Third Wave Feminist Linguistics and the Analysis of Sexism

Sara Mills

School of Cultural Studies, Sheffield Hallam University, Collegiate Crescent, Sheffield S10 2BP

Abstract: This paper critically examines Third Wave feminist linguistics, a form of anti-essentialist analysis which challenges Second Wave feminist linguistics' analysis of  the language of women and men as homogeneous groups.  Rather than assuming that men and women necessarily speak in different ways, men being direct and forceful, women being hesitant, polite and apologetic, a Third Wave feminist linguistics analyses the complex negotiations undertaken by  women and men with gendered domains (those sets of linguistic routines or contexts which appear to be gendered, for example  public speaking, intimate conversation), and  gendered stereotypes of what it is assumed that women and men should do (that is, women should be co-operative,  men should be competitive). In this way, Third Wave feminist analysis makes it possible to analyse the language use of women and men, without assuming that all women are powerless, all males are powerful, or that gender always makes a difference. Thus Third Wave Feminist linguistics examines, for example, the language of women who adopt primarily masculine forms of speaking in the public sphere. However, rather than just focusing on the individual, this form of analysis also examines the  role of context and social forces on the individual, in that these ways of speaking may be judged by others as incompetent, aggressive, unprofessional and unfeminine. Third Wave feminist linguistics is therefore concerned with moving the analysis of gender and language  away from the individual alone towards an analysis of the individual in relation to social groups who judge their linguistic behaviour and also in relation to hypothesised gendered stereotypes. However, this article does not wholeheratedly advocate the adoption of a Third Wave feminist perspective. It seems that within this type of analysis sexism becomes difficult to analyse or challenge, and this I suggest that  rather than seeing Second and Third Wave feminist linguistics as chronological, they need to be seen more as approaches which may be more or less appropriate depending on the context and social situation. In the case of sexism, for certain types of sedimented sexism a Second Wave feminist approach is more applicable, whereas in others a more locally-oriented and context-specific Third Wave approach is preferable. Thus Second Wave feminism needs to be integrated into Third Wave feminist linguistics, so that both local and global issues can be addressed.


Keywords: Third wave feminism; Second wave feminism; linguistics; sexism; stereotypes

1.0 Introduction

In this article, I contrast Second and Third Wave feminist linguistics, broadly speaking, Second Wave feminism focusing on the language of women as a subordinated group and Third Wave feminism challenging the homogeneity of women as a group and focusing instead on localised studies. [1]   I  challenge the notion that these forms of analysis are simply chronological so that Third Wave feminism supersedes and supplants Second Wave feminism; rather I argue that Third Wave feminism is best seen as a development from Second Wave feminism which nevertheless depends on the basic framework of Second Wave feminism for its theoretical integrity. [2] In order to contrast the way in which these two approaches work and to demonstrate that each tendency can be put to work in particular contexts, I examine the way the difficulties which each approach finds with the analysis of sexism.

2.0 Third wave feminist linguistics

The term Third Wave feminism has developed relatively recently to describe a form of analysis which is critical of  Second Wave feminism. [3]  It seems to be part of a wider postmodernist-influenced theoretical position where `big stories are bad, little stories are good', but, unlike some other forms of analysis, such as post-feminism, it locates itself within a feminist trajectory (Potter, 1996).  Second Wave feminism has achieved a great deal: feminist campaigning and consciousness raising in the 1960s and onwards have changed attitudes to the role of women and have resulted, in  Western Europe and the US, in equal opportunities legislation, greater access to work within the public sphere, access to childcare,  access to contraception and abortion.  However, this campaigning was largely focused on the needs of straight white middle class women. [4] The linguistic work which stemmed from Second Wave feminism focused on the stereotypical speech of these same women and made generalisations about all women's language on the basis of anecdotal evidence (Spender, 1980; Lakoff, 1975).  Thus, women were assumed to be oppressed in similar ways by men and by a patriarchal social system; research drew attention to the way in which women's use of language exhibited powerlessness.  Lakoff and Spender characterised women's speech as hesitant, deferent and polite  and suggested that elements such as tag-questions and back-channel behaviour were more likely to be found in the speech of women than in men, and that men interrupted women more than vice versa. [5]   Deborah Tannen challenged this work by suggesting that women and men's speech was characterised by a difference in style rather than a difference in power and that misunderstandings occurred between men and women because women try to establish empathy with their interlocutors in speech through the use of what Tannen terms `rapport talk', whilst men try to establish a place for themselves within a hierarchy, through the use of information-laden talk, what Tannen terms `report-talk' (Tannen, 1991). [6]   Lakoff, Spender and Tannen's  Second Wave feminist research assumed that women's and men's language are necessarily different even though they often disagreed as to the cause of that difference.  This focus on global gender differences has been criticised by a number of  feminist linguists who have suggested that what is needed is a form of analysis which is less focused on the individual woman or man  and trends of speech in the  society as a whole, and more focused on  the way that context and individual mutually shape the way that interaction takes place (Troemel-Ploetz 1998; Bergvall et al. 1996; Bucholtz, 1999).  These critiques have led to a new form of feminist linguistics which  seems to share certain tendencies.  In the following section, I examine these shared characteristics.

