Abstract
Feminists have long recognised important relationships between language and a gendered social order that disadvantages women. At the establishment of gender and language as a field of academic inquiry, work documented sexism in language—the ways words were used to ignore, narrowly define, or demean women. Using feminist conversation analysis, this article further develops that early work by considering recorded instances of gender and sexism in talk. A broad notion of “gender trouble” was used to identify 50 relevant cases from everyday interactions. Two sexist language issues that were evident in the collection are presented in this article—the derogation of women and participants’ orientations to gender inclusiveness. The analysis contributes to a better understanding of sexism in language by examining how instances of it unfold over turns of talk. The study is discussed with respect to the methodological tensions inherent in feminist conversation analysis.
Feminists have long noted the negative and unequal treatment of women and language. Lakoff (1973) famously noted that women’s secondary place in society is reflected both in the ways women are expected to speak and the ways they are spoken of. Questions about the ways language—in structure, content, and daily usage—reflect and help constitute sexism were foundational for studies in social psychology on gender and language (e.g., Kramer, Thorne, & Henley, 1978). Now, much feminist language research is discourse analytic studying the ways societal structures and systems of meaning produce gendered norms and subjectivities (e.g., Baxter, 2003; Holmes, 2006; Sheriff & Weatherall, 2009; Weatherall & Gallois, 2003). The present study uses feminist conversation analysis to examine sexism as it can be observed in interaction. The focus is on turns and sequences of talk when “gender trouble” arises.
The concept of gender trouble developed in the present article draws its sense from two very different sources. “Gender trouble” was in the title of J. Butler’s (1990) book that broke new ground in feminist theory. Part of Butler’s intellectual project was to find a way “to trouble the gender categories that support gender hierarchy and compulsory heterosexuality” (J. Butler, 1990, p. viii). With respect to the prevalence of the derogatory and demeaning names for women, J. Butler (1997, p. 2) noted they “may appear to fix and paralyse the one it hails, but . . . may also produce and unexpected and enabling response.” Speer (2005) rightfully noted that mapping queer theoretical concepts to observable phenomena in interaction is an important task, which the present study undertakes. Speer was amongst the first to bring together queer theory, discursive psychology, and conversation analysis to study gender identities and heterosexism.
A second sense of gender trouble developed in the present article is as an empirical phenomenon that resembles “repair.” The domain of repair is one of the structural systems, discovered by conversation analysis that is fundamental to the organisation of talk. Repair is a domain that organises how troubles in speaking, hearing, and understanding are addressed (Schegloff, Jefferson, & Sacks, 1977). Practices that can be used to do repair can also be used in the service of other actions—such as doing displays of doubt or disalignment (Schegloff, 1997). The collected cases of gender trouble for this study included repair and repair-like practices.
Feminist conversation analysis is a distinctive discursive approach. It is a controversial methodology (Kitzinger, 2008; Speer, 2012; Weatherall, 2012; Wowk, 2007). On the one hand, its grounded analytic mentality has been criticised as too empirically constrained to be properly critical (Wetherell, 1998; Weatherall, in press). On the other, it is too political in its orientation to be truly conversation analytic (Wowk, 2007). Nevertheless, feminist conversation analysis has produced new insights into key matters, such as how normative heterosexuality is accomplished (Kitzinger, 2005; Land & Kitzinger, 2005) and the ways in which people come out as lesbian, gay, or bisexual (Kitzinger, 2000). It is used in this article to relate emerging moments of gender trouble to feminist concerns about sexism in language.
One of the most well-documented and researched aspects of sexist language is the use of masculine generics or male terms to refer to people in general (e.g., mankind) or the use of masculine pronouns when the gender of the referent is unknown. However, studies of the use of masculine generic language in situ are rare. Stringer and Hopper (1998) found no clear instances of the generic “he” in the conversations they examined. However, they did identify “pseudo generics”—a default assumption that a referent is male although it could be a female. Stringer and Hopper presented cases of the default assumption being ignored, noticed, and corrected in an embedded way. They noted that most research masculine generics was based on written or imagined exemplars and called for further research on their use in talk-in-interaction. Their work showed that examining a phenomenon like masculine generic language as it is used in social interaction will likely produce a unique perspective on the matter.
