one guide is attached with 3 readings. Tutor needs to choose only 1. Critical Response Papers: Tutor is required to write 3 Critical Response Papers throughout the term. Each paper should be about 1

1 INTRODUCTION FEMINISM, DISCOURSE AND CONVERSATION ANALYSIS These are exciting times for the feminist study of gender talk. Since the mid-1970s there has been a rapid growth in the number and range of approaches that have set about exploring the relationship between gender and language. This is, in part, a consequence of the postmodern ‘death of the subject’ and ‘turn to discourse’ in the social and human sciences, in which language is seen, not simply as a neutral means of expression, a passive vehicle through which we report on events and experiences—but instead, as something that is central to the constructi on and reproduction of gendered selves, social structures and relations (Gergen 1985; Shotter and Gergen 1989). Few feminists would dispute that discourse is often gendered, and that it forms one of the primary means through which patriarchy and oppressive norms and social practices are instantiated and reproduced. Indeed, we are, as feminists, increasingly aware of the fundamentally political nature of discourse. When we use discourse to communicate we ‘naturalize’ and perpetuate oppressive understandings of gender and ‘gender role behaviour’—that is, we present them as timeless, rational and natural. These understandings become deeply ingrained in our commonsense views about the world, and become regarded as normative and expectable. Likewise, we’re aware that the politics of discourse is not one-dimensional. Discourse can be used to expose and ‘denaturali ze’, commonsense understandings of gender (through the use of humour and irony, for example) and to challenge ideas which create and sustain sexist and heterosexist social practices. By studying gender and discourse, and by exploring how dominant or prejudicial ideas about gender are created or resisted in discourse, we can acquire knowledge that can be used to inform social change for the better. Research on gender and language is diverse, spanning a range of disciplines. Just as there is no one feminist theory or method, but rather multiple ‘feminisms’, so too there is no one approach to the study of gender and language. The field is characterized by epistemological and methodological diversity. It draws on approaches ranging from the phenomenological and e xperiential to the positivist and experimental, each revealing different, and often competi ng, theoretical and political assumptions about the way discourse, ideology and gender identity should be conceived and understood.

Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

There now exist numerous books on gender and language, gender and discourse, feminism and discourse, sexuality and language, gender and conversation, and gender and interaction (for some recent examples, see Cameron and Kulick 2003; Eckert and McConnell-Ginet 2003; Fenster-maker and West 2002; Holmes and Meyerhoff 2003b; Hopper 2003; Mcllvenny 2002c; Weatherall 2002a). Likewise, there are an increasing range of books on DA and CA (for some widely used examples, see Atkinson and Heritage 1984; Hutchby and Wooffitt 1998; ten Have 1999; Wetherell et al. 2001a, 2001b). However, the problem that faces feminists and other researchers new to the field of gender and language research is how to go about identifying and choosing between the numerous approaches to discourse that now exist. When one reads these books, as the students on my level three elective module, Feminism Discourse and Conversation, will testify, it is rather difficult to establish precisely what the theoretical and methodological ‘boundaries’ of the various approaches are, where they diverge and overlap, and what approach or combination of approaches might prove most productive (both empirically and politically) for feminism. It’s easy to get bogged down in the detail of the different studies that are reported, at the expense of gaining a broader understanding of how the diverse theoretical and methodological models that are represented relate to each other, and a sense of their possibilities and problems.

The purpose of this introductory chapter is to provide an overview of the field of research on gender and language, and to situate my own approach in relation to it. In order to provide the reader with a sense of how my particular analytic approach fits within the broad terrain of research on gender and language, I want first to describe what I see as the main perspectives that dominate the field today. I outline some of the problems with the field as I see it at present—in particular, the reluctance on the part of many feminists and critically oriented researchers to adopt the kind of fine-grained form of analysis associated with CA. I describe five key features which distinguish the analytic perspective I argue for in this book, and which illustrate how I think the relationship between gender and language should be conceptualized. These five features serve as an organizing framework throughout the book, and provide a template for the reader to understand my particular analytic approach, and the criteria that I use to evaluate a range of other perspectives.

Feminist research on gender and language: mapping the terrain For the purposes of this book, I group feminist research on gender and language into four broad organizing frameworks or traditions: sexist language; interactional sociolinguistics and the ethnography of communication; ‘critical’ discursive approaches informed primarily by (one or more of) critical theory, poststructuralism and psychoanalysis; and finally, discursive approaches informed primarily by ethnomethodology and CA.

Research conducted within these four frameworks will be discussed in more detail in specific chapters as I work through the book. It must be emphasized that any attempt at categorization will caricature a far more complex terrain. There are no neat boundaries separating these frameworks from each other. Their boundaries are ‘leaky’, precisely because there is much cross-fertilization of ideas. While some researchers work in more Gender talk 2 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

than one tradition, others have adopted new frameworks and have moved into new research areas as their approaches have developed over time.

Sexist language Research on gender and language has traditionally been divided into two strands: the study of how gender is represented in the language (the form of language) and the study of how men and women use language (the function of language). The study of how gender is represented in the language is a vibrant body of work which starts from the assumption that language is an ‘ideological filter on the world’ (Ehrlich and King 1994:60). From this perspective, language reflects and perpetuates a sexist and heterosexist version of reality. Examples of sexist language include the purportedly generic pronouns ‘he’ and ‘man’, words such as ‘mankind’, job titles ending in ‘-man’, and the asymmetry of address terms for men (‘Mr’) and women (‘Mrs’/‘Miss’), where women are defined—not in their own right—but in terms of their relationship to a man.

Some of the earliest work by feminist linguists such as Robin Lakoff (1973, 1975) set about demonstrating a range of ways in which language is sexist, while the radical feminist, Dale Spender (1980), explores the development of what she calls ‘he/man’ language. I discuss Lakoff and Spender’s work in more detail in Chapter 2 (see also Henley 1987; Miller and Swift 1976).

There now exists an extensive body of research on sexist linguistic forms, and a range of sexist forms have been identified (see Mills 1995; Weatherall 2002a for examples).

Commentators have diverse views about how sexist language should be conceptualized and remedied. According to some feminist reformers and the writers of ‘non-sexist language guidelines’, for example, sexist talk can be eliminated through the development of linguistic innovations which replace sexist with non-sexist words (see Doyle 1995; Miller and Swift 1980). Examples of such reforms include substituting the masculine ‘generics’ ‘he’ and ‘man’ with neutral terms such as the singular ‘they’ and ‘he/she’, replacing job titles ending in ‘-man’ with neutral titles such as ‘chairperson’, ‘chair’ or ‘spokesperson’, creating neutral address terms for women such as ‘Ms’, and developing new categories that give meaning to experiences that have hitherto been ignored, such as ‘sexual harassment’ and ‘date-rape’ (Ehrlich and King 1994:61; for a comprehensive account of feminist linguistic reform, see Pauwels 1998, 2003).

