‘Gay, aren’t they?’ An ethnographic approach to compulsory heterosexuality

]odie Clark


Ethnography holds a particular appeal for researchers who wish to resist generalisations about gender and sexuality and focus instead on the local ways in which gender and sexuality affect people’s lives. Language, gender, and sexuality ethnographies have offered insights into the unique gendered and sexed practices in schools (Bucholtz 1999; Eckert 2000; Moore 2006), sports teams (Clark 2012; Sauntson and Morrish 2012), LGBTQ+ groups (Jones 2012, 2016) and drag queen performances (Barrett 1999), to name just a few. Ethnographic research can also offer a window into the sometimes small but always significant ways in which people’s lives are adversely affected by constrictive gender and sexuality norms. By focusing on individual experiences of gender and sexuality, ethnography can posit an implicit counter-argument to claims that ‘everything is so much better than it used to be’ in relation to sexism, homophobia, and transphobia.

One focus of ethnographic research is on the norms that participants in various communities and social spaces abide by. Sometimes these are explicit - they can be explained by members of the community in interviews for instance - but they are more often under the radar. Ethnographic methods are often used to identify the unwritten rules of interaction, acceptance, and belonging. Language, gender, and sexuality ethnographies frequently focus on the heteronormative frameworks that govern particular social spaces (Bucholtz 1999; Clark 2012; Eckert 2011).

This chapter focuses on what Adrienne Rich (1980) calls ‘compulsory heterosexuality’ - a set of ideologies that make it very difficult, especially for women, to imagine a life free from the power imbalances of heterosexual relationships. I draw upon two ethnographic studies - one of a university women’s field hockey team and one of a university friendship group - to explore instances in which compulsoxy heterosexuality effected both physical and ‘symbolic violence’ (Bourdieu and Waquant 1992: 166). I focus on instances in which participants were required by their peers to engage in heterosexual activity - specifically, they were compelled or coerced to kiss men against their will.

Seidman (2009) argues in favour of making a clear distinction between heteronorma- tivity (as discourses, norms, and social pressures) and compulsory heterosexuality (where heterosexuality is an official requirement). From my point of view, to focus on this distinction is to miss an important perspective. Seidman’s account presents both terms, heteronor- mativity and heterosexuality, as oppressive forces that impose themselves on individuals. I propose instead an approach that draws upon ethnographic data to identify those fleeting instances in which heterosexuality makes itselffelt as compulsory, where the result of not engaging in heterosexual acts results in violence to the self. The opportunity then presents itself to explore the structure of the social space at that precise moment, and then to ask questions about how it might otherwise have been strucmred.

The approach I describe here makes use of the analytic method described in Clark (2016), which outlines a method for analysing ‘the grammatical structures of participants’ accounts of their social worlds’ (2016: 9) as a means of imagining possibilities for social transformation. The methodology supplements ethnography with the methods of critical discourse analysis (CDA) described in Part V, ‘Sociocultural and critical approaches’, of this Handbook (for other uses of CDA within gender and sexuality research, see Koller 2015; Morrish 1997 and Peterson 2010). My approach differs from CDA in that it is designed to orient not to ideology (claims to truth), but rather to discursively constructed selves and social worlds:

I propose a CDA that asks different questions. ... Instead of, what are the claims to truth, and wherein lies the straggle? I would ask, what is the shape of the social structure here, and where is the desire for an alternative? What are the ‘selves’ that are textually constituted here and how might they be otherwise constituted? What are the possibilities for transformation here?

(Clark 2016: 38)

When these questions are applied to compulsory heterosexuality, the orientation shifts from understanding heteronormativity as an oppressive discourse (Motschenbacher and Stegu 2013), an inescapable ‘contract’ (Wittig 1992: 34), or a regulatory ‘matrix’ (Butler 2006: 208). Instead, instances in which heterosexuality makes itself felt as compulsory in participant accounts can be used as windows into the shape of the social world at that precise moment. A key focus in this chapter will be on the different ‘selves’ that are grammatically constructed in these accounts. I will explore two different ways in which selves are presented as vulnerable in the face of compulsory heterosexuality. My analysis demonstrates that grammatical constructions of the vulnerable self coincide with grammatical constructions of alternative social worlds - worlds in which vulnerable selves are protected, and heterosexuality loses its force.

The method of analysing ethnographic data I demonstrate in this chapter is unique from other approaches in that it does not focus on norms, discourses, belief systems, and ideologies that structure a particular social space or community. The analysis here is concentrated on very specific instances within the context that the ethnographic fieldwork provided - precise moments when compulsory heterosexuality is at issue in participants’ accounts. The aim here is not to chart how heteronormativity governs a particular social group, but rather to identify moments within social group interactions in which alternatives to heteronormativity present themselves.

The ethnographic fieldwork

Another way in which the analysis I present in this chapter is different from other ethnographic work is that it draws upon data collected from two different ethnographies, situated within two different social groups. The first is a women’s university field hockey team (Clark 2012) and the second a small group of students in a module I taught called ‘Language and social life’ (Clark 2016). Comparing these communities enables me to explore two very distinct moments of compulsory heterosexuality, to situate these in two different social contexts and to identify the distinct ways in which the selves and the social worlds are constructed in each. Despite the participants’ distinct responses to these moments, it became clear (as I will show) that the accounts of compulsory heterosexuality consistently reveal, first, the grammatical construction of a vulnerable self and, second, the grammatical construction of an alternative social world.