Another widely recognised aspect of sexism in patterns of language use is the entrenchment of gender stereotypes that often involve a negative valuing of women (Lakoff, 1973; Liddicoat, 2011; Weatherall, 1996; Weatherall, 2002b). Words for women regularly take diminutive forms (e.g., -ess or -ette endings). Others have negative and/or sexual connotations (e.g., madam, mistress). Masculine forms of words tend to have more positive connotations than equivalent feminine words—compare bachelor versus spinster and lord versus lady. Metaphors for women not only sexualise women but also construe women in passive object positions (Weatherall & Walton, 1999). Few, if any, interactional studies document the use of these other aspects of linguistic bias against women.
The present research uses feminist conversation analysis to further investigate sexism in language. It uniquely develops the notion of gender trouble as an empirical phenomenon in talk-in-interaction. The analysis provides an account of sexism in language that is grounded in social interaction.
Data and Method
More than 50 instances of “gender trouble” from a wide range of everyday and institutional contexts, in different varieties of English (American, British, Australian, and New Zealand), were collected. Each case was found serendipitously in an evolving set of recordings of social interaction for teaching and research in discursive psychology and conversation analysis. The cases were gathered over a period of around 7 years.
Gender trouble was broadly conceived to include possible cases of conversation repair where gender was an aspect of the trouble source or the solution, cases where practices for doing repair were used, and instances where gender or sexism became consequential in the unfolding action. Routine gendered references or gendered topics of talk were not included in the collection. In all the cases addressing the gender trouble became, at least for a moment, the relevant business for the parties in the interaction whether or not it was oriented to as sexist.
Analysis
The analysis presents six extracts representing different instances of two sexism-in-language issues: the derogatory use of female reference terms and gender inclusive language.
Negative Use of Female References
The first case shows girl being used as a criticism, albeit in a teasing context. In Extract 1, the target of the criticism—Mike—comments the categorisation is sexist. The interaction is a family discussion where food planning is being done. The possibility of a pasta dish is being raised by Dave.
In Extract 1, Dave, the son of Fran and Mike, describes his father as a bit of a girl for not liking lasagne, which negatively assesses his Dad’s food preferences. In the next turn, Dave’s mother, Fran, produces a second assessment—he is not a pasta person. In her turn, Fran evaluates Mike’s food preferences by excluding him from a general category of persons who do like pasta.
When a speaker produces some object (X) in a turn and a subsequent speaker produces an alternative (Y), the subsequent speaker is doing correcting (Jefferson, 1987). In the above case, Dave categorises Fran as a bit of a girl, which is a sex-specific description. Subsequently, Fran produced an alternative categorisation not a pasta person—an exclusion from a generic collectivity. Fran’s turn is not clearly correcting but it does modify the prior turn. It edits out the sexism inherent in the initial form of the criticism. In contrast, Mike makes the sexism of the form of the criticism explicit by his remark a bit sexist innit (Line 09).
Using the term girl as a negative assessment is a recognised sexist language practice (Weatherall, 2002a). The bias derives from its use to constitute a valued attribute as masculine one and the lack of it as feminine and deficient. For example, to be a girl’s blouse is used to refer to someone who is deficient in risk taking or sporting prowess. The above case shows a particular manifestation of that general practice and possible consequences of its use. The criticism can be modified in a non-sexist way and its sexism can be made explicit.
Extract 2 is a second example where a female reference term is used in a derogatory way. In the following case wife is used negatively. Two friends—Steve and Cameron—are planning an outing to the movies. Cameron raises the possibility of Ben joining them. Steve shows little enthusiasm for that possibility by indicating inviting Ben would also mean Ben’s girlfriend—Betty—joins them.
In talk, there is a preference for a personal name over another kind of descriptor when a speaker supposes a recipient knows a non-present person by name (Schegloff, 1996). Cameron’s suggestion of someone else joining them at the movies refers to the relevant non-present person using his first name, Ben—a word selection that invites Steve to recognise the referent as someone they know and also does simple reference. Steve displays his understanding by responding to the suggestion. It is a disaligning response to Cameron’s proposal—that is, he does not further progress the idea by accepting it as a structural feature that evidences Steve’s less than enthusiastic uptake about Ben coming along with them is the claim to insufficient knowledge (Line 13; Schegloff, 2007b).
