judith, however, the two positionings can go together because of her age: ‘It’s also about age. Because we hit puberty, and then you want to try things out, and that’s allowed I think, also if you’re a Christian.’ Judith did not feel the need to completely reject sexiness, because she identified not just as a Christian girl, but as a young Christian girl. By making this statement during the focus group meeting, she constructed age as a relevant category and performed young, Christian, sexy femininity, allowing herself some room for sexual experimentation.

At the same time, by taking a positive stance towards sexy selfies and sexual activities more generally, Judith made clear that her young, Christian, sexy subjectivity was not an ‘immature’ subject position, thus distinguishing age from maturity. For young people, the body and sexuality play a crucial role in the performance of maturity (see also Duits, 2008). Being self-confident, being able to talk about sexuality without (showing) discomfort, and being sexually active (within certain limits) were usually regarded as signs of maturity. Against this background we can understand Judith’s emphasis on being ‘open’ about her romantic and sexual relations, and her (partly) positive attitude towards sexiness and sexy selfies, as contributing to the construction of a Christian sexy self that is young, but also mature.

Conclusion

In this article, I analysed how teen girls perform intersectional identities through their navigations of sexy selfies. While feminist sociologists and cultural media studies scholars have studied dominant discourses about sexiness, only a few studies have analysed teen girls’ own reflections on the topic. My analysis demonstrates that girls navigate sexiness and sexy selfies in different ways: they may (partly) reject sexiness, (partly) embrace it, or take a more contradictory position. These positionings are not static, and they may change even in the course of one conversation.

I analysed girls’ navigations of sexiness as a form of boundary work: by labelling their position towards sexiness as markers of difference, girls (un)make differences among themselves, and perform complex intersectional identities. Axes of social difference that play a role in this process are not just gender and sexuality, which have been central to previous studies, but also other axes. Some of these are well-known (ethnicity, class, educational level, religion), whereas others have remained largely invisible in previous studies of youth, sexuality and social media (smartness, maturity, popularity). The constructions of these multiple differences interfere in ways that may be dynamic and unpredictable, resulting in a variety of possible subject positions.

My study also showed how girls’ positionings and identifications interact not only with dominant gendered, heteronormative, racialised, classed and religious discourses about sexiness, but also with the materiality of girls’ bodies (e.g. bodily shape, skin colour), with girls’ perceived social position (e.g. popularity, educational level, class) and with the specific context in which the topic is being discussed (e.g. an interview in a specific school setting, or a focus group meeting framed in a particular way). This means that while the interferences between constructions of multiple differences can be dynamic and unpredictable, they do not exist in a social and material vacuum, and they may also work to reproduce dominant categorisations, social norms, stereotypes and power relations.

These findings indicate that dominant popular and academic conceptualisations of sexy selfies as no more than a matter of sexual seduction, narcissist vanity, insecurity, or self-objectification are extremely limited and ignore the link between sexiness, social categorisations and intersectional identities, not just in adult women’s, but also in teen girls’ selfie making practices. Acknowledging these interconnections, and making them part of the discussion about sexy selfies may create promising opportunities for interrogating social norms, stereotypes and power inequalities.

Finally, my study illustrates that all axes of social difference should be analysed as social constructions. Axes such as educational level and age are still often presented as objective ‘facts’, but my study demonstrates that they are in fact performative accomplishments, that can be performed for instance through rejecting or embracing sexy selfies. For research participants, this performative nature was captured in the concepts of smartness and maturity. Therefore, I propose to use the concept of smartness in addition to educational level, and that of maturity in addition to age, much like we use gender in addition to sex, in order to facilitate a social constructionist analysis of young people’s navigations of sexy selfies, sexiness and sexuality.