Gender and sexuality normativities Using conversation analysis to investigate heteronormativity and cisnormativity in interaction

Stina Ericsson

Introduction

Research informed by queer theory assumes ‘normality’ as the object of investigation, taking a critical approach to gender and sexuality normativities (e.g. Milani 2014: 261). Thus, rather than seeing gender and sexuality as involving distinctions between ‘femininity’ and ‘masculinity’, or ‘homosexuality’ and ‘heterosexuality’, a queer theoretical stance sees the very creation of such binaries and dichotomies, and the different values and connotations being attached to various genders and sexualities, as that which is to be explained. Taking a queer approach, this chapter explores the use of conversation analysis (CA) for investigating the (re)production of normative sexualities and genders in everyday spoken interactions. Everyday interactions are interesting in this regard because they concern both individual actions and societal discourses, and everyday norms may be conveyed in such unremarkable ways that they are almost invisible - turning research into exciting detective work.

CA is informed by Garfinkel's (1967) ethnomethodological approach, where the focus is participants’ own understandings, that is, how participants themselves make sense of then- worlds, through social actions. Theoretically, the chapter takes a queer critical approach through the notions of ‘heteronormativity’ and ‘cisnormativity’. Heteronormativity refers to the ‘defining, normalising power of the heterosexual assumption, which marginalises other- sexualities at best, and invalidates them at worse' (Weeks 2010: 92). Heterosexuality is here taken as a sexuality which is specific to a particular time and place. Cisnormativity concerns the normalising power of the cisgender assumption, that is, the idea that gender is coherent over time, in perception, and through actions and identities. Just as for heterosexuality, I take cisgender to be temporally and spatially specific, and as a notion that needs to be examined critically, rather than taken for granted or essentialised.

I begin this chapter by outlining the theoretical framework that underpins the approach taken here - a framework which includes heteronormativity and cisnormativity, and which places these two notions in relation to each other and to other normativities. Next, I introduce relevant methodological tools and assumptions that underpin CA. I then illustrate the theory and methodology by applying them to empirical examples of heteronormativity and cisnormativity in a case study of recorded interactions, where children and parents talk about families and relationships. The chapter ends with a concluding discussion, together with future directions and further reading.

Theoretical framework: intersecting normativities

As a way of specifying a queer critical approach to binaries and gender and sexuality normativities, this chapter adopts the genderism model of Hornscheidt (2012, 2015). This is a framework that outlines different discriminatory structures which are related to gender, including how they interact with and depend on each other. In the version of the model that I adopt here, genderism has six different realisations: categorical gendering, binary gendering, cisgendering, androgendering, heterogendering, and reprogendering. Of these, I will focus on heterogendering and cisgendering in the case study below.

'Categorical gendering’ is the idea that all human beings are gendered. All of the other realisations of genderism are based on this idea, and it represents a powerful imperative of seeing people, properties, behaviour, etc. as gendered. Assumptions of more genders than two, or a gender system that incorporates genders that are not ‘female’ and ‘male’, all rely on the idea that gender exists as a category. 'Binary gendering' means the assumption of two, and only two, mutually exclusive gender categories - ‘female’ and ‘male’. This idea permeates dominant thinking in many contemporary societies around the globe, but balanced against both historical and contemporary non-binary gendering, a critical stance towards binaries is necessary in language, gender, and sexuality research. ‘Cisgendering’ corresponds to the notion of cisnormativity that I introduced above, and it ‘constructs women and men as coherent, natural genderings which are given at birth and which remain constant throughout a person’s life’ (Hornscheidt 2015: 37). The coherence and consistency constraints involved in cisgendering marginalise certain trans bodies and experiences, and by presupposing binary gendering, and thereby ‘continuously affirming it' (Hornscheidt 2015: 37), cisgendering marginalises intersex and various non-binary identities and experiences.

‘Androgendering’ is the ‘universalisation and the norm-setting of men/а male norm’ (Hornscheidt 2015: 32), that is, it corresponds to the notion of ‘sexism’. ‘ Heterogendering' is Hornscheidt’s term for ‘heteronormativity’, which renders heterosexuality as the norm, marginalising other sexualities. A powerful strategy of heterogendering is couple normativ- ity, which is ‘the normalisation of a couple as the standard and desired form of living in Western societies’ (Hornscheidt 2015: 34-35). Finally, ‘reprogendering’ means that reproduction is viewed as an essential part of a heterosexual couple’s experience.