A further indication of Steve’s reluctance to include Ben is that it means dragging the wife along as well. The term wife here is an alternative recognitional—a kind of descriptor selected when more than simple reference is being done. Alternative recognitionals are selected to fit with the performance of action (Stivers, 2007). Here, Steve is rejecting the suggestion that Ben join them because Ben comes as part of a couple. Steve accounts for the rejection, in part, by selecting the term dragging, which carries connotations of reluctance. By referring to Ben’s partner as the wife, he not only describes her using an alternative recognitional but categorises her in way that supports his rejection of Cameron’s proposal. As the wife Ben’s partner is a member of a category that would not be welcome company as evidenced by Steve’s own words—and I’m not quite keen on that.
Cameron has trouble understanding the wife, which he shows by initiating a repair that specifies the source of misunderstanding by repeating the words from the prior turn that occurred just before the thing he is having trouble understanding—dragging the what. Steve clarifies using the referent’s personal name, Betty, which shows that he knows and supposes Cameron to know her in that way. That clarification of who he was referring further supports the claim that Steve was using the wife to do something other than simple reference in the previous turn. He could have referred to her simply as Betty since he and Cameron know her by name, but he used an alternative recognitional reference.
Cameron claims knowledge of who Steve means (yeah I know, Line 27). However, he treats the repair solution Betty as discrepant from what he heard—I thought you said wife (Line 27). Smith (2013) established I-thought prefaced utterances routinely mark something in the prior speaker’s utterances as being discrepant with what the current speaker believes they know. Cameron uses that practice here to show he heard wife and to indicate that is discrepant with whom Steve is referring to. Steve confirms that Cameron actually heard right I did (Line 30). The alternative recognitional is a trouble source for Cameron—he cannot understand the wife to refer to Betty whom he knows as Ben’s girlfriend. Cameron then displays a change of state OHH, which is hearable as a realisation of what Steve means, which he positively and possibly ironically assesses as a nice way of putting it. The proterm it presumably refers to the undesirability of having Ben and Betty in their movie outing. The potential offensiveness of referring to Betty as the wife is evident from Steve’s request for Cameron not to repeat it.
The analysis so far has presented two cases where feminine gender categorisations (bituva girl, the wife) and were used to accomplish negative social actions, namely, criticising and excluding. In the next and third case a mistake about the actual sex of the referent occasions a demeaning female reference. The initial error is oriented to as transgressive, which points to the normativity of binary gender categories that was the target of Bulter’s (1990) discussion of the kind of social sanction and taboo that produces gender as a performative accomplishment.
The next extract is from a telephone interaction between BJ and SI—two friends who are both university students. SI is making some distracting noise while on the phone and his father—DD, in the transcript—tells him to shut up. BJ misattributes the source of that instruction to Catherine—SI’s sister. The correction then becomes the interactional business—BJ is moved to provide an account for the mistake—explaining DD sounded like a girl. That gender miscategorisation ultimately results in a negative assessment of the way the father sounds—like an old moaning wife. The analysis describes the sequential and inferential underpinnings leading to that sexist characterisation.
At Lines 04 and 05, BJ requests confirmation that it is Catherine—SI’s sister that is saying shut up. SI responds no, accounting for who it was—his Dad. BJ first registers the new information with an oh—a token that displays a change in cognitive state (Heritage, 1984), followed by one admitting a transgression oops. Steve does not understand the admission (what, Line 10), and makes sense of it by inferring that Ben thought himself at fault for the noise (no its not you its me). However, the mistake that Ben was actually conceding was his assumption about the gender of the speaker. At Line 13, BJ accounts for that mistake by saying he sounded like a girl, which is punctuated by laughter tokens. The laughter particles are post-completion stance markers, which regularly function to modulate the force of an action (Potter & Hepburn, 2010); in this case a misattribution of gender.
SI informs his father of the gender misattribution by making it a more general attribute—Dad you sound like a girl i’m afraid (Lines 16 and 17) and you sound like a girl anyway (Line 19). Teasing regularly involves using a locally relevant activity to invoke some kind of deviant identity (Drew, 1987), which seems to be what Steve is accomplishing here. BJ supports the tease by treating it as laughable (Line 21) and actually (re)voicing the activity implicated in the gender miscategorisation—when he goes shut up. Steve agrees and affiliates with BJ (laughter followed by yeah and right) by upgrading the tease implicated identity to old moaning wife. Although not oriented to by the participants as sexist, the formulation is clearly derogatory—it demeans the value of older married women’s words in a way that has been the subject of feminist critique (Lakoff, 1973; Weatherall, 2002a).