The members of the Midland University1 women’s field hockey team were straight cis women, white British middle-class, between the ages of 18 and 21. They identified as high achieving hockey players, students, and citizens. They made it clear that to achieve in all these areas requires not only securing an identity as feminine and straight but also silencing any discussions of sexual desire (Clark 2012). In other words, their identities depended upon conforming to heterosexual norms.

The participants in the second group were three students (Beth, Maiyam, and Andrew) in a module I was leading called ‘Language and social life’. Troubled by the ways in which these three friends had been alienated by other smdents in the seminar group that I was leading, I asked them, at the end of the academic year, if they would be willing to talk about their experiences of feeling marginalised as part of a focus group. During the focus group Beth, Maiyam, and Andrew revealed that what brought them together as friends was a shared resistance to the norms of the other students on their degree programme. Each of them had experienced a sense of lack of acceptance from others on their course, which they attributed to a certain extent to aspects of their identities: for Beth, it was her unapologetic working- class status; for Maryam, it was her ethnicity (Pakistani-British) and her religion (Muslim); for Andrew, it was his disability (which affected his mobility and required him to sometimes use a crutch) and his non-normative gender (transgendered man).

The similarities between the two groups (their ages, for instance, as well as their engagement in university communities of practice (Eckert and McConnell-Ginet 1992)) serve as a backdrop against which to highlight the points of contrast around compulsory heterosexuality. The hockey players expressed explicit homophobic attitudes throughout the ethnography (see Clark 2012 for an analysis of these). The ‘Language and social life’ group, on the other hand, consistently voiced their disapproval of any exclusionary attitudes and advocated in favour of diverging from the mainstream. The fundamental differences in perspective of the two groups offer a unique backdrop from which to explore their responses to specific moments of compulsory heterosexuality.

The role of the ethnographer

An important part of the complex decision-making process involved in doing ethnography involves the researcher’s relationship to the community or the social space they are studying. Most ethnographic methodologies define these relationships in terms of four types of role, defined in Gold’s (1958) work: the complete participant, participant as observer, observer as participant, and complete observer. As O’Reilly (2009) points out, however, most ethnography requires some form of participation and some form of observation, and roles are likely to shift as the ethnography progresses. In my ethnography of the university field hockey team, I initially understood myself to fit within the category of ‘observer as participant’ -1 was not part of the hockey team, nor did I know any of the team members before asking their permission to conduct the study. As I became more involved in their activities - attending socials, supporting them at matches and, indeed hosting meals -1 moved more into the ‘participant as observer’ role. The duration of the hockey-team ethnography was an academic year, during which time I observed the team's activities at training sessions, matches, and social events. I collected conversational data from ten members of the team over a period of seven ‘data-gathering evenings’, in which I served meals to self-selected groups of two or three participants and recorded their conversations. I allowed the participants to steer the conversation for these evenings, contributing just enough to stay involved in the conversation. The data-gathering evenings produced a corpus of about 18 hours of conversational data, a corpus of225,000 words.

My participation in the second ethnography followed a different trajectory, one in which I moved from being a ‘complete participant’ (a member of the community I was investigating, as their tutor) to a ‘participant observer’ (a researcher seeking answers to questions that had emerged as part of my institutional participation). ‘A complete participant’, O’Reilly explains, ‘is not an ethnographer; he or she is a participant. If she decides to research the group or culture in which she participates, she becomes a participant observer’ (2009: 154). The focus group marks the moment when I became participant observer.

The second ethnography was much smaller in scope from the first: I had known and observed the participants for five months before inviting them to participate in the focus group. As for the first ethnography, I allowed the participants to steer the conversation, but asked a few open-ended questions designed to elicit narratives about belonging or exclusion. The focus group lasted 2 hours and generated a corpus of about 20,000 words, which made it just under one-tenth the size of the hockey-team ethnography.

Beginning in the ‘observer as participant' role, as I did in the hockey-team ethnography, required me to engage in quite a lot more information gathering, explicit observation, and permission-seeking. I was an outsider to this community, and indeed, to the larger world of university spoil in which this community was situated, so I had to gain access by introducing myself to prominent members of the community, including the team captain and the president of the Student Athletic Union. I was also dependent upon the willingness of team members to allow me to participate in their team-only social events and to attend the meals I had arranged. My initial position as outsider required me to fill the gaps in my knowledge about the team and its context by engaging in a large amount of time-consuming fieldwork, research, recording, and transcription. As Eckert remarks, ethnography 'requires a considerable time investment, and frequently more of an investment than an academic [or student!] is free to make’ (2000: 74).