These different realisations of genderism interact in different ways. For instance, reprogendering has different effects depending on whether someone is privileged or discriminated against by androgendering. This may mean that people presenting as female are assumed to be ‘(potential) mothers and/or daughters’ (Hornscheidt 2015: 35) in a way that people presenting as male are not correspondingly assumed to be (potential) fathers or sons.

Having briefly introduced this theoretical framework of genderism, I next turn to the CA method employed in this chapter.

Method: conversation analysis

Queer investigations of normativities can be approached using different kinds of material and methods. The focus in this chapter is on social actions in everyday settings - how people produce, reproduce, and transform gender and sexuality normativities through mundane actions. That is, how people through the routine and unremarkable things that they do in their everyday lives, also create, adhere to, or change norms that govern gender and sexuality. For instance, referring to one’s partner while booking a dentist’s appointment may challenge or reinforce heteronormativity.

For the analysis of such social actions in everyday settings, CA proves a highly valuable tool, providing methods for both data collection and data analysis, enabling fine-grained investigation of human actions at the micro level. While the focus of this chapter is on methods for analysis, data collection methods are vital to any CA enterprise. In brief, CA scholars use audio and video recordings of actual situated activities - which contrast with data collection methods such as interviews and experiments - to gain insight into how participants themselves make sense of their social worlds. As part of this enterprise, CA methodology involves the transcription of recorded audio or video data in ways that capture sufficient information regarding the temporal and sequential unfolding of the interaction. However, it is important to remember that the primary data are always the recordings, not the transcriptions, and the researcher should always return to the primary data. CA involves listening to or watching recorded data in tandem with transcriptions several times, giving an increasing understanding of the interaction, leading to the identification of interesting phenomena and patterns.

CA methodology has been developed in tandem with theoretical assumptions about the organisation of human interaction and social life. In fact, the approach takes social action as fundamental, rather than language. A central concern is sequential organisation, referring to the way in which social actions are always carried out in the context of preceding actions or turns of talk. This means, for instance, that the meaning of a piece of talk can only be recognised through its sequential position, and not by looking at that piece of talk in isolation. Relatedly, C A work has shown how meaning in interaction is co-constructed, incrementally, by participants. Thus, any analysis of gender and sexuality norms in interaction will need to consider the actions that the expression of these norms perform, in specific contexts of other speakers and preceding talk.

One analytical consequence of CA's ethnomethodological basis, is the requirement that the analysis always be grounded in what occurs in the interaction. That is, in terms of gender and sexuality studies, the analysis cannot freely use gender and sexuality categories to interpret the interaction, but instead has to show how participants themselves make gender and sexuality relevant. For instance, knowing that a participant self-identifies as ‘asexual’ is not enough to infer that the participant makes use of ‘asexual speech’. At a general level, this is the issue of just what counts as context (Schegloff 1992), which has been seen as both a strength and a weakness of CA. As a strength, it means that CA observations have a solid scientific basis in the data at hand, staying clear of unwarranted speculations. As a weakness, critics argue that relevant observations may be overlooked because wider contextual knowledge is not made use of. This issue also turns on the CA notion of ‘relevance for participants’ or ‘participants’ orientations’, that is, the analytical imperative of demonstrating what participants themselves make relevant in interaction. These issues of context and participants’ orientations have generated productive methodological debates in language, gender, and sexuality research (e.g. Kitzinger 2008; Stokoe and Smithson 2001. See also Schegloffs debates with Billig in 1999 - all contributions from both Schegloff and Billig are freely available from the Schegloff Publications Archive online - and the chapter on feminist conversation analysis in this volume).

Two methodological points will be made here regarding the use of ethnomethodologi- cally grounded CA for the analysis of gender and sexuality. First, the analysis must indeed always be grounded in participants’ own actions and interpretations. Without this, analysis runs the risk of 'starting] out by "knowing” the [sexual] identities whose very constitution ought to have been precisely the issue under investigation' (Kulick 2000: 265) or of, for instance, reinforcing gender dichotomies. That is, the analysis runs the risk of yielding uninteresting results or of misinterpreting or giving a skewed picture of the data.