Third Wave feminist linguistics does not assume that women are a homogeneous grouping and in fact stresses the diversity of women's speech.  For example, Penny Eckert  analyses the differences between the language use of different groups of girls in a high school in America, drawing on the categories and groupings that they themselves use, such as `jocks' and `burnouts' (Eckert, 2000). Mary Bucholtz  and Nancy Henley analyse the way that Black American women's speech does not necessarily accord with the type of speech patterns described by Lakoff and Spender, since there are different linguistics resources available,  signalling potentially different affiliations (Bucholtz, 1996; Henley, 1995) The essays in the collections edited by Bergvall et al (1996) and Coates and Cameron (1988) all stress the way in which women's language differs according to context and factors such as class, ethnic  and regional affiliation. Even the notion of the status of the variable itself has been questioned; for example, Mary Bucholtz has argued that in Second Wave feminism  `locally defined groupings based on ongoing activities and concerns were rarely given scholarly attention; if they were, members were assigned to large scale categories of gender, race and ethnicity and class' (Bucholtz, 1999:8). In contrast, in Third Wave feminism, these large scale categories are now questioned, so that rather than gender being seen as a stable unified variable, to be considered in addition to  race or class, gender is now considered as a variable constrained and constituted by them and in turn defining them in the context of local conditions. Indeed, feminist linguistics now seems to have turned away from these more established identity categories to an analysis which focuses on ` a whole set of identity features (being a manager, someone's mother, a sensible person)' which might be potentially relevant (Swann, 2002:49) Furthermore, identities are now seen as plural and potentially conflicting even within a specific individual in a particular interaction.  Third Wave feminist linguistics does not make global statements about women's language but rather focuses on a more punctual analysis, that is one which can analyse the way that one's gendered identity varies from context to context.  However, Swann has argued that this contextual focus in relation to variables has almost invalidated the notion of the variable; she argues `if gender identity is something that is done in context, this begs the question of how an analyst is able to interpret any utterance in terms of masculinity (or working class, white, heterosexual masculinity).  How does an analyst assess whether a speaker is doing gender, or another aspect of identity?' (Swann, 2002:48)  What Swann goes on to argue is that rather than seeing Third Wave (or as she terms it Postmodern) feminism as a simple reaction to  Second Wave feminist linguistics, we need instead to see the way in which Third Wave feminism depends on early feminism; the contextualised studies are interesting `partly because they qualify, or complexify, or introduce counter-examples' (Swann, 2002:60).  Thus, the localised studies should be seen against the background of the earlier global (and problematised) claims of Second Wave feminism, which they can perhaps help to modify and temper.

Much Third Wave feminist linguistics draws on the work of Judith Butler, particularly the notion of performativity (Butler, 1990;1993; 1997). [7]   Gender within this type of analysis is viewed as a verb, something which you do in interaction, rather than something which you possess (Crawford, 1995). Gender is constructed through the repetition of gendered acts and varies according to the context.  In many readings of Butler's work, gender is seen almost like a set of clothes that one puts on - the individual chooses what sort of identity they would like to have and simply performs that role.  However, it is clear that institutional and contextual constraints determine the type and form of identity and linguistic routines which an individual considers possible within an interaction. [8]   Whilst Second Wave feminist linguistics assumed that gender pre-existed the  interaction and affected the way that the interaction developed, Third Wave feminists focus on the way that participants in conversation bring about their gendered identity, thus seeing gendering as a process. [9]   This focus on the orienting of participants to gender is clearly influenced by heated debates between  Conversation Analysis and Critical Discourse Analysis about whether extra-textual factors such as gender and race can be considered if they are not specifically addressed by participants (Schegloff, 1997; Wodak and Meyer, 2001; Mills, forthcoming)  However, it could be argued this more process-oriented feminism still has a very clear notion of what gender is, bringing that pre-constructed notion to their analyses of the way that participants orient to gender within interactions (Mills, forthcoming; Swann, 2002) I now discuss certain issues or shared concerns which seem to define a Third Wave feminist linguistics: meaning; power; questioning the notion of a women's language; the relation of the individual and the social.