The second study I conducted provides some ways of using an ethnographic approach when it is not possible to engage in such intensive, time-consuming fieldwork. As I mentioned above, my role as tutor on the participants’ course of study made me a ‘complete participant' in the community I chose to later to observe, which meant there were fewer issues of access and information gathering. The challenge for smaller-scale ethnographic studies is less about time and more about perspective and critical distance. As Hammersley and Atkinson explain, 'it is in the “space” created by this distance that the analytic work of the ethnographer gets done. Without that distance, without such analytic space, the ethnography can be little more than the autobiographical account of a personal conversion’ (2007: 90).

It is worth pointing out, however, that to do research in language, gender, and sexuality is already to engage in the intellectual and critical distance that Hammersley and Atkinson describe. The priority within Queer approaches, for instance, is to see heterosexuality as constructed in discourse. The term ‘Queer Linguistics’ (Motschenbacher and Stegu 2013) is often used to encompass research that critically analyses how language use reinforces or challenges heteronormativity. Many scholars working within Queer linguistics have drawn upon ethnographic methods as a means of achieving this critical angle.

Because ethnography privileges sustained engagement with participants over a longer period of time than survey methods, they usually generate richer data in relation to each participant in the study. The rich data my own ethnographies produced made it possible for me to draw connections between participant narratives and community norms and practices, and thus to recognise instances in which symbolic violence was being chronicled. As Bourdieu and Waquant (1992: 166) point out, 'symbolic violence’ can work more ‘efficiently’ than physical violence as a form of social control: it can cause individuals to undermine those aspects of self that they consider to be valuable. Symbolic violence is not as easy to detect as physical violence, because participants may not be able to recognise it as such. The ethnographic data from both studies offered a rich source of participant narratives in which compulsory heterosexuality inflicted symbolic violence.

Grammatical analysis and social change

A familiar difficulty for the ethnographer is how to analyse the vast amounts of data that ethnographic methods generate. As Angrosino points out, the analysis of ethnographic research is most often ‘“custom-built” to suit the needs of particular projects’ (2007: 74). In an earlier work (Clark 2012) I discuss how heteronormativity serves as a structuring principle for the local and institutional practices of the hockey team. In this chapter, I am focusing on participant narratives in both groups that depict some form of compulsory heterosexuality - what I have called ‘moments’ of compulsory heterosexuality. The shift from analysing norms to analysing participant accounts of their social worlds is motivated by a desire for empirical research to orient more explicitly towards social transformation. My argument is that 'critical readings of individual accounts reveal traces of new, yet-to-be-imagined social structures’ (Clark 2016: 8) and that by bringing these new possibilities to light, researchers will be contributing to social change.

As I mentioned above, the critical readings I offer are informed by grammatical analysis, specifically Halliday’s (2014) Systemic Functional Grammar, which describes language in terms of three metafunctions: ideational (how human experience is constmed), interpersonal (how social roles and relationships are constmed), and textual (how textual coherence is achieved). I argue that this type of grammatical analysis ‘can reveal the structure of a given text (textual metafunction), the specific constitution of each “self’ constmed or invoked in a text (interpersonal/ideational metafunction) and the “shape” of the world that is produced in a segment of discourse (ideational metafunction)’ (Clark 2016: 36). In this chapter I focus particularly on the ideational metafunction, drawing upon Halliday’s framework for analysing transitivity. An analysis of the transitivity of a chunk of language categorises how different types of experience are constmed as processes (through verb phrases) and participants (usually through noun and prepositional phrases). I will demonstrate how these categorisations construct different configurations of selves and social worlds.

The grammar of compulsory heterosexuality

As I mentioned above, drawing upon two ethnographies offers the possibility of exploring distinct responses to similar moments - in this case, moments of compulsory heterosexuality. I was fascinated to discover that the conversational data from both the hockey team and the ‘Language and social life’ ethnographies offered moments in which participants told stories about women being coerced into kissing men. At first glance, the narratives seem rather different. Consider, for instance, Extract 1, which is part of a conversation between three members of the hockey team - Speedo, Eimna, and Sullivan. They are all middle-class cisgender women between 20 and 21 years old. Speedo and Emma are from the English Midlands; Sullivan is from Noxthern Ireland. Speedo is team captain and is in her third and final year at university; Emma and Sullivan are in their second year. Speedo is telling Emma and Sullivan about travelling home on the bus after a match with the women’s and men’s football teams, where the 'football boys’ tried to get the ‘football girls’ to kiss them. When the women refused to kiss them, the men came to the conclusion that they were gay, which Speedo confimxed.

Extract l2

  • 1 Speedo: they ((the football boys)) were trying to make the football girls snog
  • 2 one of them [and I was like oh, it’s not gonna happen]
  • 3 Sullivan: [Oh were there football girls in the bus]
  • 4 Speedo: It’s not gonna happen. They’re like gay, aren’t they. And I was
  • 5 like, lmnhm, yeah? [((laughs))] I KNOW them
  • 6 Sullivan: [((laughs))] [All of them are!]

It takes Speedo only a few seconds to tell the story, and the response to it is light-hearted, as evidenced by the laughter on Lines 5 and 6. Indeed, as will be shown later, the focus of the subsequent conversation is on which players in women’s football and hockey are gay - not on the men’s sexual advances.