Second, making sense of just what participants achieve in any given interaction may involve making use of different kinds of analytical resources. One way of doing this is by seeing wider knowledge of gender and sexuality as another layer that is added to the analysis, such as the ‘post-analytic’ application of the term ‘heteronormativity’ to interactional patterns found in data (Kitzinger 2008: 202-203). The argument here is that to say something that is more specific and interesting from a gender and sexuality point of view, ‘one must [sometimes] go beyond describing data in participants’ own terms’ (Stokoe and Smithson 2001: 232). It should be noted in this regard that while language, gender, and sexuality studies, including queer theoretically informed studies, may use CA in combination with various other methods and theories, other CA scholars may not take favourably to this (e.g. Kitzinger’s 2008 debate with Wowk and the Schegloff-Billig 1999 debate in the Schegloff Publications Archive). In the case study below, I will make use of Hornscheidt’s (2012, 2015) theoretical model of genderism as a way of adding another layer to the analysis, thus going beyond describing data in participants’ own terms, in order to further show how the data is interesting from a gender and sexuality perspective.

A second way of making use of different kinds of analytical resources in order to make sense of what participants do in interaction, is by considering the multiple things to which participants may be orienting (Kitzinger 2008), and calling for more precision regarding just what is an orientation and what is not (Wilkinson and Kitzinger 2014: 156). For instance, participants in a medical help-line conversation orienting to requesting a home visit by the doctor, may also be shown to be orienting to a cultural understanding of couples living together, e.g. through a doctor asking a husband who’s calling on behalf of his wife, ‘where do you live?’ and not ‘where does she live?’ (Kitzinger 2008: 200). When it comes to the queer enterprise of investigating normativities, this issue of context and participants’ orientations is highly pertinent, since norms in many ways are ‘invisible’ and may therefore, at first glance, be expected to be something to which participants do not orient. However, in the analysis of empirical examples below, we will see just how an ethnomethodological approach can reveal participants’ orientations to normativities.

Given that gender and sexuality research brings certain theoretical perspectives to the analysis, an important critique of CA’s tenet of approaching data without any preconceptions, concerns the background knowledge that analysts nevertheless bring to the analysis. For instance, Stokoe and Smithson (2001: 200) argue that ‘culture and common-sense knowledge, of both members and analysts, are largely unacknowledged and unexplicated resources in CA’. Thus, for language and gender scholars using CA methodology, it is pivotal to be explicit about background assumptions and theoretical knowledge. In this, self- reflexivity is an important tool, just as a critical stance and self-reflexivity in relation to theory is important for any language and gender research. In practice, the gender and sexuality scholar that uses CA can employ such techniques as being open to what data shows and collecting also cases that are troubling for one's theoretical perspectives, perhaps even specifically looking for such cases; working with one’s data in data sessions with different groups of people, both within and outside of academia (the latter may require less technical transcriptions); spelling out and critically examining one’s own assumptions and how one is variously privileged or discriminated by different power structures.

Turning now more concretely to the methodology of CA, this basically involves identifying and describing patterns in participants’ actions in interaction. In accordance with ethnomethodology, analytical proof is found in the data itself: evidence for identified patterns and what participants’ actions do, is located in what participants themselves do and how they respond to each other’s actions. From this, two kinds of scientific proof method can be formulated: the identification of cases that follow the pattern and cases that deviate (Sidnell 2013). For instance, say that we wish to establish whether a question followed by an answer constitutes an ‘adjacency pair', that is, whether an answer is in general interac- tionally expected subsequent to a question. Evidence indicating that a question-answer is indeed an adjacency pair, can be found in ‘the great many cases’ (Sidnell 2013: 80) of question-answer examples in data (cases that follow the pattern). Evidence can also be found in examples where a question is not followed by an answer, but where either the questioner or the non-answerer shows that an answer should have followed (deviant cases). Examples of the latter that Sidnell mentions include the recipient of the question apologising for not giving an answer, or the questioner asking a follow-up question. Sidnell (2013: 80) argues that deviant cases ‘often provide the strongest evidence for the analysis because it is here that we see participants’ own orientations to the normative structures most clearly’.

Concluding from the above, the method to be illustrated in this chapter involves CA-informed ethnomethodological analysis of interactional data, to identify what sequentiality and participants’ orientations reveal about normative structures, where these normative structures here concern genders and sexualities. The next sections are concerned with a case study of gender and sexuality normativities in a data set of child-parent conversations.