Although the English language is certainly evolving to contain fewer sexist forms (Weatherall 2002a: 12), some feminist linguists, most notably Deborah Cameron (1992, 1998b), have written extensively on the problems that are associated with linguistic reform efforts. In particular, Cameron is critical of any approach which implies that sexist meanings reside in, or come attached to certain words. For her, it is deeply problematic to imply that there are a limited number of context-free, derogatory terms that are ‘essentially’ sexist, and that by pinpointing them and substituting them with ‘non-sexist’ words, one can somehow rid the language of sexism. As she states, ‘we cannot simply change a word’s meaning for the whole community by fiat’ (1992:110, emphasis in original). According to Cameron, many institutional reform efforts and attempts at what has been termed ‘verbal hygiene’ (Cameron 1995) treat sexist talk as a linguistic rather than as a social and a contextual problem, and ignore the ‘context sensitivity’ of actual language use. Thus, arguments underlying linguistic reform work by stripping talk of its Introduction 3 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

contextual subtleties and by caricaturing what might count as sexism to a significant degree (see also Cameron 1998b).

For many feminists, discourse analysts and sociolinguists, including Cameron, the meaning of words is not fixed but fluid. Linguistic meanings are socially constructed, contextually variable and continually subject to negotiation and modification in interaction. It follows that specific words need not always be sexist or egalitarian in their function. If the meaning of words is (at least partially) dependent on their context of use, then the whole idea that we can legislate ‘non-sexist’ language into existence— essentially fixing the meaning of ‘approved’ versus ‘sexist’ words, must be questioned.

As Speer and Potter (2000, 2002) and Speer (2002b) show in their research on the discursive construction of sexism and heterosexism, just as purportedly derogatory words (e.g., ‘dyke’, ‘queen’, ‘queer’) can, under certain circumstances (e.g. when they are reclaimed by lesbians and gay men, or used humorously or ironically), be invested with new meanings, and used to non-derogatory purposes, likewise seemingly benevolent, nonsexist or practically oriented descriptions and evaluations which do not index sexist words (for example, the claim that ‘women should not play rugby because they might get injured’) can be built and used in order to justify inequality and sexism (see also Gill 1993; Wetherell et al. 1987). For these researchers, the precise meaning of what is said can be established only by exploring what utterances are doing in specific contexts. As many discourse analysts now show, in contemporary society, there may even be a norm against explicit forms of prejudice, such that we are all apparently—overtly at least— ‘politically correct’ (Suhr and Johnson 2003) and hence ‘liberal in our views’ (Clarke 2005). Thus, speakers can frequently be heard to preface some arguably prejudiced claim, with a disclaimer (e.g., ‘I’m not sexist but…’, or ‘I’ve got nothing against gay people, my best friend is gay but…) (Stokoe and Smithson 2002:91; see also Potter and Wetherell 1987; van Dijk et al. 1997). This body of work highlights the problems that derive from any study which confines itself to the analysis of discrete word forms, and how linguistic meanings are more malleable, and more ‘context-sensitive’ than many researchers have hitherto assumed. I discuss this work further in Chapter 6.

Other gender and language researchers, while acknowledging that linguistic meanings are not fixed, are nonetheless concerned that, if taken to its logical conclusion, such ‘linguistic relativity’ may be taken to imply that the meaning of words is infinitely malleable, and that we can take them to mean whatever it is that we want them to mean.

A ‘communities of practice’ perspective works to address this concern by locating sexism, not in specific words or individuals, but in distinctive social and political contexts—or ‘linguistic communities’ (see Eckert and McConnell-Ginet 1992; Ehrlich and King 1994; Holmes 1999; McConnell-Ginet 1989). Here, ‘communities of practice’ are defined as ‘an aggregate of people who come together around mutual engagement in an endeavour. Ways of doing things, ways of talking, beliefs, values, power relations—in short, practices—emerge in the course of this mutual endeavour’ (Eckert and McConnell- Ginet 1992:464). For proponents of this approach, the meaning of an utterance ‘is a matter not only of individual will but of social relations embedded in political structures’ (McConnell-Ginet 1998:207). Thus, linguistic meanings can be fully understood only when one considers the nature of the social context and, specifically, the background social knowledge and ‘mutually accessible cultural beliefs’ (Ehrlich and King 1994:60) available to the linguistic community in which words are uttered. From this perspective, Gender talk 4 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

linguistic reform efforts will not always succeed, because ‘terms initially introduced to be non-sexist and neutral may lose their neutrality in the mouths of a sexist speech community and/or culture’ (Ehrlich and King 1994:59; see also Ehrlich and King 1992; McConnell-Ginet 2003).

Although there are problems associated with the application of this ‘communities of practice’ approach (for example, how might one decide where a ‘community’ begins and ends, or account for variable, or contradictory linguistic meanings within the same community?), its value, nonetheless, lies in the way in which it helps to account for how linguistic meanings can ‘congeal’ or ‘sediment’ through time in specific institutional and cultural contexts, and how some non-sexist linguistic innovations may take hold in certain groups in society, among some individuals (e.g., young feminist women, gay men) and not others. It deals, in other words, with the perennial problem of how to account for how individuals are neither totally determined by language, nor totally free to make words mean whatever it is that they want them to mean.

Interactional sociolinguistics, and the ethnography of communication The study of how men and women use language (the function of language) has its roots in linguistics and anthropology. Interactional sociolinguistics developed in the work of John Gumperz (1982a, 1982b) and focuses on the relationship between gender, language and culture. Sociolinguists are of the view that variations in patterns of language use are not random but are conditioned by macro-social and demographic features such as a person’s gender or class, and the situation or context in which they find themselves. Unlike their variationist and quantitative sociolinguistic colleagues who use statistics to explore male- female patterns of linguistic (that is phonological and grammatical) variation (for an overview see Romaine 2003), proponents of interactional sociolinguistics use predominantly qualitative methods to study male-female variation in patterns of interaction and communicative style. These communication patterns are learnt during socialization, or emerge as a result of gender-segregated play during childhood.

Examples of this approach can be found in the work of Coates (1986), Fishman (1978), Holmes (1995), Maltz and Borker (1982) and Tannen (1990, 1994b, 1997). Their findings have been buttressed by a range of work on children and child development in feminist psychology (e.g., Gilligan’s (1982) theory of women’s moral development). I discuss Tannen’s work in more detail in Chapter 2.

A complementary approach—the ‘ethnography of communication’ (previously termed the ‘ethnography of speaking’) was developed in the work of Dell Hymes (1962, 1974) and describes research which uses ethnographic methods to explore how language using is done and understood differently by men and women in different cultural groups.