2.1 Meaning

Second Wave feminist linguistics was concerned with analysing the inherent meanings of words and often made statements about the abstract meanings of words, constructing dictionaries of sexist language and advising on the avoidance of certain words (Kramarae & Treichler, 1985; Miller and Swift,1981). There was also a tendency to assume that certain words or ways of speaking were simply more powerful than others; thus, interrupting was seen as a powerful strategy, and hesitating was seen to be a powerless strategy.  After Cameron's work on the multifunctionality of tag-questions and Michael Toolan's work on the difficulty of assigning clear functions to specific formal features, the notion that there was a clear link to be made between power and, for example, talking time was made more problematic (Cameron; Toolan, 1996; see for a discussion Thornborrow, 2002) Third Wave feminist linguistics focuses on the way that words are made to mean in specific ways and function to achieve certain purposes in particular contexts (Christie, 2001).  Thus, rather than discussing oppressive global social structures  such as patriarchy, Third Wave feminists analyse the way that gender and conflict are managed by women at a local level  (Cameron, 1998) . It is still possible to refer to structural inequality and to highlight instances of discrimination, but Third Wave feminist linguistics is more concerned with variability and resistance than on making global statements about the condition of women in relation to language use. Thus, whilst a Second Wave analysis might focus on the use of the generic pronoun `he' to refer to both men and women, or the derogatory terms used to describe women such as `bitch' or `slag', a Third Wave feminist analysis might focus on the way that within a particular context, a certain hesitation and ironic intonation might be considered to be  sexist when articulating the word `chairperson' to describe a female chair.  However, whilst this local focus helps women to describe practices which discriminate against them, Third Wave feminists find it difficult to refer to global, structural and systematic forms of discrimination.

Rather than meanings being imposed on women, Third Wave feminists consider meanings to be  co-constructed.  Thus within particular contexts, women and men engage in the contestation and affirmation of particular types of practices and interpretations.  What something means in a particular context is the result of  the actions of all of the individuals concerned,  negotiating with the institutional constraints of status and institutionalised linguistic  routines. For example, Joanna Thornborrow, in her analysis of an interview between a woman and two police officers, where the woman claims that she has been raped and the police try to throw doubt on the veracity of her claim, by suggesting that she is mentally ill, the woman plays an active role in contesting their assertions (Thornborrow, 2002).  A Second Wave  feminist analysis would analyse this interaction as the police oppressing and silencing the woman; however, this woman seems to have accrued to herself a certain amount of what I have called interactive power, that is, she has drawn on  linguistic resources which were available  within that particular context, using questions and rebuttals to challenge her characterisation by the police as an untrustworthy person (Mills, forthcoming). Ultimately, however, the police officers' version of events seems to be the one which holds sway, even though the woman's interventions are important in defining the way that the interview takes shape -  the institutional status of the police officers plays an crucial role in their version being seen as the `truth'. (see also, Potter, 1996)  We cannot see this woman as simply powerless as a Second Wave feminist analysis might have done. However a Third Wave feminist analysis does not seem to be able to argue for a change in the way that police interviews are carried out, or call for training for police officers in the type of language which it is appropriate to use with rape victims.

2.2 Power

Most Third Wave feminists have been influenced by Michael Foucault's theorisation of power (Foucault, 1978). Power is seen as a net or web of relations not as a possession; thus power is enacted and contested in every interaction (Thornborrow, 2002). Power becomes a much more mundane, material and everyday element rather than something abstract and intangible which is imposed from above.  Thus, there is now a concern with the local management of  power relations, the way that individuals negotiate with the status which they and others  have been allotted or which they have managed to achieve, and which within particular contexts they can contest or affirm, through their use of language and through their behaviour. Many feminist theorists draw a distinction between institutional status (that is the status that you are allocated through your position within an institution) and local status (that is the position that you manage to negotiate because of your verbal skill, confidence,  concern for others, `niceness' and so on ) and whilst these are clearly interconnected, it is now often the local status which is focused on by feminist theorists (Manke, 1997; Diamond, 1996) . However, this move away from the analysis of institutional rank to that of local status, whilst important in challenging the characterisation of women as powerless speakers, means that feminists no longer concern themselves so much with the way that institutional rank and gender relate, and the way that the basis on which local rank is negotiated may be heavily determined by stereotypes of gender and gendered practices. This means that the analysis of the speech of men in positions of authority will only focus on the way that their speech is negotiated at the local level and will not consider the way that particular styles are authorised with reference to factors outside the local context.

2.3 Questioning the notion of a `woman's language'

Second Wave feminists assumed that all women were more deferent, polite, more concerned with the welfare of others and more co-operative. Third wave feminist linguistics suggests that this type of speech style is perhaps only available to a very small number of white middle class women, and even then only within very specific contexts. Holmes (1995) and Coates (1998)  also call for a re-evaluation of  co-operative speech styles and question whether they necessarily denote powerlessness; both argue that concern for others in speech should be valued and Holmes in particular claims that women's greater use of positive politeness within the work environment leads to more productive discussions.  However, whilst all women are not powerless, we have to accept that, for many people, powerlessness and deference are stereotypically associated with women and therefore when women speak assertively, their speech will be considered aberrant and aggressive because they are judged against a stereotypical norm of deference.  This is why many women, as Crawford has shown, rather than being assertive, decide to temper their speech by using politeness strategically: `"unassertive" speech, rather than being a (female) deficiency in social skills, may reflect a sensitivity to the social impact of one's behaviour.  Tentative and indirect speech may be a pragmatic choice for women.  It is more persuasive, at least when the recipient is male, less likely to lead to negative attributions about personality traits and likeability, and less likely to provoke verbal attack' (Crawford, 1996:68)  Thus, rather than asserting that women are more  polite or indirect than men, Third Wave feminist linguists argue that women engage in a complex process whereby they assess others' stereotypical beliefs about gender and then strategically adopt strategies which will be most likely to achieve their ends; some of those strategies may be ones stereotypically associated with feminine language. Third Wave feminists tend to avoid discussing the notion of stereotype since it is one of those problematic concepts often drawn on unproblematically by Second Wave feminists in their analysis of the workings of a global gendered ideology; however, it is clear that a hypothesised stereotype of gendered behaviour informs interaction.  For example, Queen has shown that lesbian speech is often produced in a parodic,  ironic playing with hypothesised stereotypes of `straight' feminine speech and masculine speech (Queen, 1997)  Halberstam, in her analysis of masculine women has  tried to prise apart the relation between masculinity and men and has shown that, rather than stereotypes being fixed and either accepted or rejected by individuals, they are played with, parodied and used for particular strategic ends, and in the process of being changed and ironised  by individuals they are inevitably globally changed  (Halberstam 1998).