Case study: normativities in interaction

The data that I use to illustrate the theory and method in this chapter, are drawn from the research project ‘Daddy, daddy, child: linguistic negotiation of family, parenthood and relations in conversations between children and adults’ (Boyd and Ericsson 2017; Ericsson 2018). This Swedish project investigated children’s and parents’ conversations about families, living arrangements, love, and marriage, specifically focusing on gender and sexuality normativities. 24 children, mainly aged 5 to 8 years, participated in the project together with their parents. Participating families included single mothers by choice through insemina- tion/IVF (in vitro fertilisation), same-gender and different-gender parental couples, binary and non-binary gender presentations and identities, parents cohabiting or not, as well as urban-rural and geographical variation across Sweden.

Data were collected using a purpose-designed tablet application (‘app’), which was developed as part of the project. With the help of an interactive character called Moi, who speaks using utterances recorded by a human child, the app asks questions of the children which they are then required to think about and discuss with a parent. The app records the child- parent conversations, both in the form of audio and through logging screen events, such as what images the users select. The app was designed using a norm-critical perspective, which involves taking a critical stance to norms in the sense of explicitly showing or challenging norms which may otherwise typically remain hidden and taken for granted. In the app such an approach was used to probe and challenge children’s views of gender and sexuality, for instance through app characters with ambiguous gender presentations. The families used the app in their homes whenever they wanted, without the presence of a researcher.

Traditional CA research places great emphasis on the recording of naturally occurring activities, that is, activities that would have occurred even without the recording or the research study. This is motivated by CA's interest in people’s ordinary lives and actions. CA studies of naturally occurring activities have been successfully used for language, gender, and sexuality research (e.g. Kitzinger 2005; Speer and Stokoe 2011). The use of the app for generating the data in the present chapter meant stepping away from the requirement of naturally occurring activities, in that these interactions only occurred because of the research project. However, in several important ways, the participants in these interactions draw on the same resources as in any other conversational setting, such as turn-taking. They also access discourses on normative genders and sexualities. As Cameron et al. (1993) argue, research data - in their case interaction between researcher and researched - constitute one form of 'normal communication’ that ’provide[s] important insights into the way social relations and identities are constructed through interaction’ (1993: 87, emphasis in original). As a general remark, it is important to be aware of the specific conditions of one’s data collection, and to critically examine the consequences of data collection methods for the data at hand, and for the analysis. With regard to the data used for the case study in this chapter, it may for instance be important to investigate how parents and children cany out their conversations for an audience (the researcher) rather than as part of their own daily activities. It is also important to carefully consider one’s research design before the collection of any data, in light of the kinds of research questions one wishes to address. Here, the present research project had a focus on ideological norms rather than naturally occurring social actions, which helped motivate the research design using the app.

All child-parent interactions with the app were transcribed in full in the project, but using a rough transcription mainly aimed at capturing the content of participants’ contributions. Parts of the data were then transcribed in more detail, for specific smdies and publications. In practice, the transcription of one’s recorded material is a matter of resources, as transcription is highly time-consuming. All families spoke Swedish with the app, with the exception of one family who mainly used English.

I will now turn to a ‘hands-on’ illustration of the use of CA, first analysing sexuality nor- mativities in the data, and then cisgender normativities. These two sections are structured in the same way: I first consider evidence from cases that follow the pattern, then evidence from deviant cases, and finally cases that challenge the normativities. This is typically not the way such an analysis would be presented in a research paper, but it is used here to explicate the method as clearly as possible. I also want to emphasise that it is not a matter of a complete analysis, but rather intended to illustrate the application of theoxy and method. Names of all participants and the people they talk about, as well as some other personal details, have been anonymised in the transcriptions.

Analysing sexuality normativities

This section investigates sexuality normativities in the data, addressing the overarching questions of which sexuality norms are produced by the participants, and how these norms are interactionally achieved. The analysis of sexuality normativities will here focus on participants’ replies to Moi’s question ‘Do you know someone who’s in love?’, where Moi is the fictional app character. An inductive, bottom-up analysis is used to make claims about the particular data at hand.

Translations into English are provided on a separate line, and a transcription key can be found at the end of the chapter.