Typically critical of the ‘separate worlds hypothesis’ developed in interactional sociolinguistics, and the ‘polarizations of gendered norms of social interaction and communication’ associated with it (see Goodwin 2003:231; Kyratzis 2001), proponents of this approach deem ethnicity, social class, and context to be central to an analysis of gender talk. Focusing primarily on language use in either non-western, non-industrialized societies, or in culturally distinctive small groups within western societies, proponents of this approach focus on the ‘ways of speaking’ and the ‘discourse genres’ that are exhibited by the members of the culture being studied, and the diverse forms of social Introduction 5 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

organization that are possible within that group. They are also concerned to explore cultural variability in gendered patterns of language use or communicative style, and this sets their work apart from that associated with many other approaches to discourse mentioned later in this chapter. This perspective is closely associated with the work of Marjorie Harness Goodwin (1990, 2001, 2002, 2003), who combines ethnographic methods with CA, and Elinor (Ochs) Keenan (1989; Ochs 1992; Ochs and Taylor 1995; see also Danby 1998; Kyratzis 2001; Sheldon 1990; Thorne 1993).

So far I have described research which explores how gender is represented in the language (the form of language), and how men and women use language (the function of language). Although both strands continue to be pursued as somewhat separate forms of inquiry, the field has developed such that many now regard them as part of the same process: the social construction of gender (Cameron 1998a; Crawford 1995). As some of the research discussed above highlights, meaning is not necessarily tied to specific words.

Likewise, patterns of variation in our use of language or communicative style are not straightforwardly determined by the sex of the speaker, or by one’s membership of a cultural group. Instead, sexism, gender and cultural difference are constructed in and through our interactions with one another. Central to this new understanding is the view that gender and sexism is ‘best analysed at the level of discourse’ (Cameron 1998c:87). It is discourse rather than individual words, which, it is argued, constitutes the ‘main locus’ (Cameron 1998a:962), or the key site, for the reproduction and resignification of gendered meanings. This focus on discourse has, in turn, led to a gradual shift away from research which analyses sexist word forms and ‘decontextualized sentences’, or which searches for the linguistic or cultural correlates of gender difference, toward an analysis of more extended sequences of language use, and its role in naturalizing specific understandings of gender.

Discourse analysis is a collective term for a diverse body of work spanning a range of disciplines. It is now possible to identify a broad variety of different types of DA, which derive from widely varying theoretical traditions (see Cameron 2001; Wetherell et al.

2001a, 2001b, for overviews). Feminist applications of DA are equally diverse: the first text on feminism and DA from a psychological perspective contains exemplars of a range of different approaches (Wilkinson and Kitzinger 1995). What does not help matters, of course, is that the label ‘discourse analysis’ tends to be applied uniformly, regardless of differences in theoretical focus and the level of analytic specificity exemplified across different studies. Moreover, some discourse analysts appear inconsistent in their approach, shifting between different perspectives in different papers, thus making it hard to identify coherent discourse ‘types’; compare, for example, the different styles of analysis in Wetherell and Potter (1992) and Antaki and Wetherell (1999). The recent move by some discourse analysts to embrace the perspective that some are now calling ‘feminist CA’ has only served to exacerbate this lack of clarity (see, for example, the special issue of Discourse and Society on ‘Gender, language, conversation analysis and feminism’: Stokoe and Weatherall 2002; see also Kitzinger 2000a, 2002).

Many of the sociolinguistic researchers whose work I mentioned above pursue, or have gone on to pursue, questions about the relationship between gender and language by conducting some form of DA. Although it is widely acknowledged that such a distinction oversimplifies matters somewhat, and that the different types of DA overlap in many important respects, it is nonetheless common nowadays to divide the field into two Gender talk 6 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

different ‘camps’ (Edley and Wetherell (1997), ‘strands’ (Widdicombe and Wooffitt 1995) or ‘styles’ (Wetherell 1998) of discourse analytic work (for a range of perspectives on the ‘two strands’ debate, see Burr 2003; Parker 1997; Widdicombe 1995; Willig 2001). For the purposes of this book, and for the sake of clarity, I divide the field into ‘critical’ discursive approaches informed primarily by (one or more of) critical theory, poststructuralism and psychoanalysis, and discursive approaches informed primarily by ethno-methodology and CA.

‘Critical’ discursive approaches informed primarily by (one or more of) critical theory, poststructuralism and psychoanalysis The first strand of DA incorporates work which applies ideas from a range of disciplines.

Intellectual precursors to this work can be traced variously to poststructuralism (e.g., Foucault 1971, 1972; Kendall and Wickham 1999; Weedon 1997), feminist theories of the performative constitution of gender (Butler 1990a, 1993), positioning theory (Davies and Harré 1990; Harré and Moghaddam 2003), the Frankfurt School of critical theory (Habermas 1984), critical linguistics (Fairclough 1989, 1992, 1995; Fair-clough and Wodak 1997; Kress 1985; van Dijk 1993), philosophy (Derrida 1976; Wittgenstein 1953) and psychoanalysis (Lacan 1989). Feminists have combined one or more of these approaches to develop their own distinctive brand of DA.

Although individual studies differ slightly in their precise analytic emphasis, advocates of this strand of DA tend to conduct broad-based, topic or theme-focused analyses focusing on power, ideology and the self. They explore the ‘constitutive’ power of discourse, and seek to identify the ‘broad meaning systems’ invoked in talk, variously termed ‘global patterns in collective sense-making’, ‘interpretative repertoires’, ‘practical ideologies’ and ‘psycho-discursive’ resources. Drawing on relatively lengthy excerpts of talk from both spoken and written texts, researchers working within this framework commonly transcribe data to a level which represents the general content of the words spoken, as opposed to the dynamics of turn-taking or the characteristics of speech delivery. Since advocates of this approach tend to focus on how gendered subjects are positioned through or by discourses, and how powerful social structures, norms, ideologies and conventions shape and constrain individuals’ actions ‘from above’ or ‘outside’ the text, critical forms of DA are commonly referred to as ‘top-down’, ‘macro- level’ forms of analysis. Proponents of this approach are influenced strongly by political aims. Indeed, this form of DA is commonly referred to as critical DA (or CDA), because researchers have a ‘leftist’ or ‘socialist’ political and analytic stance, and their primary goal is to examine texts which naturalize unequal power arrangements and ideologies.

Researchers are driven by a belief that insights gained from the analysis of discourse and ‘engaged scholarship’ will help us to change society for the better.