Because of the change in focus in relation to power, there has been a move away from the analysis of subordinated women. Mary Bucholtz argues that in the past :`much of the scholarship in language and gender has been what might be called "good-girl research" - studies of "good" (that is normatively female - white straight middle class) women being "good" (that is normatively feminine)'. (Bucholtz, 1999:13)  Now rather than analysing women's indirectness or lack of assertiveness, many linguists focus on strong women speakers and women's resistance to masculine forms of speech, such as interruption or aggressiveness (Mills, 1999)  Clare Walsh  has analysed the language of women working within masculinist or male-dominated environments, for example, women priests,  MPs and environmental campaigners. (Walsh, 2000) She has  found that women within institutions are often viewed very negatively and if they use direct, confrontational language they are often criticised (Walsh, 2001).  Sylvia Shaw (2002) has also analysed the language use of women MPs and has shown that whilst women are very able to adopt the type of aggressive formalised Parliamentary debating techniques which have been developed by male MPs, they may be judged differently to men when they do so. She has also shown that women MPs tend to adhere to the speaking rules very strictly, observing Parliamentary forms of address, protocol and etiquette, whereas the male MPs often manage to achieve certain advantages for themselves by breaking the rules. Marjorie Harness Goodwin has analysed girls' language in play and has contested the notion that girls' language is necessarily more co-operative or nicer than boys', showing that girls use direct and confrontational language (Goodwin, 1998).  In her most recent work, she argues that it is expertise, for example in play, which determines who uses assertive language, rather than sex difference.  Bonnie McElhinny has analysed the language of women police officers in Pittsburgh and found that they  feel obliged to adopt particular masculine ways of speaking simply to appear to be doing their job in a professional way (McElhinny, 1998).  They adopt what she calls `an economy of affect' because disinterestedness is demanded of police officers by the public, since it signifies authoritativeness and impartiality.  McElhinny argues `that women who move into powerful and masculine institutions sometimes adopt the interactional behaviour characteristic of these institutions might disappoint some feminists.  But it seems clear that who we think can do certain jobs changes more rapidly than expectations about how these jobs should be done. The process by which women enter a masculine workplace necessarily includes some adoption as well as adaptation of institutional norms.' (McElhinny, 1998: 322) Thus, all of these studies suggest that women, when entering primarily masculine environments, adopt the language styles prevalent in those institutions, and those styles themselves are both an indicator of masculinity and also of professionalism.  McElhinny states that ` masculinity is not referentially (or directly) marked by behaviours and attitudes but is indexically linked to them (in mediated non-exclusive probabilistic ways)' (ibid). Alice Freed, in her analysis of the language styles of intimate conversation,  suggests that masculinity and femininity should be seen as a characteristic of the context or situation, rather than an attribute of individuals (Freed, 1996) She argues that intimate self-disclosing conversation is associated with stereotypical femininity and therefore when males engage in such conversations, they may tend to display the same `feminine' speech styles as women. [10] Thus, these Third Wave feminist analyses are interested in analysing the way that masculinity and femininity can be seen to exist at an institutional level, linked in some ways to particular institutional contexts rather than simply at the level of the individual and can be associated stereotypically with attributes such as professionalism and competence.

There are certain contexts, however, where women do seem to have brought changes into the predominantly masculine norms in institutions.  Wendy Webster's analysis of Margaret Thatcher's speech styles demonstrates that rather than simply adopting the speech norms associated with the role of Prime Minister, Thatcher integrated more feminine elements into her overall  style, incorporating elements of self-disclosure and informality with a more public authoritative discourse (Webster,1990)  Thus, women's negotiations with the speech norms of the context within which they find themselves should not be seen as simple capitulation to dominant forms.  However, generally as yet these feminine forms do not seem to have been adopted by male politicians. [11]   Furthermore, we should see women's adoption of masculine dominant forms as strategic and perhaps argue that women's adoption of positions of institutional status may result in the use of language styles which are characterised by a different approach to `doing power'.  Thus, as Diamond has argued in her analysis of group dynamics in a group of psychotherapists, in certain contexts, those in positions of institutional authority in fact do not use direct commands and assertiveness, preferring to use indirectness (Diamond, 1996). [12] Third Wave feminist linguistics forces us to reconsider the way that we think that power is exercised through language, but perhaps does not enable us to describe adequately the way that rank within an organisation may influence our localised interactions.