Consider the following example, where Oskar and his parent talk to Moi:

Example 1. ‘Henry and Edith’

593 Moi: vet du nan som e kar

do you know someone who's in love

594 Osk: jao ((smackljud)) Henry och Edith dom e kara

yieah ((lip smack)) Henry and Edith they’re in love

595 Par: m hm:: ha dom sagt de eller

m hm:: have they said so or

596 Osk: Henry ha sagt de ti mej

Нету has said so to me

597 Par: mhm::

In this example, Oskar answers the yes/no question of whether he knows someone who’s in love - the Swedish nan (‘someone’, line 593) formally requesting just one person - by giving an example of two people he knows, ‘Henry and Edith’ (line 594). We learn elsewhere in the conversation that they are friends of Oskar's from school.

In his reply, Oskar describes a couple: Henry and Edith. The mention of a couple indicates an understanding of being in love as a relation that takes an object (cf. if Oskar had replied ‘yieah Henry he’s in love’), and only one object (cf. ‘yieah Hemy, Edith, and Fred they're in love’). Further, the couple that Oskar names is a different-gender couple. What Oskar does, then, I argue, can be seen as an instance of what Kitzinger (2005) describes as the production of heterosexual couples, or as an example of Hornscheidt’s (2015) couple normativity as a strategy of heterogendering.

In line 595, the parent acknowledges Oskar’s reply through ‘m hm::,’ and then poses a question. The question concerns the type of evidence that Oskar has for his claim, but does not question the idea of Henry and Edith as a couple, or of different-gender couples, as such. As another example, consider Stella’s reply:

Example 2. ‘You and mummy’

052 Moi: vet du nan som e kar

do you know someone who's in love

053 Ste: manga ©du a mamma© o::ch assa de=e sa inanga

many ©you ’n ’mummy’© a::ndwell it’s so many

054 Par: okej

okay

055 Ste: Jessica a Davi:::d o::ch

Jessica 'n’Davi:::dav.nd [removed talk about how Stella sits on her chair]

058 Ste: (a sen) Eimnas mamma a pappa a Susans mamma a pappa
(’n’then) Emma’s mum ’n’dad ’n’Susan’s mum ’n’dad
059 Par: e de bara mammor a pappor som e kara

is it just mummies ’n ’daddies who ’re in love

060 Ste: och barn kan me vara kara

and children can also be in love

061 Par: okej

okay

Stella gives several different couples as part of her response (lines 53, 55, 58). Her response is acknowledged by the parent, explicitly through ‘okay’ (line 54), and implicitly through a follow-up question (line 59). All couples are different-gender couples (the ‘you’ in line 53 referring to Stella’s pappa (daddy)).

The parent's question in line 59 concerns whether anyone else can be in love, apart from ‘mummies ’n' daddies’. Stella understands this as a generational issue, and states that children can also be in love, which the father agrees to, possibly indicating that Stella gave an expected reply (lines 60-61).

Regarding participants’ orientations, as Kitzinger’s (2008) analysis of orientation to cultural norms shows, such an analysis relies on inferences and requires careful and critical analysis of just what it is that participants treat as common knowledge. For the two brief examples considered here, it can be argued that Oskar, Stella, and their parents orient to a cultural norm of different-gender couples as part of their understanding of being in love. Also, participants’ understanding is co-constructed by the children and parents, through Oskar and Stella describing the couples in their replies to Moi’s questions, and the parents acknowledging these replies and producing follow-up questions.

In the data, there are ‘great many cases’ (Sidnell 2013: 80) of different-gender couples being described in response to the question of knowing someone in love, similar to the two examples which we have seen above. This is one type of argument for seeing such examples as cases that follow a pattern.

Summing up, cases that follow the pattern of sexuality normativities in the data involve participants producing and treating different-gender couples and living arrangements as the norm. We have seen two examples in some detail here, and the analysis as a whole would build upon many other, and different kinds of, examples in the data.