Extract 2, from the ‘Language and social life’ ethnography, seems to tell a very different story. Beth is retelling (for my sake) an account she has already shared with Andrew and Maryam, about her experiences of being bullied at school as a child. Beth describes being paid of a ‘wrong crowd’ when she was in school. The ‘wrong crowd’ was a group of girls with a set of rules about how to fit in, what to wear, how to do your hair, who to go out with. The consequences for not abiding by these, for Beth, was physical and emotional abuse.

Beth, Andrew, and Maryam are all 20 years old. Beth and Andrew are white British, from the same working-class village in the north of England. Maryam is Pakistani British from a working-class city in the north of England. Beth and Maxyam are cis women; Andrew is a trans man. Extract 2 comes from Beth’s account of her school friends trying to force her to kiss a boy. When she refused, her friends became physically violent.

Extract 2

  • 1 Beth: Erm (0.4) they tried to make me go out with this boy that I didn't want to-
  • 2 (0.7) like, I didn’t want to be near him, he were nasty. (0.4) And then when I
  • 3 wouldn't kiss him, cos I-1 didn't want to, I mean, I were like "(thirteen and he
  • 4 were a dick (0.7) a:nd er when that didn’t happen they shoved me off a kerb?
  • 5 (0.4) And I cut all me leg, I’ve still got a scar.

The tone is very different in Beth’s account of compulsory heterosexuality than in Speedo’s. Beth’s story has significance to her personally - it represents a specific memory from her childhood that she has shared at least once before. Beth presents herself as the victim of the sexual coercion here. Speedo’s story, in contrast, is of something that happened only a few days before she tells it, and she presents herself in Extract 1 as an observer, not as a main character in the narrative.

A grammatical analysis of some of the clauses in each of the accounts, however, reveals some interesting similarities. In both accounts the ‘compulsory’ moment is grammatically construed in what Halliday calls a ‘three-participant causative’ (2014: 579), where one participant - the ‘initiator’ - causes another - the ‘actor’ to act on a final participant - ‘the goal’, as shown in Table 8.1.

A further similarity is revealed through an investigation of the process types in the clauses of each account. Material clauses (processes of doing and happening) predominate in Beth’s account, but there are also some mental (thinking and feeling) and relational (being) clauses, as shown in Table 8.2.

Speedo’s story, too, consists of mostly material clauses, with one mental and one relational clause. Her account also includes verbal (saying) clauses, as shown in Table 8.3.

The most significant parallel between the two stories comes into view through a comparison of the material processes in each, as shown in Table 8.4. Restricting the accounts only to their material clauses suggests a striking similarity in the grammatical form of compulsory heterosexuality. In both cases ‘heterosexuality’ takes two forms: in the first instance, as a transitive verb indicating sexual interaction: ‘go out with’, ‘kiss’, and ‘snog’; in the second as the intransitive verb 'happen'. The 'compulsory' element takes the form of the causative ‘to make’. In addition, in both stories the causative is extended by a process of conation, or trying to do something (Halliday 2014: 572).

Table 8.1 Grammatical analysis of some clauses



they were trying to make

the football girls


one of them

they tried to make


go out with

this boy




Table 8.2 Investigation of the process types in the clauses of each account

Clauses in Beth's account

Process types

they tried to make me go out with this boy


when 1 wouldn't kiss him


when that didn't happen


they shoved me off a kerb


1 didn't want to


1 didn't want to be near him


1 didn't want to


he were nasty


1 were thirteen


he were a dick


Clauses in Speedo's account

Process types

they were trying to make the football girls snog one of them


it's not gonna happen


it's not gonna happen


1 KNOW them


gay, aren't they


1 was like


they're like


1 was like, mmhm, yeah


Table 8.4 Material clauses in Speedo's and Beth's accounts

Speedo's account

Beth's account

they were trying to make

they tried to make

the football girls snog one of them

me go out with this boy when 1 wouldn't kiss him

it's not gonna happen (x2)

when that didn't happen they shoved me off a kerb

Social worlds and vulnerable selves

Having identifying some parallels in the grammatical structure of these two accounts, we are now in a position to inquire about the shapes of the social worlds these accounts offer, including how ‘selves’ are constructed. Consider first how each account construes the participants in the scene. Although it is clear that there were multiple people in each scene, the number of clausal participants in both accounts is reduced to three, as shown in Table 8.5. In Anna’s account, these consist of the plural noun phrases ‘the football boys’ and ‘the football girls' (with their associated third-person plural pronouns, ‘they’ and ‘them’), and the first- person pronoun T. In Beth's account the participants are reduced to ‘they’ (Beth’s school friends), ‘this boy’ (and the associated third-person singular pronouns, ‘he’ and ‘him’), and the first-person pronoun T. Note the tendency in each account to treat the groups as discrete entities; Beth never singles out any of the members of her friendship group (here or at any other time in the conversation), and Anna only does so one time, with her use of the indefinite pronoun ‘one’ as it is used in relation to the group (‘one of them’).