2.4 The Relation between the Individual and the Social

For many Third Wave feminist linguists, the notion of the community of practice has been important in terms of trying to describe the way that group values affect the individual and their notion of what is linguistically appropriate (Eckert and McConnell-Ginet, 1999; 1998).  The community of practice is a group of people who are brought together in a joint engagement on a task and who therefore jointly construct a range of values and appropriate behaviours, for example, a community of practice might be a group of people who meet to plan an event, or a group of people who go out drinking together. Thus, rather than focusing on the role of an oppressive social system, ideology or patriarchy in relation to individual linguistic production and reception, Third Wave feminists focus on the interaction at the level of the community of practice.  Individuals hypothesise what is appropriate within the community of practice and, in speaking, affirm or contest the community's sense of appropriate behaviour.  In this sense, one's choice of words and one's speech style,  can be seen as defining one's position within a group or community of practice.  Bourdieu's notion of `habitus' has also been extensively drawn on by Third Wave feminist linguists:   'habitus'  is the set of dispositions which one draws upon and engages with in order to perform one's identity through discourse (Bourdieu, 1999). This set of attitudes or practices which are seen as constituting a norm by individuals are then discursively negotiated by individuals in terms of their own perception of what is acceptable for their own behaviour within a particular community of practice. Gino Eelen, drawing on Bourdieu's work, argues that we assume that there is a common world, that is, a set of beliefs which exist somewhere in the social world and which are accepted by everyone, which we as individuals need to agree with or contest:   'On the one hand, collective history creates a "common" world in which each individual is embedded. On the other hand, each individual also has a unique individual history and experiences the "common" world from this unique position. The common world is thus never identical for everyone. It is essentially fragmented, distributed over a constellation of unique positions and unique perspectives' (Eelen, 2000: 223).   Thus, this view of the relation between individuals and others moves us significantly away from notions of  society as a whole influencing the linguistic behaviour of individuals to an analysis of the way that at a local level, individuals decide on what type of language and speech style is appropriate.  This local focus of Third Wave feminism is one of its benefits, but it does make it extremely difficult to discuss the impact of the values and pressures of the  wider society; talking about society above the level of the community of practice is almost impossible, and it is clear that the wider society as a whole needs to be discussed in terms of the impact it has on practices within communities of practice. Third wave feminist linguistics tries to maintain a balance between a focus on the local and an awareness of the negotiations at the local level with structures which are largely imposed. Mary Bucholtz characterises the concerns of Third Wave feminism within the following themes: `that language users' identities are not essential to their natures but are produced through contingent social interactions; that those identities are inflected by ideologies of gender and other social constructs; that speakers, writers and signers respond to these ideologies through practices that sometimes challenge and sometimes reproduce dominant beliefs; and that as new social resources become available, language users enact and produce new identities, themselves temporary and historical, that assign new meanings to gender' (Bucholtz, 1999: 20).  However, perhaps this quotation draws our attention to the difficulties encountered by Third Wave feminist linguistics since it does not seem possible to maintain both a focus on contingent social interactions and wider societal notions such as ideologies of gender, without some fundamental rethinking of our models of language and gender.  I'd like now to move to a testing out of some of these ideas about Third and Second Wave feminism by analysing the way that they can be brought to bear on the analysis of sexism.

3.0 The Analysis of Sexism

Because of this move away from the top-down model of Second Wave feminism, Third Wave feminism finds it difficult to discuss sexism, since sexism as a concept is based on the idea that discrimination against women is systematic and that sexism is imposed on women by those in positions of power, is ingrained in social structures and works to the benefit of all men. Sexism as a topic of analysis is distinctly unfashionable at the moment and has a slightly anachronistic feel to it.  Sexist language was broadly defined within 2nd wave feminism as the use of statements which 'create, constitute, promote or exploit an unfair or irrelevant distinction between the sexes', (Vetterling-Braggin, 1981: 3). Thus, studies of sexism concerned themselves with the use of the so-called generic pronoun `he' to refer to both males and females, and the use of the form `lady' or `female' with generic nouns such as `doctor' when they are used to refer to females.  This type of analysis showed that there was a systematic tendency within English and other languages to assume that males were the norm and to associate women with trivial, sexualised or non-serious topics (Cameron, 1998; Pauwels, 1998). [13]   Many Feminist analysts of sexist language argued that this type of language use should be reformed to reflect the changes in women's position in society.  However, some feminists  questioned this  determinist position and suggested that perhaps sexist language did not itself determine women's oppression; reform of the language alone would not alter the way that women were treated. [14]   Feminists drawing on social constructionism argued that changes in women's position would lead to a change in the way language was used.  Neither of these views is accurate, as it is clear there is a complex  dialectic process going on in language, whereby language items both affirm and contest the status quo, and changes in social structures necessitate the development of new vocabulary and forms of expression. Language is a site where challenges to the status quo through challenges to sexism can take place and these changes at the local level may lead to changes in the overall meanings of words and also wider changes at a societal level.