Evidence from deviant cases

The two examples above with Oskar and Stella, respectively, can be contrasted with the following:

Example 3. ‘Kasper and Vincent’

247 Moi: vet du nan som e kar

do you know someone who's in love

248 Lud: eh:: tva

eh:: two

249 Par: vilka da

who

250 Lud: Kasper a Vincent

Kasper 'n ’ Vincent

251 Par: e dom kara i varann

are they in love with each other

252 Par: naha vilka e dom kara i da

no okay who are they in Jove with then

253 Lud: nara i en annan klass

some in the other class

254 Par: jaha

okay

255 Par: e de tjejer eller

are they girls or

256 Lud: m
257 Par: ha

okay

Highly reminiscent of the examples that follow the pattern - such as Examples 1 and 2 above - Ludwig here uses a conjunctive structure, ‘Kasper 'n' Vincent’ (line 250), in reply to the question of knowing someone in love. However, unlike the parents’ next contributions in those two other examples, the subsequent contribution by Ludwig’s parent questions the very coupledom of Kasper and Vincent (line 251). From the ensuing contributions by Ludwig and his parent, lines 252-257, it is clear that Kasper and Vincent are not to be understood as a couple in love with each other (presumably Ludwig shakes his head or gives some other non-verbal response, between lines 251 and 252).

What makes this a deviant case, then, is partly seen from sequential evidence, in the parent’s response to the child’s answer, and the participants' mutual meaning construction of ‘Kasper ‘n’ Vincent’, despite formal appearances, as not a couple. Additionally, one thing that differs between Ludwig’s conjunctive structure and those used by Oskar and Stella above, is the inclusion of two people of the same gender in Ludwig’s utterance. One of the ways in which heterogendering works is by making sexualities other than heterosexuality invisible. It may be this mechanism that makes the parent question a couple reading here. As possible further support of sexuality and gender being at issue here, gender is explicitly brought up by the parent in line 255, and Ludwig (line 256) confirms that Kasper and Vincent are in love with girls.

Another sequential deviation in Example 3 comes in the position immediately following Moi’s question (here line 248). As can be seen from Examples 1 and 2, this is where an answer is given to Moi’s question, enumerating exemplifying couples. Ludwig’s response differs from this, giving instead the number of people that he knows are in love, ‘two’. This answer may also be what contributes to the parent’s inference that ‘Kasper ‘n’ Vincent’ are to be understood as two people in love with others and not each other.

In this way, Examples 1-3 have given an illustration of how a collection of examples that are similar in some way - in this case through being the utterances that follow a specific question - can be put together and examined, revealing normative structures through cases that follow and deviate from an identified pattern, respectively. There are also several other kinds of examples in the data that can be examined for what they reveal about heterogendering through deviant cases. One type of example involves participants expressing breaches to their expectations, such as Ebba exclaiming ‘WHAT can you get married to a girl’ when her parent probes who she wants to get married to. Another type of example is discriminatory or negative utterances, such as Hannes expressing that two male-presenting characters in the app cannot get married or that a boy going out with a boy is ‘odd’.

Challenges

Based on evidence from cases that follow and deviate from a pattern, it can thus be argued that heterogendering is prevalent in the data. However, there are also other kinds of examples, which complicate the picture. Consider Example 4:

Example 4. ‘Agnes and Vera’

201 Moi: vet du nan som e kar

do you know someone who's in love

202 Mor: Agnes a Vera

Agnes 'n’Vera

203 Par: du fa prata ordentligt sa dom hor

you need to speak properly so they can hear

This example follows the pattern of the child - here Morgan - describing a couple in the second-pair part (Stivers 2013) of the question-answer adjacency pair (line 202). It can also be said to follow the pattern of the parent not questioning the couple as such (line 203), meaning that Morgan and the parent interactionally co-construct the couple as unremarkable. The couple here is a same-gender couple, involving Morgan's sister Agnes. In other parts of the data, Agnes herself also mentions Vera as the person that she is in love with, and this is also interactionally conveyed as something ordinary. Similarly, Siri and her father Jesper jointly construe 'daddy Peter' as the person that Jesper is in love with, and Siri also mentions ‘Vilgot’s mummies’, among other couples, as being in love.

What do such examples mean for heterogendering? Would it be inaccurate to connect empirical evidence to the theoretical notion of heterogendering here? Well, yes and no. No, because of the great number of normatively conveyed other-gender assumptions that permeate the data, attesting to the pervasiveness of heterogendering. Here, actual numbers of same-gender and other-gender couples in the data set as a whole could be stated to support the argument. Yes, because it would not give the full picture of the data. Considering Examples 1-4 again, in light of the whole material, what comes across even more strongly than heterogendering is couple normativity, irrespective of sexuality or gender constellations. That is, in the data as a whole (based on many more cases than Examples 1-4), the couple as a way of organising love, sexuality, living arrangements, etc., is superordinate to, and can thus be theoretically partly detached from, heteronormativity. Then, given the interdependence between different realisations of genderism as shown by Hornscheidt (2012, 2015), couple normativity can be expected to interact with heteronormativity in different ways.