So, in the above extract, BJ misattributed the gender category of SI’s father on the basis of how he sounded when he said shut up. That misattribution became a resource for a tease that miscategorised Dad as a girl. However, the membership category girl with its locally occasioned activity—demanding quiet, is not actually a good fit for a married man with a university student aged son. A feature of membership categorisation devices is they operate on the basis of consistency (Sacks, 1992; also see C. Butler & Weatherall, 2011). The categories old and wife are more consistent with father’s age and marital status than girl. A derogatory construal of making demands or complaining is nagging or moaning, which are attributes that have been historically linked to the description of women’s speech (Lakoff, 1973). Thus, the rather insulting portrayal of SI’s father as an old moaning wife was achievable by the negative attributes associated with the talk of women.
The above extracts show three different ways the derogation of terms for women emerged in interaction—as teasing criticism (Extracts 1 and 3) and to justify exclusion from an outing (Extract 2). In Extract 1 the sexism was made explicit. However, in Extracts 2 and 3 it was not explicitly noted to as sexist. Rather, they were oriented to other concerns. In Extract 2, Steve shows he may have overstepped by referring to Ben’s girlfriend as the wife by saying don’t repeat it will you? In Extract 3, a gender misattribution resulted in teasing the father for sounding like a girl.
The next three extracts are examples where the gender trouble is an orientation to being gender inclusive.
Being Gender Inclusive
Using sex-specific terms to refer to people in general or to a non-specific referent has long been criticised as a sexist practice. Regularly it is women who have been excluded. The non-sexist alternative typically recommended in publishing guidelines or language policies is being gender inclusive. The extracts in this section show examples where speakers correct an exclusive reference to be inclusive. The first two examples are cases taken from public radio where there is an initial non-specific referent followed by a gendered subsequent reference form, which then has a co-class pronoun added. The first example is where the move is to including females and the second is to include males. The third example is taken from a sexuality consciousness-raising group where the gender trouble being addressed is a more contemporary matter—the invisibility of transgender individuals. It is an example of grassroots language activism. Together these cases show participants’ orientations to being neutral or not biased.
The first extract comes from American public broadcasting radio. The interviewer is discussing in general terms what candidates require for political office. The use of a masculine pronoun as a subsequent reference form for candidate is oriented to as a trouble source and the interviewer adds an or-prefaced feminine pronoun.
The interviewer is describing what might make a successful candidate for political office. The term candidate is an initial reference to non-specific person. However, the subsequent reference form the interviewer uses is he. The interviewer displays her orientation that he does not adequately refer to all possible people who could be candidates because she adds or she has got viability. The repetition of has got viability provides a right frame to both the trouble source he and the resultant repair solution that parses as “he or she has got viability.”
The use of masculine pronouns to refer to non-specific persons has long been considered sexist because they exclude and make invisible women. Stringer and Hopper (1998) found the use of true masculine generic terms in a gender-inclusive way was rare. Instead, they noted that when masculine pronouns were used as a subsequent reference to a non-specific person, it was regularly noticed by a next speaker for excluding possible female referents. In the above example, the interviewer both initiates and conducts the repair, resulting in a non-sexist formulation. Nevertheless the initial default assumption is that candidates are male.
The next example is also from a radio interview, but from the British Broadcasting Corporation. It is taken from a show with a focus on women’s issues. Inter-marriage between Shia and Sunni Muslims is being discussed because of the problems associated with it for the parties involved. As in the previous case, an initial reference is made to non-specific persons Shia and Sunni marriages and the subsequent reference form is gender specific. However, in this case that subsequent reference is to women, which obviously excludes men. The interviewer orients to women as a trouble source and repairs the reference to add men.
The subsequent reference form to Shia and Sunni marriages is women (Line 13), which reflects the focus of the show on women’s issues. It is also largely women from mixed marriages who are pressurised to divorce to the selection so the word choice reflects the gender category that is the major concern. However, men as well as women get pressurised. The interviewee displays her orientation that women does not adequately refer to everyone who is pressed to divorce because she adds and sometimes men have been pressurised. The repetition of have been pressurised provides a right frame to both the trouble source women and the resultant repair solution, which parses as “women and sometimes men have been pressurised.”