I would like now to discuss the ways in which analysing sexism within Third Wave feminism has been made more complex,  and to analyse the reasons  that sexism has become difficult to discuss.  One of the major factors in the current difficulty in discussing sexism is the result of very effective feminist campaigns over language: in the public sphere, sexism is often viewed by employers and employees to be incompatible with equal opportunities in the workplace. Publishing houses, trades unions, public corporations, public service providers, universities and so on, have issued guidelines on appropriate language. [15]   Feminists have developed alternative terms, so that instead of `chairman; 'chair', can be used.  Instead of referring to `air hostess' which some find demeaning, one can use  'flight attendant', and so on.  Cameron has argued that in fact by challenging the use of sexist words, 'the radicals have effectively politicised all the terms, so that, in any interaction, the choice of certain words will announce your political stance in relation to women' (Cameron, 1994b: 31). [16]   It is important that these feminist campaigns have led to language policies being adopted by institutions.  Whilst many of the policies on sexism and racism seem to have largely fallen into disuse, the fact that there is institutional support changes the status of an individual's complaint about language use (Pauwels, 1998).  But the very success of the campaigns to change the language used at work has meant that certain forms of sexism rather than being seen as neutral forms have become marked and associated with conservatism-  sexism thus seems to have been driven underground.  Therefore, rather than sexism being overt as in the past, sexism has become much more indirect. [17]

3.1 Indirect sexism

It seems as if there is now a certain instability within sexism itself, so that  whilst Second Wave feminism saw sexism as  a clearly defined set of practices which reflected a particular set of attitudes towards women, in fact now sexism has a range of meanings for different people. This makes sexism much more difficult to context. Now it seems that sexism in English is largely  "indirect sexism", that is, sexism which manifests itself at the level of presupposition, and also through innuendo, irony and humour, or which is prefaced by disclaimers or hesitation (Mills, 1998)  For example, in the British television programme Men Behaving Badly, the two central male characters use the term `top totty' to refer to women. This is such an exagerrated form of sexism that within the terms of the programme it cannot be objected to as sexist as it is intended to be humourous and tongue-in-cheek. Sexism at the level of presupposition is also much more difficult to challenge as Christine Christie has demonstrated, since it is necessary to make overt the assumptions upon which the sexism is based; the reason this indirectness is in fact chosen is to mask the sexism and to give the speaker the potential for denying any intended sexism (Christie, 2001). For example in the phrase `So have you women finished gossiping?' there are a number of presuppositions about women and talk which would need to be unpacked before the phrase could be responded to (for example, that women's talk is trivial, that women engage in gossiping more than men, that two women talking together can be assumed to be gossiping, and so on). Imelda Whelehan has described the difficulty for feminists of watching television programmes such as Men Behaving Badly and Never Mind the Buzzcocks,  or listening to radio programmes such as Radio 1's Chris Moyles, where overtly sexist statements are made about women, but because these statements are made in a very knowing ironic way, it is to be assumed by viewers that they are not taken to be as sexist, or at least not in any simple way (Whelehan, 2000). For many feminist viewers, not wishing to be seen as puritanical and lacking a sense of humour, there is little possibility of contesting these ways of presenting sexist ideas, even though sexism is still kept in play by these means. To give another example of the instability within sexism at the moment, we might consider the television advertisements for Yorkie chocolate bars.  The advertisements, following on from the association of Yorkie bars with truck-drivers, claim that Yorkies are `Not for Girls'; a woman disguised as a male builder with a hard hat and false moustache goes into a sweet shop and tries to buy a Yorkie bar.  The shopkeeper tries to test whether she is a man or not by asking her to define the off-side rule in football, to decide whether stockings or tights are better, and  finally, he manages to show that she is female because she responds to flattery. If this advert had been shown in the 1980s, the feminist response would have been clear; classifying the product as `not for girls', suggesting that women are not `man enough' to eat large chunks of chocolate would have been seen as sexist.  But this advert is playing with stereotypes; the woman is not disguised convincingly as a man; the advertisement  ridicules men as much as women, suggesting that men are obsessed with football and sex. So if we laugh at this advert  because we think it is ironising sexism,  we could be seen to be buying into sexism, i.e.  rejecting femininity and valuing masculinity, or if we don't laugh at the advert and take it as sexist, we could be seen as humourless and unable to see the overt playfulness  and critique in  the advert (Mills, 2003).