An important contrast can be seen in how each account constructs the participants involved in the act of coercion. Both set up an antagonism between two parties. For Speedo these consist of a heterosexual binary between ‘boys’ trying to kiss ‘girls' who do not want to. In Beth's account, on the other hand, the antagonism is not between male and female participants, but between two female participants, Beth and her group of female friends. What is perhaps more noteworthy is that the two parties in Speedo’s account are two groups (the two football teams), whereas Beth’s narrative constructs a singular self (T) as at the mercy of the group she is ostensibly a part of. In other words, Beth’s account is shaped in relation to a singular self that is vulnerable to the violence caused by a group. Speedo’s account of compulsory heterosexuality does not seem to involve a vulnerable self.

Participants in Speedo's account

Participants in Beth's account

(the football boys)/they/them the football girls/they/them

they (Beth's school friends)

one of them (one of the football boys)

this boy/he/him

1 (Speedo)

l/me (Beth)

Table 8.6 A comparison of the clauses that serve the 'justification' function

Beth's account

Speedo's account

  • 1 didn't want to [go out with this boy] (mental) 1 didn't want to be near him (mental) he were nasty (relational)
  • 1 didn't want to [kiss him] (mental)
  • 1 were thirteen (relational) he were a dick (relational)

gay, aren't they (relational)

Consider next what happens in both accounts when the sexual act doesn’t happen. In Beth’s account, the consequence is physical harm: ‘when that didn't happen, they shoved me off a kerb'. This ‘not happening' is paralleled in Speedo’s story, but with the future modality in the verbal clause where Speedo’s talking to the football boys: T was like, “oh, it’s not gonna happen. ... It's not gonna happen’”. Rather than describing the consequences of it not happening, Speedo offers an explanation as to why 'it’s not gonna happen’. The two ‘sayers' (Halliday 2014: 303) in the dialogue (T and ‘they’) co-construct a justification: 'They’re like, “Gay, aren’t they?” And I was like, “Mmhm, yeah?”’

Beth’s story, too, provides a justification for why 'that didn’t happen’, but hers is expressed through mental and relational clauses, all of them oriented to her own reasons why she’s not kissing the boy. A comparison of the clauses that serve this ‘justification’ function, shown in Table 8.6, is revealing.

Justification for her refusal to do the heterosexual act requires a good deal of discursive reinforcement for Beth. In the first instance, it occurs through the repeated use of the mental clause T didn't want to’, and in the second through relating undesirable attributes to the boy (‘he were nasty', ‘he were a dick’). The only attribute she relates to herself as a justification for not kissing him is her age (T were 13’). Speedo’s justification, on the other hand, requires no mention of what the football girls desire, nor whether the football boys are desirable; she needs simply to assign the football girls (‘they’) with the attribution ‘gay’. When Speedo makes it clear to the football boys that the football girls are gay, they seem to stop pushing it. The selves here do not seem vulnerable to the type of violence Beth describes having to suffer. Indeed, at this point in Speedo's narrative the heterosexuality may not even seem compulsory: the boys tiy to kiss the girls, the girls refuse, no one gets hurt - or no one seems to, anyway.

A closer look, however, reveals that a vulnerable self does emerge in Speedo’s account. It occurs in the bit of the conversation that occurs just prior to Extract 1, where Speedo describes the boys inquiring and speculating about the sexualities of the other women on the bus.

Extract 3

  • 1 Speedo: All the boys kept going on, how many of the girls are
  • 2 lesbians in the bus and I was just like, oh and they eventually
  • 3 came and they were like you clearly are. ((laughing)) And I
  • 4 was like, [what?]
  • 5 Emma: [((laughs))]
  • 6 Sullivan: I’ve got a [boyfriend!]
  • 7 Speedo: [You’ve] done it! (0.4)
  • 8 And I was like [nnyah]
  • 9 Emma: These [weren’t] hockey [guys though]
  • 10 Speedo: [nnyeah]
  • 11 Emma: Oh.
  • 12 Sullivan: But they don’t know [Speedo]
  • 13 Speedo: [They were] freshers
  • 14 Sullivan: They were freshers
  • 15 Emma: Oh they were [freshers]
  • 16 Speedo: [But they were] with the- football

Compulsory heterosexuality does not show up straightforwardly in this extract as a grammatical construction with a three-participant causative. Nevertheless, it is here that we can find a parallel to the vulnerable self that shows up in Beth’s story in Extract 2. Consider the participants in Speedo’s account, as listed in Table 8.7. The singular self occurs here as the first- and second-person singular pronoun, and, as in Beth’s story, it is pitted against an antagonistic group. The antagonism here occurs not through material processes, but through a series of verbal clauses, in which 'the boys’ interrogate Speedo's sexual identity, as can be seen in Table 8.8.

It is worth pointing out, however, that the verbal processes here do not depict a straightforward antagonism between the self and group, as they did in Beth’s account. Beth’s narrative constructed her resistance to the group’s coercion through a series of mental processes: 'I didn’t want to’, ‘he were nasty’, ‘he were a dick’. Speedo's account does not offer any insights into her own mental processes: she does not indicate whether she agreed or disagreed with the boys’ assessments about her sexuality, and the verbal responses she reports suggest a lack of commitment. Indeed, the self constructed here is one that resists being labelled as either straight or gay. I would argue that it is this self - the self that resists such labelling - that is vulnerable in this account. It is vulnerable not only to the questioning of the boys on the bus, but also to the responses to her account that Emma and Sullivan offer in the rest of the extract.