Feminists who employ this approach, who have backgrounds in socio-linguistics, and who are influenced primarily by poststructuralism or theories of performativity, include Baxter (2002, 2003), Bucholtz et al. (1999), Cameron (1997a), Coates (1996, 1997, 1999), Hall and Bucholtz (1995) and Sunderland (2004). Feminist ‘critical discourse analysts’ who are influenced primarily by research in the Frankfurt School and critical linguistics include Talbot (1997, 1998, 2000) and Wodak (1989, 1997, 2003). Within psychology, feminists influenced primarily by theories of social positioning or Introduction 7 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

psychoanalysis, include Frosh et al. (2003), Gough (2004), Henriques et al. (1984), Hollway (1989, 1995, 1998) and Kulick (2003). Finally, proponents of a feminist ‘critical discursive psychology’ (or CDP), who are informed by an eclectic mix of poststructuralism, theories of social positioning, social constructionism, linguistic philosophy and ethnomethodology, include Burman and Parker (1993), Crawford (1995), Edley and Wetherell (1997, 1999, 2001), Gavey (1989), Korobov and Bamberg (2004), Weatherall (2002a), Wetherell (1998, 1999a, 2003) and Wetherell and Edley (1998, 1999).

Discursive approaches informed primarily by ethnomethodology and CA The second strand of discourse work takes its primary influence from ethnomethodology and CA. Ethnomethodology developed in the work of Harold Garfinkel (1967) and takes as its topic for study ‘members’ methods’ for producing their everyday affairs. Members’ methods consist of the routinized, taken-for-granted procedures individuals employ as they go about their everyday lives and tasks. CA was developed in the pioneering lectures of the American sociologist, Harvey Sacks, between 1964 and 1972 (Sacks 1995), and has its roots in Garfinkel’s (1967) ethno-methodology, Goffman’s (1983) theory of the interaction order and linguistic philosophy (Austin 1962; Wittgenstein 1953). Harvey Sacks and his colleagues, Emmanuel Schegloff, Gail Jefferson and Anita Pomerantz, were among the first to translate ideas from these perspectives into an empirically grounded, data-driven, and highly systematized research agenda. Building on the ethnomethodological critique of the structural functionalist, Parsonian idea that order is achieved because individuals act on the basis of an internalized system of constraining social norms (Parsons 1937), conversation analysts instead theorize members as active participants who produce and orient to social order as they interact with one another and engage in ongoing, interpretative, meaning-making work Thus, for conversation analysts, the interest is in the in situ organization of conduct, and the local production of order (for more on the relationship between ethnomethodology and CA, see Heritage 1984, 2001).

Drawing on analyses of first-hand, transcribed examples of everyday interactions, CA is primarily concerned to describe the methods speakers use to coordinate their talk to produce orderly and meaningful conversational actions. It studies the design of both individual utterances or ‘turn constructional units’, as well as the organization of turns into sequences of utterances. Finally, it seeks to explicate how talk is implicated in broader forms of social organization at the ‘institutional’ or ‘social structural’ level.

While CA developed primarily in sociology in the United States in the 1960s, the related approach of discursive psychology (DP) was developed in social psychology in the United Kingdom in the 1980s by Potter and Wetherell (1987; see also Gilbert and Mulkay 1984). DP has developed an angle which applies principles from ethnomethodology, CA, linguistic philosophy and the sociology of scientific knowledge, to rethink a distinctive set of concerns around cognition, and a range of other psychological concepts, processes and questions. Thus, where CA respecifies the dominant sociological concept of social structure, DP respecifies the dominant psychological notion of cognition. Indeed, DP has a very distinct way of conceiving of the psychological world and research into it. Discursive psychologists are critical of the idea that mental entities such as ‘attitudes’ or ‘feelings’ can simply be ‘read off from talk, Gender talk 8 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

or that talk is a relatively transparent ‘window’ on our inner worlds. Instead, they are interested in what people do with attitude talk and feeling talk, and how a whole range of ‘mentalist notions’ are constructed and used discursively by participants in interaction.

This focus on action, not cognition, entails a reformulation of traditional ‘theories of the subject’, or identity, in practical, discursive terms (for examples of this ‘anti-cognitivist’ approach, see Edwards 1997; Edwards and Potter 1992; Potter 1996b; Te Molder and Potter 2005). Indeed, within both DP and CA, identities are treated, not as demographic facts or variables that condition or constrain behaviour, and which people just ‘have’ (cf.

much sociolinguistics). Rather, features of identity (including gender), are treated as locally occasioned resources, whose relevancies may or may not be oriented to, or ‘procedurally consequential’ for talk (Antaki and Widdicombe 1998a).

Although the relationship between CA and DP is a contentious point for those who are concerned with defining, defending and policing their disciplinary boundaries—see, for example, the debate between Hammersley (2003a, 2003b, 2003c) and Potter (2003a, 2003b)—CA and DP are becoming increasingly hard to separate on methodological and conceptual grounds. As David Silverman (1998:193) points out, since there is evidence that some discursive psychologists ‘pay considerable attention to the turn-by-turn organization of talk…we may end up in a pointless debate about whether such work is DA or CA!’ Indeed, Silverman (1998:193) notes that the distinction between DP and CA often rests on whether the author pays their ‘disciplinary dues’ to psychology or sociology (see also Wooffitt 2005). Moreover, although Margaret Wetherell—one of DP’s central proponents—has shifted more firmly toward the more ‘eclectic’ understanding of discourse associated with CDP, other key DP figures, including Jonathan Potter and Derek Edwards, have moved strongly towards the CA end of the analytic spectrum, and have recently come together with the conversation analysts John Heritage, Paul Drew, Anita Pomerantz and others, to write about the relationship between talk and cognition (Te Molder and Potter 2005).

Proponents of this CA strand of DA align with Harvey Sacks’ (1995) view that much of what is going on in interaction occurs in its particulars: in the details of pauses, turn taking organization, hesitations, word choices, repairs and overlaps. CA has developed a comprehensive system for transcribing such features, and a simplified version of these transcription conventions are included in the Appendix. Analyses are typically conducted using a number of short extracts, transcribed and analysed to a high level of technical detail. Sometimes researchers analyse longer stretches of talk—or ‘single cases’—to explore how a range of conversational devices figure in its production.

Finally, since advocates of this approach tend towards more finegrained, action-driven analyses which stick closely to an analysis of participants’ orientations and what is happening within the data—from ‘ground level’, so to speak—it is commonly referred to as a ‘bottom-up’, ‘micro’ (and, more controversially for the perspective developed in this book), ‘non-critical’ approach.

Neither CA nor DP are explicitly feminist in orientation. However, there is a long tradition of feminist work utilizing ideas from both perspectives to study a variety of topics relevant to feminist concerns. For example, a number of feminists have used CA to explore how patriarchy and male dominance ‘is realized at the micro-level of interaction’ (Stokoe 2000:556; see Ainsworth-Vaughn 1992; Davis 1988; Fisher 1986; Fishman 1977, 1978; Todd 1989; West 1979, 1992; West and Garcia 1988; West and Zimmerman Introduction 9 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

1983; Zimmerman and West 1975). Others have used CA to highlight a range of gendered interactional patterns, for example, women’s conversational competence (West 1995) and men and women’s use of directives (Goodwin 1990; West 1998). What is interesting about this work, however, is that even while these studies demonstrate clearly how ‘the inspection of authentic conversational materials might reveal more about women’s and men’s speech than armchair speculation (cf. Lakoff 1975)’ (Mcllvenny 2002c:15), they are nonetheless guilty of using CA for predominantly ‘non-CA purposes’ (Stokoe 2000:556; ten Have 1999). As Elizabeth Stokoe (2000:556) points out, since ‘these studies link gender to, for example, interruption, talk time, topic initiation and topic maintenance’, then they retain the idea that on some level, talk can and should be mapped onto demographic variables or gendered attributes.