As a brief conclusion of the analysis of sexuality normativities, my aim here has been to show how the identification of cases following and deviating from a pattern can be useful analytical tools, and also how micro-level actions by participants can be connected to macrolevel language, gender, and sexuality theories. At the same time, my aim has also been to show how a critical approach, investigating challenges to one’s analysis, is necessary.

Analysing cisgender normativities

Having shown how an analysis of sexuality normativities can be carried out, I will here consider more briefly an analysis of cisgender normativities, addressing the overarching questions of which cisgender norms are produced by the participants, and how these norms are interactionally achieved.

In the example below, Siri and her friend Felix are doing the app activity of giving Moi a family, by choosing from a large set of characters that are shown visually in the app (but otherwise not verbally described in any way by the app). Having chosen a set of characters, they talk about the ages of these fictional family members:

Example 5. He and she

205 Moi: hu gamla a vi

how old are we

206 Sir: [m::: ]
207 Fel: [eh: han] ska va kanske:

[eh: he] should be maybe:

208 Sir: ksch arton

ksch eighteen

209 Fel: na@tju- [tjuge::]

no @ twe- [twenty::]

210 Sir: [@ @ ]
211 Sir: [a]

[yeah]

212 Fel: [nej] trett-tretti:uie

[no] thir- thirty. nine

213 Sir: a trettinie [a hon]

yeah thirtynine [’n ’she]

214 Fel: [a hon] ska va trettiatta

[’n ’she] should be thirtyeight

Siri's and Felix’s use of gendered pronouns to refer to the characters (lines 207, 213, 214) attest to their assumption of coherent genders as something that can be read off their perception of the visually presented characters. Coherence over time can be seen through participants’ repeated use of the same pronoun for the same character on different occasions when using the app (not shown in the extract). In this way, Siri, Felix, and other participants can be seen to be orienting to an unnamed and taken-for-granted norm (cf. Kitzinger 2008) of coherent gender.

The prevalence of gendering, and thereby the continuous reproduction of gender as a relevant category, is further seen through numerous examples in the data (not shown in this chapter). For instance, participants use gendered pronouns, gendered proper names, and gendered family roles such as ‘mummy’ and ‘daddy’ in their utterances. These also reproduce a binarily gendered worldview. Thus, categorical and binary gendering are both firmly attested to in the data. Cisgendering presupposes both categorical and binary gendering, and further involves the assumption of gender coherence over time, in perception, and through actions and identities.

It might be argued that pronoun gendering is unavoidable in languages like Swedish and English - as opposed to e.g. languages like Finnish where the single third-person pronoun han does not encode gender - that is, that the language system itself constructs categorical, binaiy, and cisgendering. However, language use always involves choice, and analysing the linguistic choices that language users make is precisely what underlies discourse and interaction analysis. Alternatives for Siri and Felix here could have been to use referring expressions like ‘this/that one’.

In interaction with a parent, Matilda asks ‘but why does she wear a tie’, and ‘why doesn’t she wear a dress’, about a bride character in the app. These two examples involve cisgender incoherence being constructed as a clash between ‘gender assignment’ (‘she’) and ‘gender presentation’ (‘wear a tie’, ‘not wear a dress’), using terms from Zimman (2015). A similar example is Marika asking her little sister Sara ‘why did you say daddy to a girl’, holding Sara accountable for what Marika sees as a clash between the ‘gender role’ of ‘daddy’ and the ‘gender identity’ of being a ‘girl’. ‘Accountability’ (Scott and Lyman 1968) is the idea that it is the unanticipated and non-normative which needs to be explained, and requests for accountability in relation to gender constitute deviant cases which give good evidence for participants’ orientations to normative structures regarding cisgender normativities. Consider next the following exchange, which occurs a few turns after Example 5:

Example 6. She you mean

222 Sir: aHA:Nskava::

'n’HE: should be::

223 Fel: eh hon menar du

eh she you mean

224 Sir: ho[n:]

sh[e:J

225 Fel: [h-] hon ska va

[shj- she should be

226 Sir: typ tre

like three

In this example, Felix initiates a repair sequence in line 223, where ‘repair’ is the CA term for participants’ treatment of problems in interaction which have to do with hearing and understanding, and the like. In Example 6, the repair concerns the gender of the character they are talking about. Felix's utterance conveys a perceived gender incoherence between Siri’s assignment of ‘he’ and the presentation, role, or identity that Felix assumes for the character. Thus, in addition to accountability, repair of assigned gender is another way in which deviant cases regarding gender attest to cisgender normativities.