The above two extracts show speakers’ displayed orientation to being gender inclusive. In both cases an initial general person reference form was used and a subsequent reference was to only one gender. The gender specific subsequent reference reflected was treated as being a trouble source and addressed by adding the missing gender in order to include all possible members. The subsequent reference pointed to a default gendered association of the general personal reference. In Extract 4, the default assumption is that a candidate for political office is male, and in Extract 5 it is that women who are pressurised to divorce. In both Extracts 4 and 5, the speaker interrupted their turn in progress to add in the gender category that has just been excluded, albeit using different prefaces (and, or).
The next example is another that shows an orientation to gender-inclusive language albeit of a different type from the previous extracts. It is from a feminist consciousness-raising group for university students. The group operates with “Safe Space Rules,” which are guidelines put in place with the aim of creating a trusting, non-judgemental, and inclusive environment. One of the safe space rules pertains to language and asks for the use of gender-neutral language. The group was encouraged by the facilitator to use the terms “male-bodied” and “female bodied” to refer to men and women, respectively. The idea behind such language use is that a female bodied person might not identity as female and a male bodied person may not identity as male. The guideline is an example of language planning with the aim of being more inclusive to transgender, intersex, and other gender variant people.
Extract 6 comes from a moment in the group discussion where a question is posed about sex work. Maya is telling a story about an actor, Hugh Grant, who was found to be having sex with a prostitute.
Maya uses the term woman to make an initial reference to the prostitute that Hugh Grant picks up off the street (Line 22). The facilitator corrects Maya in the next turn with female bodied (Line 24). A third turn in a correction sequence regularly accepts or rejects the correction (Jefferson, 1987). In this case, Maya accepts the correction by repeating it. Further trouble arises with Maya’s selection of a female pronoun her as a subsequent reference term, which the facilitator corrects with their, which was also accepted by Maya by repeating the term. That acceptance was acknowledged and supported by the facilitator by the use of finger snaps—another interactional practice promoted in the group as a way of supporting what was being said without interrupting the speaker.
The next gendered pronoun use she (Line 31) is not oriented to as a trouble source, perhaps because it is in the aftermath of the two previous correction sequences. However, the one after that—she (Line 33)—is oriented to as a trouble source. Unlike the previous two gender repair sequences, where the facilitator both initiated and did the correction, this third is both initiated and addressed by Maya—the current speaker. The gendered pronoun is oriented to as a trouble source and addressed by replacing it with that person. The repair segment is noticed and positively receipted by the facilitator with snaps, which—despite intentions to the contrary—do disrupt the smooth progressivity of the speaker’s turn.
Extract 6 illustrates the difficulties in using gender-neutral terms in English. Many person reference words make the gender of the referent available (e.g., woman, man). However, the use of lexically gendered terms does not necessarily make gender as a social categorisation relevant. Rather, gender marked nouns are regularly used in a straightforward referential way (Kitzinger, 2007; Schegloff, 2007a). By calling attention to the unmarked used of lexically gendered terms, the facilitator makes gender relevant rather than simply available, which was not the intention of the facilitator.
The analysis of this section shows that gender trouble can arise when either an initial or subsequent person reference is oriented to as being referentially inadequate. The trouble can be addressed by the current or the next speaker by adding reference forms that complete a gendered category set (as in Extracts 4 and 5) or by replacing with a term that does adequately describe the referent. In all cases, the trouble being oriented to as relevant is that of exclusion and the solution is to add or change the person reference to be gender inclusive—a central tenet of non-sexist or gender-fair language policies.
Discussion
Feminists have highlighted a fundamental relationship between language and a gendered social order that disadvantages women. At the establishment of gender and language as a legitimate field of inquiry, work documented the ways language can be understood as ignoring or narrowly defining and demeaning women. Cognitive social psychology has confirmed various psychological implications of sexism in language (Douglas & Sutton, 2014; Ng, Kam Kuen, Weatherall, & Moody, 1993). Social constructionist–informed discourse analysis details how gendered subjectivities and moral orders are produced and resisted (Baxter, 2003). However, there is little research that examines how aspects of sexist language actually unfold in social interaction—a situation the present study sought to rectify.
Using feminist conversation analysis, this article developed J. Butler’s (1990, 1997) theoretical notions of gender trouble and excitable speech as empirical phenomena. It was operationalised as moments in social interaction when repair or repair like practices made gender or sexism observably relevant. Operationalising concepts in observable phenomena is counter to a pure conversation analytic approach where the methodology dictates observations of members’ displayed concerns are made, for example, by the next-turn proof procedure. In Extract 1, for example, Mike displays his understanding of a bit of a girl as sexist. However, in Extract 3, old moaning wife is analysed as sexist when there is no evidence the participants are treating it that way.