Wary of such shortcomings, more recently, a critical current has developed among feminists with backgrounds in feminist psychology, which has set about treating gender, sexuality and prejudice as emergent phenomena that are made relevant in interaction, and which are constructed and oriented to as participants’ concerns (see, for example, Kitzinger 2000a; Kitzinger and Peel 2005; Speer 2001b, 2002d, 2005; Speer and Potter 2000; Stokoe 1998; Stokoe and Smithson 2001, 2002; Weatherall 2002b). A related strand of work has begun to merge this focus on participants’ orientations to gender, with insights from Sacks’ (1995) approach to membership categorization analysis (or MCA), in order to examine the construction of gender as a category in discourse, and how it is used and oriented to by members (see Edwards 1998; Fenstermaker and West 2002; Kitzinger and Wilkinson 2003; Speer 2002b; Stokoe 2003, 2004, in press; Stokoe and Edwards 2005; Wowk 1984).

Some problems with the field of research on gender and language The analytic approach I argue for in this book is closely associated with this second strand of ethnomethodological and CA-inspired discourse work. However, while this approach is becoming more popular among feminists, many gender and language researchers have been sceptical of the value of CA for feminist work. For the most part, the majority of gender and language researchers, and others with a critical, political agenda, have favoured the broader forms of DA influenced strongly by critical theory, poststructuralism, and psychoanalysis.

There is, in Schegloff’s terms, an ‘impatience, and often intolerance, of close analysis’ (1997:180; see also Widdicombe 1995). This is primarily due to misgivings about the political efficaciousness and practical utility of a technical and fine-grained approach to feminist issues and concepts—an approach which many believe focuses on the mundane and the trivial aspects of social life, at the expense of an analysis of the more politically consequential aspects. As Kitzinger (2000a: 173; see also Kitzinger and Frith 1999:311) observes, CA is often viewed ‘as nit-picking, obsessively concerned with the minute details of in-breaths and hesitations, and as unable to see beyond the “micro-” level of the 0.2 second pause, to the “macro” level of oppression’. Critics often raise the objection that CA—as a’micro’ approach to talk-in-interaction that limits itself to the study of members’ perspectives and the analysis of short extracts—cannot account for the ways in which gender norms and ‘wider, macro power structures’ exert a determining effect on Gender talk 10 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

action. Neglecting this ‘top-down’ or constraining feature of culture, society and context, some commentators argue, leads to apolitical and reductionist forms of analysis that leave us few opportunities to comment on patriarchal power or the oppressive constraints on women’s lives. As such, this approach is not only unnecessarily limiting, but also has little practical relevance to feminism, or the world beyond academia (Billig 1999a, 1999b; Edley 2001a, 2001b; Weatherall 2000; Wetherell 1998).

For the majority of feminists, feminism is a politics which rests on the belief that women as a group are oppressed by men as a group. It is dedicated to social and political change. Since participants themselves might not be conscious of the impact of the broader social and ideological context within which their utterances are embedded, an approach which limits itself to an analysis of participants’ orientations, and which relies on participants explicitly attending to the topic under investigation, offers little by way of resources to advance core feminist aims. As Hannah Frith (1998:535) notes, it is highly doubtful whether all the dimensions relevant to a piece of interaction, such as participants’ ‘shared whiteness’, will be ‘interactionally displayed’, or made explicitly manifest in the ‘micro-interactional’—that is, the small-scale, discursive features of an interaction. This view has led some feminists to suggest that Schegloff’s approach ‘limits admissible context so severely that only the most blatant aspects of gendered discursive practice, such as the overt topicalizing of gender in conversation, are likely candidates for Schegloffian analysis’ (Bucholtz 2003:52). For some critically oriented researchers, neither CA nor DP should be treated as self-sufficient paradigms (Hammersley 2003a:

751). Rather, both need subsidizing with other methods.

Others argue that the ‘symmetrical’, seemingly ‘neutral’ form of analysis associated with a participants’ orientations approach, is apolitical, ‘invites missed opportunities’ and ‘risks a form of ideological complicity’ (Edley 2001b:137). Mick Billig (1999b:554–6), for example, claims that CA’s ‘participatory rhetoric’ encourages the analyst to treat participants’ contributions equally, thus ignoring the power differential between—say— rapist and victim.

Finally, some assert that, contrary to the conversation analyst’s injunction that we must stick closely to an analysis of ‘participants’ orientations’, that CA analyses already involve—indeed rely on—the articulation of members’ and analysts’ cultural and commonsense knowledge as ‘largely unacknowledged and unexplicated resources’ (Stokoe and Smithson 2001, 2002). From this perspective, a ‘pure’ form of CA which does not go beyond the text and an analysis of members’ perspectives is not only an inaccurate and unrealistic portrayal of the actual practice of CA, but is also impossible (Stokoe and Smithson 2001, 2002).

Thus, for most of these researchers, we must venture beyond the limits of the text and the micro-analysis of members’ perspectives, in order to be able to observe, and say anything politically effective about, the ‘constraining’ or ‘enabling’ impact of a range of large-scale, ‘extra-discursive’, ‘macro’ social structural factors. Indeed, similar concerns do not just preoccupy feminists, but are part of a much broader and longstanding social scientific debate concerning the relative importance that we should give to the macro- social structural and micro-interactional realms, and which realm is primary or determinate of the other (Giddens 1981, 1984; Knorr-Cetina and Cicourel 1981). Introduction 11 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

How should the relationship between gender and language be conceptualized?

A view that I articulate in this book is that there is nothing intrinsic to a strongly CA- aligned discursive approach which would prevent feminists and others with a critical agenda from using it to ask politically motivated questions, or to reach politically efficacious outcomes. Indeed, I suggest that we should take seriously precisely those features of CA (for example, its focus on ‘participants’ orientations’ and the fine-grained analysis of talk), which have frequently been dismissed as anti-feminist (see also Kitzinger 2000a). We should see how far we can get with such analyses, and what they can offer us in terms of advancing our understanding of the constitution of gender, sexuality and prejudice in talk, and in terms of ‘grounding’ our feminist politics.

I argue that an adequate feminist discursive approach that draws on insights from the sociological perspective of CA, and the closely related psychological perspective of DP, would be characterized by five key features: (i) a constructionist approach; (ii) discourse of mind and world as topic, not resource; (iii) language as a form of social action; (iv) analysts’ claims are grounded in participants’ practices; and (v) a relativist approach.