Table 8.7 Participants in Speedo's account

Participants in Speedo's account in Extract 3

all the boys/they the girls ... in the bus l/you (Speedo)



the boys

how many of the girls are lesbians in the bus

you clearly are

you've done it





This vulnerability becomes visible with a closer look at how Speedo’s narrative becomes a collaborative telling at line 6. Note the different constructed dialogue responses to the moment of the story in which Speedo is labelled a lesbian (‘you clearly are’). Sullivan joins in at this moment, supplying a line of dialogue on Speedo’s behalf - ‘I’ve got a boyfriend’ - which is likely designed to put an end to further questioning about her sexuality.

Perhaps more revealing than the dialogue that Speedo attributes to herself are those she attributes to the footballer who is interrogating her. In line 3 she has the footballer’s accusation take the form of a relational clause: ‘you clearly are [a lesbian]’. Once Sullivan provides what might be considered evidence that Speedo is not a lesbian (having a boyfriend), Speedo repeats the initial accusation to take the form of a material clause in line 7: ‘you’ve done it’. This shift can be interpreted as a subtle way of countering Sullivan’s assumption that being in a heterosexual relationship precludes the possibility of being gay.

Speedo’s narrative in Extract 3 can be understood as an act of resistance to Sullivan’s attempt to impose a heterosexual identity upon her. Not only does Speedo resist being straightforwardly labelled as straight, she also resists the silencing of her non-normative sexual experiences, as becomes clear with the continuation of Speedo’s narrative, in Extract 4. Here Speedo is revealing to Emma and Sullivan that she has ‘tried it’ (in other words, she has experimented sexually with women in the past), but her previous team members never judged her for doing so.

Extract 4



But then it is- you do- the reason I- (0.2) a lot of the time I tried


it was because the other people [were]



[Everybody] did it, it’s peer


pressure, I understand [that]



[It wasn’t] peer pressure!


[It was just what I was seeing] I was like [oh, you know]



[Not peer pressure it’s just] [you’re in that]


environment [aren’t you]






Yeah, but it’s not WRONG, is it [if s-]





if eveiyone else did it it wasn’t wrong, it didn’t matter, who


cares, if somebody did it now in front of like half of our friends,


they’d all be like oh my God

Note Speedo’s attempt to give voice to her non-normative sexual experiences in Lines 1-2: 'a lot of the time I tried it was because the other people were'. Emma’s response attempts to reframe Speedo’s experience to make it understandable to her. By calling it "peer pressure’, Emma can make sense of it (’I can understand that’) because it fits within a framework that she and other teaimnates have formulated in which university students experiment with gay sex because they are insecure and have not yet matured (Clark 2013).

With Speedo's rejection of Emma’s move toward intelligibility (‘It wasn’t peer pressure!’, Line 5) comes a discussion of an image of community that Emma may not have ever considered: a community in which sexuality is not constrained by norms. In this view of community, heterosexuality is not a structuring principle, nor are sexual experiences subject to ‘pressure’; instead ‘it wasn’t wrong, it didn’t matter, who cares’ (Lines 12-13).

Speedo’s acts of resistance here reveal a ‘self’ that her friends attempt to silence in this conversation. In my discussion of Beth’s account, I described a 'vulnerable self’, vulnerable by virtue of being subject to physical violence from a norm-enforcing community. I would argue that Speedo constructs a self in her narrative that is vulnerable to symbolic violence from the silencing and from disavowing her sexual experiences.

Alternative social worlds

My approach here has been to identify instances in which heterosexuality makes itselffelt as compulsory, where heterosexuality is compulsory by virtue of its imposing violence to the self. I have argued that such accounts help us understand grammatical constructions of selves that are vulnerable in the face of compulsory heterosexuality.

After having constructed an account of a vulnerable self in the face of what she calls 'the wrong crowd’, Beth goes on to describe and construct another type of social world, one that protects her from the types of attacks she encountered with her former friendship group. She recounts joining a drama group, where she met people with whom she was to form ‘lasting friendships’ (Clark 2016: 64). ‘I just knew,’ she said, ‘that I needed to keep with these friends that I made, cos they were- they still are so protective over me’ (Clark 2016: 65). With the arrival of the new friendship group, Beth said, the ‘bullying just stopped. [...] It’s like they saw me get confident and get a good group of friends around me and just backed away’ (Clark 2016: 65).

The image Beth presents here is of use to gender and sexuality scholars, especially in relation to the account of compulsory heterosexuality discussed in this chapter. The social world Beth describes is not one that resists the force of compulsory heterosexuality, but rather one that supports and protects the integrity of each member’s sense of self.

Speedo’s account also offers a new way of thinking about community. In the same way that for Beth, her 'good group of friends’ served as ‘an envelope to protect the integrity of the self’ (Clark 2016: 71), Speedo’s nostalgic account of a more sexually progressive hockey team serves as a means of making visible a disavowed, unintelligible self.