Challenges

In the following example, Gabi is talking to their parent:

Example 7. Mapa

061 Moi: who are the people in your family
062 Par: can you say
063 Gab: °ma:pa:°
064 Par: say it a little bit louder
065 Gab: u:o[: ]
066 Par: [who] am I
067 Gab: та- тара
068 Par: and who else is in our family

069 Gab: mama
070 Par: and who else
071 Gab: Cam
072 Par: okay

Gabi here uses the term ‘тара’ for their parent (lines 63, 67). This is a term created from the Swedish ‘mamma’ and ‘pappa’, so in one sense it builds upon binary gendering, but in another sense, it challenges it, by naming something outside of the binary. Cisgender being definition- ally restricted to a binary system, ‘тара’ then also challenges cisgendering. Thus, this example constitutes a challenge to the cisgender normativities shown by the other examples above.

In a general way, challenges to cisgendering would involve examples of participants embracing gender incoherence in various ways. Examples of categorical gendering being challenged exist in the data, in the fonn of participants using non-gendered expressions such as ‘that person’ or ‘that/this one’, that is, here participants do not assign gender at all. Such examples can perhaps also be analysed as challenging cisgendering, as cisgendering relies on gender being assigned. An interesting case in this regard is also utterances where ‘he or she’ is used with a specific reference, such as the Moi character. Such a referring expression seems to allow for gender fluidity (while categorical and binary gendering remain intact). However, such examples in the data are typically offset and overruled by the same app characters being gendered on other occasions in the interactions. Examples that more explicitly and firmly embrace gender incoherence and fluidity axe rare in the data, with the use of ‘тара’ in one family as an exception.

Conclusion

This chapter has employed the notions of ‘heteronormativity’ and ‘cisnormativity’ in the enterprise of using ethnomethodologically grounded CA for the analysis of gender and sexuality normativities. Evaluating this approach, there are several distinct advantages to using CA for gender and sexuality research in general and normativities in particular: it shows what people are doing rather than what they believe they, or others, are doing; it can reveal how discourses at the societal macro level are formed, used, and challenged at the interactional everyday micro level; it can make seemingly invisible norms visible and thereby help denaturalise the seemingly natural orders of cisgendering and heterogendering. By using CA, the gender and sexuality scholar can also take full advantage of several decades’ work on interactional practices in other areas. Disadvantages of CA as a method include practical obstacles: transcription is highly time-consuming, and, as with any field work, finding participants and obtaining access may be difficult.

Future directions

In recent years, multimodality and embodiment have received increasing attention in CA research, with studies of gaze, gesture, bodily posture, and movement, in relation to objects and physical space. This can be expected to develop as a highly productive line of investigation also for language, gender, and sexuality research, with detailed micro-level analyses of, for instance, embodied gender expressions and assignments in specific interactional contexts; embodied interactions in sexual encounters; and gender and sexuality normativities conveyed in embodied ways.

An area which has received increasing attention in language, gender, and sexuality research in recent years involves gender and sexuality non-binaries. Here, queer-theoretically informed CA smdies ought to be able to generate significant new knowledge of human genders and sexualities in practice, such as through analyses of how binary norms are upheld and challenged in various ways in different interactional settings.

Transcription conventions

((...)) meta-comments and non-verbal actions

lengthening of preceding sound (...) uncertain transcription

@ laughter, each token marks one pulse

[... speech produced in overlap

°...° quiet voice

interruption

word emphasis

WORD strong emphasis

Gender and sexuality normativities Using conversation analysis to investigate heteronormativity and cisnormativity in interaction

Text Analysis I: Interpretive Analysis, Narrative Analysis, Performance Analysis, and Conversation Analysis

Conversation Analysis

Linguistic approaches: Genre analysis, corpus analysis and conversation analysis

Conversation analysis

Gendered Understandings and Normative Heterosexuality

Teacher Gender and Social Norms

Non-binary approaches to gender and sexuality

Perception of gender and sexuality

Ethnography and the shifting semiotics of gender and sexuality