These five features serve as an organizing framework throughout the book, and provide a template for the reader to understand my particular analytic approach, and the criteria that I use to evaluate a range of other perspectives. It’s important to note that these criteria do not map onto some preexisting feminist discursive type, and that not all researchers who are influenced by CA or DP would agree with the criteria that I have chosen (for example, as I will show later in this book, many conversation analysts would reject the suggestion that their work can be described as constructionist, or even relativist in its focus). I have chosen to combine elements from both perspectives because I believe that combined, they provide the most productive and ‘complete’ framework for reconceptualizing the relationship between gender and language.

(i) A constructionist approach In many social scientific studies, researchers adopt an essentialist approach to analysis, in which they treat people as having relatively fixed ‘traits’, ‘attributes’ or ‘essences’ residing inside them that condition what they do or say. Essentialist assumptions about sex and gender manifest themselves most clearly in the variationist and sociolinguistic studies of the type described earlier in this chapter. In sociolinguistic ‘sex differences’ research, for example, research participants are divided into groups of males and females, and sex differences in their use of language are mapped and measured accordingly. Sex and gender are treated as pre-given traits or ‘natural facts’, that reside in individuals and which determine the linguistic resources men and women use to speak.

Many feminists problematize essentialism on the grounds that it sustains and reproduces ‘binary thinking’ (Bing and Bergvall 1998; see also Ferree et al. 1999; Hare- Mustin and Maracek 1994; Hollway 1994; Kitzinger 1994b). Essentialist studies will always support arguments that men and women are fundamentally different, because they start from the assumption that the sex of the speaker both causes and accounts for male- female linguistic differences. Gender talk 12 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

In order to avoid this kind of circular reasoning, the approach I adopt throughout this book, is, by contrast, a constructionist one (Burr 2003; Gergen and Gergen 2003). In a constructionist analysis, both sex and gender are treated as fluid accomplishments.

Gender is something one does rather than as something that one has (see Bohan 1993; Cealey Harrison and Hood-Williams 2002; Harding 1998; Lorber 1994; West and Zimmerman 1987). Researchers who adopt a constructionist perspective focus on how gender identities are achieved, and treat the coherence of gender as something which is produced and reproduced in the course of social interaction.

In constructionist analyses, the focus of inquiry shifts away from studies which correlate linguistic variables with demographic variables, and which claim that ‘men talk like this’ and ‘women talk like that’, toward analyses of the dynamic ‘processes by which people come to describe, explain or otherwise account for the world (including themselves) in which they live’ (Gergen 1985:266, emphasis added). The focus is on how language users produce speakers as male and female, and construct, orient towards, and use gendered identities in their talk. This paves the way for a more detailed analysis of the ways in which people use language to produce gender difference, and to construct gender dualism as natural, inevitable and timeless (Cameron 1992; Fenstermaker and West 2002). I discuss the relationship between constructionist and essentialist theories of gender and language further in Chapters 2 and 3.

(ii) Discourse of mind and world as topic, not resource Mainstream sociology and psychology (and their feminist derivatives) work with a model of the world that is divided into three realms: the macro-social structural realm, the cognitive-psychological realm, and the micro-interactional, discursive realm. In mainstream sociology, for example, explanations for activity, discourse and other social phenomena (such as crime and prejudice) tend to be sought in terms of the ‘broader’, ‘extra-situational’, ‘macro-social’ contexts within which they are embedded. These contexts may include social institutions (such as the family and marriage), social structures (such as the law, economy or education system) or the social norms and conventions which are learnt through socialization and which give rise to particular ideologies and patterns of social behaviour (ideas about appropriate ‘masculine’ and ‘feminine’ behaviours, for example) (Burr 2003:8–9). Psychologists working within the dominant ‘cognitivist’ paradigm add a further level to this explanation, in that they treat such macro-social structural contexts as things which, in order to be perceived, experienced, and understood—that is, in order to have their effects—they must pass through the ‘interior’ cognitive psychological realm of mental states, cognitive structures and processes (Potter 1998c:33).

From within this framework, macro-social structures (‘the world’) and cognitive- psychological processes (‘the mind’) are treated as primary ‘inputs’ to social action, while activity, discourse and other social phenomena (e.g. crime or prejudice) are treated as something secondary—a by-product or ‘output’ of the system (Edwards and Potter 2001:15). In both cases, discourse—to the extent that it is studied ‘first hand’ at all—is treated as an analytic resource which, barring certain kinds of methodological bias and distortion or error, reflects the world and people’s perceptions of it (Edwards and Potter 2001:12). Introduction 13 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

The perspective I adopt in this book inverts this distinction. Instead of treating discourse as a secondary resource that can be used to access facts or information about the macro-social structural, and cognitive-psychological realms, world and mind (and their corresponding ‘realms’) are treated as phenomena that are constructed and oriented to by people in discourse as they go about their everyday lives and tasks (Edwards and Potter 2001:15). It follows that discourses of mind and world become studiable as the primary topic, or domain for analysis in their own right (Edwards and Potter 2001:15). I demonstrate what this approach looks like in Chapters 4 to 6.

(iii) Language as a form of social action As I showed in my discussion of sexist talk above, some researchers have sought to isolate and remedy problematic linguistic terms, and treat ‘discourses’ as having fixed meanings. In this book, by contrast, I treat discourse as a social practice rather than a thing (Potter et al. 1990). One of the central tenets of both CA and DP is that talk and texts have an action orientation—that is, the precise way we construct the world, and the import of an utterance, depends on the specific action or business that talk is designed to achieve. ‘Action’ here can refer to a whole range of practical, technical and interpersonal tasks that people perform as they go about their everyday lives and tasks. For example, as both CA and DP researchers have shown, discourse can be used to constitute events and identities, to manage issues of responsibility and stake, to present oneself in a favourable light, to account for one’s actions (to offer excuses, for example), to make invitations, requests, offers and assessments, to persuade and argue, and to achieve and manage justifications, mitigations and blamings. The analytic approach I adopt focuses on what discourse is doing, how it is constructed to make certain things happen, and the conversational resources that are drawn on to facilitate that action or activity (Potter 2004b).

The idea that language does things—that it creates rather than reflects meaning—is closely tied to the concept of indexicality. The notion of indexicality captures the ethnomethodological (Garfinkel 1967) idea that we settle or ‘fill in’ the meaning of an utterance on any given occasion, by noting information about the context in which the words are uttered. ‘Context’ here refers to the sequential or interactional environment of the talk itself, in which events unfold turn-by-turn, as well as the local context—the setting in which an activity takes place.