Future directions

I have argued that ethnographic methods make it possible to identify the adverse effects of gender and sexuality norms at particular moments in particular social spaces. I have analysed two instances of compulsory heterosexuality from my own ethnographic data, both resulting in violence, one physical, one symbolic. As I mentioned above, instances of symbolic violence are harder to spot, especially in largely norm-enforcing communities. Identifying these moments is well worth the effort, though, because it sheds light on the specific, localised, and largely unnoticed ways in which people are adversely affected by heteronormativity.

More importantly, though, focusing on accounts of symbolic or physical violence makes it possible to ask questions about how the self and the social world are grammatically constructed in these accounts, and to identify the alternative structures that these accounts produce. My analysis of Beth’s and Speedo’s accounts of compulsory heterosexuality reveals a fascinating pattern. They both first constructed a self that is ■vulnerable to the norms of the community. They then made reference to an alternative type of community, one that is accepting and protective of this vulnerable self.

Much gender and sexuality ethnographic research focuses on community-level resistance to compulsory heteronormativity. That said, as Hall (2013) points out, such research focuses on how these social spaces are structured according to locally constructed norms. In Barrett's (1999) ethnographic study on drag queen performances, for instance, the norms take the form of a culture-specific notion of authenticity, which become the criteria upon which these performances are evaluated. Similarly, Jones’s (2012, 2014) ethnography of a British lesbian hiking group reveals a set of norms that centre around community-specific ideas about what it means to be an authentic lesbian. In Jones’s (2016) ethnography of an LGBT youth group the normative pattern is oriented to the racialised marginalisation of others. These studies present a pattern in which social spaces will inevitably be governed by sets of norms.

I would encourage language, gender, and sexuality ethnographers to widen their focus, such that they look not only to the norms that structure the community, but also to those fleeting moments in the data in which participants construct alternatives to oppressive structures. A close look at Beth’s and Speedo’s responses to instances of compulsory heterosexuality offers a new way of thinking about how selves might be situated in social spaces. Their accounts make it possible for gender and sexuality researchers to consider the possibility that social worlds might be configured in such a way as to protect the integrity of a vulnerable self, rather than to enforce norms upon individuals.

Such a perspective, combined with the ethnographic depth and the analytic rigour of grammatical analysis, offers gender and sexuality researchers new ways of understanding the complexities of heteronormativity and the possibility of imagining how these might be transformed.


  • 1 The name of this university, and the names of all participants are pseudonyms.
  • 2 The transcripts follow my modified form of the conventions used for conversation analysis (CA):

Overlapping speech Underlining Emphasis

CAPITALS Loud speech

  • (0.4) Length of a pause in seconds
  • (.) Pause less than one-tenth of a second
  • ((laughs)) Transcriber’s descriptions or comments, contextual infonnation ((...)) Words or lines omitted
  • () Indecipherable
  • (word) Transcriber’s best guess at what was said

sto::p Colons indicate elongation of a sound (number of colons corresponds to length

of elongation) hhh Out-breaths

.hhh In-breaths

(as with colons, number of h’s corresponds to length of out-breaths or in-breaths) , Weak, ‘continuing’ intonation ? Rising, ‘questioning’ intonation . Falling intonation

Further reading

Clark, J. (2012) Language, sex and social structure: analysing discourses of sexuality. Basingstoke: Palgrave.

A detailed explanation of how I conducted the hockey-team ethnography can be found here. The book describes the structuring principles of the heteronormative practices of the women’s hockey team.

Clark, J. (2016) Selves, bodies and the grammar of social worlds: reimagining social change. London: Palgrave.

This book contains more background on the ethnographic study of Beth, Maryam, and Andrew that I describe in this chapter. It offers insights into how to draw upon ethnographic and discourse analytic research to reimagine social structure.

Livia, A. and Hall, K. (1997) Queerly phrased: language, gender, and sexuality. New York: Oxford University Press.

This edited collection offers a wide range of contexts in which questions of identity, community, selfhood, and normativity are explored through the lens of language, gender, and sexuality. Many of the chapters in this volume draw upon etlmographic research.

Motschenbacher, H. and Stegu, M. (2013) ‘Queer linguistic approaches to discourse’, Discourse & Society, 24, pp. 519-535.

This article outlines the basic principles of Queer theory and reviews its applications in linguistic and discourse-oriented research.

O’Reilly, K. (2009). Key concepts in ethnography. London: Sage.

This text offers a concise and critical explanation of issues and terminology associated with etlmographic research.

Related topics

Language, gender and sexuality: reflections on the field’s ongoing critical engagement with the sociopolitical landscape; applying queer theory to language, gender and sexuality research in schools; identity construction in gendered workplaces; using communities of practice and ethnography to answer sociolinguistic questions; gender and sexuality normativities.


Angrosino, M. (2007) Doing etlmographic and obseivational research. Los Angeles: SAGE.

Barrett, R. (1999) ‘Indexing polyphonous identity in the speech of African-American drag queens’, in Bucholtz, M., Liang, A. C., and Sutton, L.A (eds) Reinventing identities: the gendered self in discourse. Oxford: Oxford University Press, pp. 313-330.