Added to this instability within sexism which results in difficulties countering sexism, there is also an instability within anti-sexism. Anti-sexist campaigns have been destabilised in recent years because of the existence of   "political correctness".  Many people feel that there is a confusion or overlap  between anti-sexism and "political correctness". To clarify, "political correctness" is often seen as an excessive concern for the sensibilities of minority groups (women, the disabled, lesbians and Black people) which is manifested in  a set of media-invented absurd, terms, (such as `vertically challenged' instead of `short' ; `follically challenged' for `bald' ; `personhole cover' instead of  `manhole/inspection cover') which no anti-sexist or anti-racist campaigners have argued should be adopted.  These are often  listed alongside  `Ms'  and `chairperson'  which feminists have campaigned to be adopted.  This overlap and confusion has led to an undermining of attempts to reform language; some argue  any intervention is impossible or politically inexpedient (Cameron, 1995).  It is necessary to distinguish  anti-sexist practices from "political correctness", which is an abstracted set of rules extrapolated by the media from these practices and generalised to absurdity. However, for  anti-feminists,  "political correctness" is perceived to be the same as anti-sexism and consists of a real set of rules which  should be challenged in the name of free speech (Matsuda, et. al. 1993).

To sum up, linguistic practices can only be interpreted as sexist in particular contexts but these local meanings depend on a notion of an outdated and highly problematic form of overt sexism against which these indirect sexist meanings are negotiated.  However, we must also differentiate between different types of sexist practice, so that some sorts of linguistic routines can be seen to be more sedimented than others, such as the use of the generic `he' pronoun to refer to men and women. [18]   It is only through the use of a Second Wave feminist analysis which can describe global systematic uses of language that these uses of language can be combated and changed.  In other contexts, where the sexism is a particularly local context-specific type, where for example, the sexism is ironic or difficult to generalise about, then a Third Wave feminist linguistic approach is more productive.  However there has to be a close relation between these different forms of analysis.  Whilst one demands a general campaigning and reform, the other demands a more local and immediate response.  Anti-sexist practices are therefore necessarily complex and feminists differ on what they see as the most effective way of dealing with those elements or practices which they consider to be discriminatory.  It is not possible to agree on what is sexist; in that sexism is an evaluation rather than an inherent quality there will be disagreement about what constitutes sexism. Vetterling-Braggin was one of the first to remark upon the fact that labelling someone's statements as sexist involves taking a moral position in relation to them and their beliefs, and may provoke a breakdown of relations with that person (Vetterling-Braggin, 1981). However, it is not quite as simple as this, since often sexism, anti-sexism and "political correctness" are hypothesised positions which we attribute to others and which then act on our own sense of what it is possible for us to do or say.  Thus, in forming our own assessments of  what is sexist, we try to map out the parameters of the beliefs of others which would allow our own beliefs to be acceptable (Volosinov, 1973). Rather than seeing sexism solely in terms of abstracted general sets of words where the sexism is considered to reside in the words themselves, we must be able to see that there  are also local interpretations and strategic responses to what is evaluated by participants as sexist.  Thus, rather than seeing Second and Third Wave feminist analysis as simply chronological, we might perhaps see them as each suited to particular types of sexism.  Second Wave analysis can analyse those sedimented forms of sexism which seem to be embedded within the morphology of the language system itself, whereas Third Wave feminism is better able to analyse the ambivalences and uncertainties about and within sexism, within particular contexts.

4.0 Conclusion

A Third Wave  feminist linguistics is an anti-essentialist analysis of the way that gender relations are negotiated within particular contexts, but it seems that there needs to be a concern with the way that those local contexts are themselves structured by societal constraints.  Some Second Wave feminist analysis perhaps overgeneralised about the nature of systematic language patterns such as sexism, but unless we consider the wider context, above the level of the community of practice, we risk formulating a feminism without politics. We therefore need to consider the possibility of, not necessarily a Fourth Wave feminism as Kaplan (2002) has suggested, but a form of analysis which combines the global concerns of Second Wave feminist analysis with the local concerns of Third Wave feminism, using perhaps modified quantitative analysis alongside more contextualised qualitative studies, as Swann (2002) has suggested.  Thus, instead of viewing these two positions as antagonistic, we will be able to see their complementarity, and draw on them to help us to theorise and analyse gender and language more adequately.


Bergvall, V. Bing, J. & Freed, A. eds. (1996)  Rethinking Language and Gender Research: Theory and Practice,  London, Longman. [cited] [cited]

Bourdieu, P. (1999) Language and Symbolic Power, London, Polity. [cited]

Bucholtz, M.(1999) `Bad examples: transgression and progress in language and gender studies,' pp. 3-20, in Bucholtz, M, et al. eds. Reinventing Identities: The gendered Self in Discourse, Oxford and New York, Oxford University Press [cited] [cited] [cited] [cited] [cited]

Bucholtz, M. (1996) `Black feminist theory and African American women's linguistic practice,' pp.267-290, in Bergvall, V, Bing, J. and Freed, A. eds. Rethinking Language and Gender Research: Theory and Practice,  London, Longman. [cited]

Butler, J. (1997) Excitable Speech: A Politics of the Performative, London, Routledge.