If the import of an utterance is tied to its context, then it follows that the same statement can be used to perform different actions, depending on the interpretative context in which it is uttered. For example, consider the evaluative phrase ‘I love dancing’: (a) when used in a conversation with a friend, the words ‘I love dancing’ may be an attempt to elicit an invitation to a night club (an implicit request); (b) when said to a friend who accuses you of dancing too much, and not spending enough time chatting to them in a night club, may be used as a justification or an excuse; and (c) when said to a partner who has just cancelled your ballroom dancing lesson, may be part of a blaming, or an attempt to make that person feel guilty. Thus, in order to understand what an utterance is doing, we need to analyse the local contextual and sequential environment in which it is situated. I flesh out what an ‘action orientation’ approach to discourse looks like, in Chapters 4 to 6. Gender talk 14 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

(iv) Analysts’ claims are grounded in participants’ practices Feminist research is epistemologically and methodologically diverse. However, a strong trend among second wave feminist writing on methodology has been a critique of mainstream social scientific methods. Many feminists reject ‘masculine’ notions of objectivity, value neutrality and scientific detachment, because they are thought to reinforce the objectification, exploitation and subordination of women (Cook and Fonow 1990:72). Consequently, many feminists adopt data collection practices that foster egalitarian and non-hierarchical research relations. Methods are chosen that will minimize harm to respondents, and which will ‘shift the balance of power and control toward the research participants’ (Wilkinson 1999:233). The overriding concern is to avoid imposing the researcher’s own analytic categories and concepts on what respondents say, and to encourage them to ‘assert their own interpretations and agendas’ (1999:233). In this way, the researcher gains access to participants’ own language, meanings and vocabulary, their ‘opinions and conceptual worlds’ (1999:233).

While the majority of feminists agree that data collection practices should cultivate non-hierarchical research relations, and that respondents ‘must be in the driver’s seat of research’ (Campbell and Salem 1999:67; see also DuBois 1983), the concern to adopt ‘respondent centred’ data collection practices in which the researcher’s role is minimized, tends not to be translated at the analytic level, where the relative importance given to the perspective of the respondent over that of the researcher is reversed.

Indeed, many feminists argue that since women are not always in a position to see— and thus problematize—their own oppression, that when it comes to our analyses, our role is not simply to act as a ‘neutral conduit’ through which participants speak, or to uncritically accept and ‘give voice’ to the generally non-feminist, non-politicized arguments of our participants (Kitzinger and Wilkinson 1997; see also Kitzinger 2003).

Instead, we are morally and politically obliged to ‘go beyond’ our data, to prioritize our feminist political agenda, and to treat our analyses as an occasion for doing politics.

Rather than simply reflecting and validating ‘whatever women tell us about their experience’, many of us use our analyses specifically in order to challenge and criticize ‘the way in which women’s experience is constructed under (hetero) patriarchy’ (Kitzinger and Wilkinson 1997:573).

In the feminist analytic approach I advocate in this book, by contrast, I avoid producing analyses which are driven, in the first instance, by my politics, and by my assumptions about the constraining or enabling features of a range of ‘extra-discursive’, ‘macro-structural’ or ‘cognitive-psychological’ factors. Instead, following Schegloff (1997), I consider, first and foremost, what is going on from a member’s perspective, and how the social and political is constituted and oriented to (if at all) in participants’ talk. I explore the notion of participants’ orientations further in Chapters 4 to 6.

(v) A relativist approach Many social scientists working within the mainstream positivist and inter-pretivist paradigms adopt a realist approach which supports the view ‘that there is a reality independent of the researcher whose nature can be known, and that the aim of research is to produce accounts that correspond to that reality’ (Hammersley 1992:43). From this perspective, providing that all sources of methodological bias and extraneous influences Introduction 15 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

are eliminated from our data (through the control of ‘contaminating’ variables, for example), then the discourse that the researcher collects can be treated as a transparent medium through which they can gain unmediated access to facts about the world and respondents’ perceptions of it, or the objectively measurable ‘reality’ that lies beyond or beneath talk. By contrast, the relativist approach I adopt in this book can be defined as:

a stance of systematic and thoroughgoing doubt about objectivist, essentialist and foundational positions. It is not so much a position as an anti-position or a meta-position. Relativist arguments emphasise the inescapable role of rhetoric in constructing claims as objective and foundational, and the contingency of tests, criteria, rules, experimentation and other procedures that are claimed to guarantee objective foundations. (Potter 2004c:951–2) The relativist argues that all knowledge (including ‘scientific facts’) is historically and culturally relative. It follows that there are no independent means of determining what is ‘true’ (Edwards et al. 1995). There are infinite possible ways of constructing the world, and an infinite number of ways in which the same events can be recounted and constituted. As such, there is no way of knowing whether any particular version offers an accurate description of ‘what really happened’, or whether it corresponds to what took place ‘in the real world’.

Since, as Wetherell and Potter (1992:62) put it, ‘there is no versionless reality’, a relativist approach urges us to take a critical stance towards all knowledge claims. It encourages us to question our commonsense assumptions about the way the world works, and particularly about those objects, events and categories that are presented to us as ‘naturally occurring givens’—including ‘scientific facts’ about biology, subjectivity and mind. Indeed, this approach ‘is fundamentally anti-intuitive: it specifically aims to deconstruct those things we ‘just know’ on the basis of personal experience or introspection’ (Kitzinger 1992:224).

Sex, for example, is one of our primary means for classifying and organizing the world, and sexual dimorphism, the idea that the world consists of two and only two sexes, is a distinction upon which heterosexuality depends (Butler 1990a). While it may seem obvious, or even sensible, that we divide the world up according to male-female ‘reproductive differences’, a relativist would question whether there is anything intrinsic to the nature of persons that requires us to divide the world up in this way—and to develop a whole host of institutionalized practices based on gender dualism (e.g., sex- demarcated toilet cubicles, sports teams, leisure pursuits, bicycle frames, gym shoe colours, dress codes, children’s toys, and so on). The relativist might ask why we cannot equally organize the world on the basis of some other classification system based on categories of weight, height or eye colour, for example (see Burr 2003:3). Thus, just because biological ‘sex differences’ are made to be important in modern western societies, it does not mean that they must inevitably be that way. Institutionalized gender demarcation is a choice rather than an inevitability. For the feminist relativist, the goal of social scientific inquiry is not ‘truth-seeking’. Rather, the focus shifts to an examination of how certain truths, ways of seeing the world, or versions of reality are constructed, and Gender talk 16 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

the social practices that sustain some oppressive, sexist and heterosexist versions over others.

A relativist approach ultimately leads us to reflexivity, where the researcher must recognize that their own practice of constructing knowledge and studying the world is itself socially constructed and organized. From this perspective, the researcher must self- consciously acknowledge and pay attention to their own role in practices of knowledge construction—for example how the role of the researcher and their methods for collecting gender talk in part shape the nature of the gender talk obtained (Speer 2002d). I will explore these issues further in Chapters 4 and 7 and in the Postscript.