Bourdieu, P. and Waquant, L. (1992) An invitation to reflexive sociology’. Cambridge: Polity Press. Bucholtz, M. (1999) “‘Why be normal?”: language and identity practices in a community of nerd girls’, Language in Society, 28, pp. 203-223.

Butler, J. (2006) Gender trouble: feminism and the subversion of identity. New York: Routledge. Clark, J. (2012) Language, sex and social structure: analysing discourses of sexuality. Basingstoke: Palgrave.

Clark, J. (2013) ‘“Maybe she just hasn’t matured yet”: politeness, gate-keeping and the maintenance of status quo in a community of practice’, Journal of Politeness Research, 9, pp. 211-237.

Clark, J. (2016) Selves, bodies and the grammar of social worlds: reimagining social change. London: Palgrave.

Eckert, P. (2000) Linguistic variation as social practice. Oxford: Blackwell.

Eckert, P. (2011) ‘Language and power in the preadolescent heterosexual market’, American Speech, 85, pp. 85-97.

Eckert, P. and McConnell-Ginet, S. (1992) ‘Think practically and look locally: language and gender as community-based practice’, Annual Review of Anthropology’, 21, pp. 461-488.

Gold, R. L. (1958) ‘Roles in sociological field observations', Social Forces, 36, pp. 217-23.

Hall, K. (2013) ‘“It’s a hijra!” queer linguistics revisited’, Discourse & Society, 24, pp. 634-42.

Halliday, M. A. K. (2014) Introduction to functional grammar. London: Routledge.

Hammersley, M. and Atkinson, P. (2007) Ethnography: principles in practice. London: Routledge.

Jones, L. (2012) Dyke/girl: language and identities in a lesbian group. Basingstoke: Palgrave.

Jones, L. (2014) “‘Dolls or teddies?”: constructing lesbian identity through community-specific practice’, Journal of Language and Sexuality, 3, pp. 161-190.

Jones, L. (2016) ‘“If a Muslim says ‘homo’, nothing gets done”: racist discourse and in-group identity construction in an LGBT youth group’, Language in Society, 45, pp. 113-133.

Roller, V. (2015) ‘The subversive potential of queer pornography: a systemic-functional analysis of a written online text’, Journal of Language and Sexuality, 4, pp. 254-271.

Moore, E. (2006) “‘You tell all the stories”: using narrative to understand hierarchy in a community of practice’, Journal of Sociolinguistics, 10, pp. 611-640.

Morrish, E. (1997) “‘Falling short of God’s ideal”: public discourse about lesbians and gays’, in Livia, A. and Hall, K. (eds) Queerly phrased: language, gender and sexuality. Oxford: Oxford University Press.

Motschenbacher, H. and Stegu, M. (2013) ‘Queer linguistic approaches to discourse’, Discourse & Society’, 24, pp. 519-535.

O’Reilly, K. (2009) Key concepts in ethnography. London: SAGE.

Peterson, D. J. (2010) ‘The “basis for a just, free, and stable society”: institutional homophobia and governance at the family research council’, Gender and Language, 4, pp. 257-286.

Rich, A. (1980) ‘Compulsory heterosexuality and lesbian existence’, Signs, 5, pp. 631-660.

Sauntson, H. and Morrish, L. (2012) ‘How gay is football this year?: identity and intersubjectivity in a women’s sports team’, Journal of Language and Sexuality, 1, pp. 151-178.

Seidman, S. (2009) ‘Critique of compulsory heterosexuality’, Sexuality’ Research & Social Policy, 6,

pp. 18-28.

Wittig, M. (1992) The straight mind and other essays. Boston: Beacon Press.


Anthropological discourse analysis and the social ordering of gender


Susan U. Philips


In this chapter I explain how I carried out anthropological discourse analysis of gender ideology in the Polynesian Kingdom of Tonga, a tiny nation in the South Pacific. I show how anthropological discourse analysis examines the relationship of forms of talk to their position within the broader social order, as well as to their immediate contexts of speech production. The ability to provide such analysis depends on a methodology that records not just one form of talk, but multiple forms of talk. When this approach is applied to the study of gender ideology, it reveals how ideas about women vary systematically across social domains. Anthropologists are also interested in how the same gender ideology occurs over and over again in repetitions of a single form of talk, or in the reproduction of culture in discourse. In this respect anthropological discourse analysis complements contemporary approaches to the study of gender and language that stress the processual, emergent, and contingent orienting to gender in face-to-face interaction.

In the discussion to follow I first explain how diversity in gender ideology came to be an issue in feminist cultural anthropology. I then focus attention on the ethnographic methodology used by anthropologists to study gender. This includes the role of recording units of interaction or discourse units within that ethnographic methodology. Finally, after a description of the design of my research project in Tonga, I provide an account of gender ideology in Tonga. I contrast two published microanalyses of gender ideology from two recorded contexts: courtroom cases of bad language and the singing, and comment on singing, in a workgroup where women were putting pieces of bark cloth together and decorating the larger piece they were making. The main purpose of this comparison is to show how different kinds of analysis can be done using different parts of a data base. The possibilities for analysis depend on how the larger project was designed and on aspects of the original data collection and processing of data in the field.

< Prev   CONTENTS   Source   Next >