Feminist conversation analysis is less restrictive than pure conversation analysis. It enables a focus on interactional practices that, from a researcher’s perspective, may be implicated in the accomplishment of a gendered social order. However, it also protects against the kind of gender essentialism plaguing; for example, sociolinguistic-informed forms of discourse analysis (see Speer & Stokoe, 2011). Feminist conversation analysis, like pure conversation analysis examines concrete practices in talk—grounding analytic description in what is visible in the talk. In doing so, it nuances understanding of talk-in-interaction as an institution that accomplishes gender and sexism. The present study developed the notion of gender trouble into empirical phenomena in talk that could be interrogated using conversation analysis.
Repair was a structural domain of organisation of talk-in-interaction that was central to gathering instances for examination. Word selection emerged as an important domain in the analysis—in particular membership categorisation devices (MCDs). MCDs were clearly operating in cases where there was a miscategorisation of the referent—a mapping of a member into a category to which they did not belong (Mike as a girl, Extract 1; a girlfriend as a wife, Extract 2; and Dad as a girl and an old moaning wife, Extract 3). The miscategorising invoked negative connotations of words for women, which is one of the documented aspects of sexism in language. It also confirms MCDs as part of the infrastructure of gender and the inferential order explored by Stokoe’s work on gender categories and the systematicity of their use (Stokoe, 2012).
The issue of gender inclusion is an important matter. The examples presented in this study involved an orientation to the referential adequacy for a subsequent reference to a general category of persons. One of the cases presented (candidate/he; Extract 4) was what Stringer and Hopper (1998) identified as a pseudo-generic where there is a default assumption that a referent is male. That default assumption was oriented to as a trouble source and resolved in the same turn by further specifying that referents could be female too. It appears the default assumption is driven by common-sense, which is inherently ideological (Billig et al., 1988). Political candidates are typically male (Extract 4); women being pressurised to divorce out of inter-sect marriages (Extract 5), and gender identity usually fits with the corresponding sexed body (Extract 6). It seems reasonable to suggest that these default next references are related to some cognitive infrastructure with just micro-seconds to interrupt that in order to progress perhaps more intentional order and action at an interactional level. Support for that suggestion is the two cases taken from investigative interviews on American and British broadcasting radio where unbiased and neutral reporting is an institutional mandate (Clayman, 2002).
Also part of word selection as a generic domain of organisation are person reference practices. These have an established set of preferences such as using a first name to refer to someone known to all parties (Schegloff, 1996). Stivers (2007) documented breaches to normative preferences for person referencing can operate to increase or decrease social distance between the speaker and the referent. That was clearly the case in Extract 2 where the category descriptor the wife was used to refer to someone both parties knew as Betty. The speaker was doing this to reject a proposal of possible additional company on a trip to the movies.
The study of person reference and its repair is relevant to the study of social identities in talk. A focus on a larger collection of gender trouble would reveal further insights into what they show about social order and action. One barrier to larger scale studies is that cases of gender trouble are relativity scarce and are not collectable by targeted data collection methods typical in social psychology. Rather, cases are found serendipitously in the course of a general programme of research that involves developing corpora for studies of social interaction. A collaborative approach to forming large collections of specific phenomena has been tried, with some success, in the international conversation analytic research community (Lerner & Kitzinger, 2007).
The relative rarity of the phenomena collected in the present study should not erode its significance. Mundane social life is a practical accomplishment structured and organised by the everyday practices of its members (Sacks, 1992). The practices that underpin some major forms of social organisation such as heteronormativity largely pass-by unnoticed requiring a critical eye to reveal them, as Kitzinger’s (2005) feminist conversation analytic work on heteronormativity so clearly established (Land & Kitzinger, 2005). The intellectual project of making the gendered social order visible in order to subvert it is core to feminist language research—feminist conversation analysis can make an important contribution to that by illuminating new details of sexism in action.
Footnotes
Author’s Note
Earlier versions of this article were presented at the 2008 Australasian Institute of Ethnomethodology and Conversation Analysis conference, Australia; the 2012 International Gender and Language conference, Brazil; and the 2013 Nordic Conference on Gender and Language, Sweden. Some of the data in this study were drawn from the Wellington Corpus of Spoken New Zealand English. See 
. Some other cases were from a study of insertion repairs with Prof. Sue Wilkinson, Loughborough University.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.