Together these five features generate a distinctive feminist analytic approach. My aim in the rest of this book is to demonstrate what this approach might look like, and how and why it offers the most productive form of analysis for feminism.

Chapter overviews In Chapters 2 and 3 I introduce some important work on the relationship between gender, language and identity that has developed within sociolinguistics, poststructuralism and ethnomethodology. This research underpins, and forms a precursor to, much research on gender and language, and can be broadly divided into two strands or types: the first strand of research, which has been conducted primarily within sociolinguistics, explores sex differences in language. The second strand of research, which has been conducted primarily within poststructuralism and ethnomethodology, explores how gender identity is constituted—how people ‘do’ gender. Research conducted from within these two traditions represents often competing theoretical and methodological assumptions about the nature of the relationship between gender and language, and how it might best be grasped analytically. My aim in Chapters 2 and 3 is to provide a focused overview and critique of research conducted within both perspectives, with a view to illustrating the distinctiveness of my own position.

I begin, in Chapter 2, by interrogating the theoretical and methodological assumptions that underpin three classic studies within the socio-linguistic ‘sex differences’ paradigm. I argue that even while work within this tradition has had a significant impact on the field of research on gender and language, that the sex differences framework nonetheless suffers a number of problems when it comes to the empirical analysis of gender and language. Specifically, this research has a tendency to reinforce dualistic understandings about sex and gender, and take us further away from, rather than closer to, an understanding of how gender and sexism are constituted and reproduced in interaction.

In Chapter 3, I contrast the ‘sex differences’ approach to gender and language considered in Chapter 2, with some important work on ‘doing’ gender that developed concurrently outside linguistics within poststructuralism and ethnomethodology. This work, while concerned more with gender identity than with language, challenges us to see sex and gender as a performance, or an accomplishment, rather than an essence.

Poststructuralism and ethnomethodology have had a significant impact on the development of the two discourse strands that I identified earlier in this chapter: the ‘critical’ discursive work informed by critical theory, poststructuralism and psychoanalysis on the one hand, and the discursive work informed by ethnomethodology Introduction 17 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

and CA on the other. By inspecting the theoretical and methodological assumptions underpinning poststructuralist and ethno-methodological studies of gender identity construction, I begin to flesh out some of the similarities and differences that underpin these two discourse strands, and the rather different feminist analyses of gender and language that derive from them. I suggest that despite the radical political potential this poststructuralist and ethnomethodological work has offered feminism, that neither of these perspectives, in and of themselves, offer an adequate empirical programme for the analysis of talk-in-interaction. Instead, I argue that an approach which draws on insights from CA, and the closely related constructionist perspective of DP, is the most fruitful analytic framework for feminism.

In Chapter 4 I demonstrate how insights from CA and DP contribute theoretically and methodologically to the five criteria for the development of an adequate feminist analytic approach set out in this chapter. I show how feminists and others with a critical agenda have used insights from both perspectives to interrogate the relevance of gender in talk, and to explore ‘what counts’ as gender, or an ‘orientation to gender’ in an interaction. I suggest that these studies challenge both the essentialist gender-typing story associated with the vast bulk of variationist and sociolinguistic research conducted to date, and the ‘broader’ forms of analysis associated with poststructuralist, ‘top-down’ approaches to discourse. I argue that an approach which is concerned with the turn-by-turn analysis of taped and transcribed segments of talk, provides the tools with which we may begin to produce an empirically grounded form of feminism, and ultimately to rethink what we mean by ‘gender-talk’.

This discussion sets the scene for Chapters 5 and 6, where I argue against the anti-CA critics, to show how an analytic approach which sticks closely to an analysis of participants’ orientations, can be applied to, and used to rework our understanding of, the relationship between the purportedly ‘external’ macro-social structural context of gender norms, ideology and ‘the world out there’, on the one hand (Chapter 5), and the purportedly ‘internal’, cognitive-psychological context of gender identity, subjectivity, prejudicial attitudes and beliefs, and ‘the world in here’, on the other (Chapter 6). These chapters demonstrate how traditional sociological and psychological understandings of the relationship between macro-social and cognitive-psychological realms (particularly as they are formulated in ‘critical’ and poststructuralist approaches to discourse) can be rethought using a micro-interactional framework, and without losing a political thrust.

I make this case in Chapter 5 by reviewing Wetherell and Edley’s (1999) critical discursive work on the widely used ‘macro-analytic’ concept of hegemonic masculinity.

Using data from two informal interviews with men in their early twenties, I explore how participants construct masculinity and situate themselves and others in relation to those constructions. I ask whether participants themselves orient toward something that analysts have glossed—in more abstract, theoretical contexts—as hegemonic masculinity, and consider what such orientations may be doing interactionally. Specifically, I look to see just how far an approach which does not go beyond participant orientations can take us in our understanding of the discursive constitution of gender identity. I argue that even while participants may align with, and differentiate themselves from, a version of masculinity that they define in similar ways across extracts, that hegemony and hegemonic masculinity are not participants’ categories, and that, in its particularities, Gender talk 18 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.

masculinity is defined in variable ways that are appropriate for the local interactional context, and the work that needs to be done to invoke or manage a particular identity.

In Chapter 6, I show how a participants’ orientations approach can be applied to rethink ‘cognitive-psychological’ understandings of prejudice and—specifically— heterosexist talk. I begin by criticizing current psychological work on heterosexism, highlighting the way its operationalization tends to obscure flexible discursive practices and settle them into stable, causal attitudes within individuals. Then, drawing on extracts from a variety of sources where sexuality is made relevant, I examine whether participants themselves orient to their talk as heterosexist or problematic in any way, and consider what such ‘attending to’ may be doing interactionally. I suggest that speakers use various conversational and interactional resources to manage the potential trouble that their remarks may engender, and to foreclose possible counter-arguments and challenges. Importantly, I argue that it is in this very management—this attending to potential trouble—that we can find the constitution of what the participants take to be prejudicial, accountable ‘heterosexist talk’. Finally, I show (cf. much critical discursive and ‘sex differences’ research) that heterosexist utterances do not have their negativity built into them, but become prejudicial, troublesome or otherwise for participants in situ, as their sense is produced and negotiated.

In Chapter 7 I summarize some of the main themes to come out of the book, and consider their implications. My discussion is framed in terms of several questions that readers may still have about the five key features of the feminist analytic approach I advocate. My responses to these questions point to some common misunderstandings of both CA and DP. I explain why I believe my approach is a theoretically and methodologically fruitful one for feminism, and how, contrary to popular belief, it may be politically and practically consequential.

Finally, in the Postscript, I consider some methodological issues that will influence the future development of feminist CA, and which researchers interested in using this perspective might usefully consider. Introduction 19 Speer, S. A. (2005). Gender talk : Feminism, discourse and conversation analysis. Taylor & Francis Group.

Created from umanitoba on 2022-01-31 03:03:40.