Butler, J. (1993) Bodies that Matter: On the Discursive Limits of Sex, London, Routledge

Butler, J. (1990) Gender Trouble: Feminism and the Subversion of Identity, London, Routledge.

Cameron, D. ed. (1998a)  The Feminist Critique of Language: A Reader, 2nd edition,  London, Routledge

Cameron, D. (1998b) `Is there any ketchup, Vera?': gender, power and pragmatics', pp. 435-455, in Discourse and Society, 9/4

Cameron, D. (1997) `Performing gender identity: young men's talk and the construction of heterosexual masculinity', pp.86-107, in eds. Johnson, S. and Meinhoff, U. Language and Masculinity, Oxford, Blackwell.

Cameron, D. (1995) Verbal Hygiene, London, Routledge

Cameron, D. McAlinden, F. and O'Leary, K. (1988) `Lakoff in context: the social and linguistic functions of tag questions', pp.13-26, in  Coates, J. and Cameron, D. eds. Women in their Speech Communities, Harlow, Longman [cited]

Christie, C. (2001) Gender and Language: Towards a Feminist Pragmatics, Edinburgh, Edinburgh University Press

Coates, J. & Cameron, D. eds. (1988) Women in Their Speech Communities, London, Longman. [cited]

Crawford, M. (1995) Talking Difference: On Gender and Language, London, Sage.

Diamond, J. (1996) Status and Power in Verbal interaction: a study of discourse in a close-knit social network, Amsterdam/Philadelphia, John Benjamins [cited] [cited]

Eckert, P. (2000) Linguistic Variation as Social Practice: The Linguistic Construction of Identity in Belten High, Oxford, Blackwell [cited]

Eckert, P. and McConnell-Ginet, S. (1999) `New generalisations and explanations in language and gender research', pp.185-203, in Language in Society, 28/2 [cited]

Eckert, P. and McConnell-Ginet, S. (1998) `Communities of practice: where language, gender and power all live,? pp.484-494, in Coates, J. ed. Language and Gender: A Reader,  Oxford, Blackwell.

Eelen, G. (2001) A Critique of Politeness Theories, Manchester,  St.Jerome's Publishing.

Fairclough, N (2000) New Labour, New Language, London, Routledge [cited]

Freed, A. (1996) `Language and gender research in an experimental setting,? pp.54-76, in Bergvall, V. Bing, J. and Freed, A. eds. Rethinking Language and Gender Research: Theory and Practice, Harlow, Longman. [cited]

Goodwin, M Harness (1998) `Games of stance: conflict and footing in hopscotch,'. pp.22-46, in eds. Hoyle, S and Adger C. eds. Kids Play: Strategic Language use in Later Childhood, Oxford and New York, Oxford University Press

Greater Manchester Police (2001) Mind our Language,

Greater Manchester Police  Appropriate Language Working Group (2000)  The Power of Language: A Practical Guide to the use of Language,

Greater Manchester Police (2001) Reporting a Hate Crime

Foucault, M. (1978) History of Sexuality: An Introduction, Vol. I,  Harmondsworth, Penguin.

Halberstam, J. (1998) Female Masculinity, London, Routledge [cited]

Henley, N. (1995) `Ethnicity and gender issues in language', pp.361-396 in Landrine, H. ed. Bringing Cultural Diversity to feminist Psychology, Washington, American Psychological Assn.

Johnson, S. and Meinhoff, U. eds. (1997) Language and Masculinity, Oxford, Blackwell.

Kaplan, E. Ann (2002) `Plenary paper: 9/11 and  feminist theory', Third Wave Feminism conference, Exeter University

Kramarae, C. and Treichler, P. (1985)  A Feminist  Dictionary, London, Pandora [cited]

Lakoff, R. (2000) The Language Wars, Berkeley, University of California Press,

Lakoff, R. (1975) Language and Woman?s Place, New York, Harper and Row [cited]

Lewis, R. and Mills, S. eds. (forthcoming) Post-colonial Feminist Theory:  A Reader, Edinburgh, Edinburgh University Press

Liladhar, J: (2000) `Making and remaking femininities', unpublished PhD thesis, Sheffield Hallam University [cited]

Manke, M. (1997) Classroom Power Relations: Understanding Student Teacher Interaction, New Jersey and London, Lawrence Erlbaum Associates [cited] [cited]

McElhinny, B. (1998) `"I don't smile much anymore": affect, gender and the discourse of Pittsburgh Police Officers', pp.309-327, in Coates, J. ed. Language and Gender: A Reader, Oxford, Blackwell. [cited] [cited]

Miller, C. and Swift, K. (1981) The Handbook of Non-sexist Writing for Writers, Editors and Speakers, London, Women's Press [cited]

Mills, S. (in progress) Third Wave Feminist Linguistics and the Analysis of Sexism Cambridge, Cambridge University Press.

Mills, S. (2003) `Caught between sexism, anti sexism and "political correctness", feminist negotiations with naming practices', Discourse and Society, vol 14 (1) pp.87-110 [cited]

Mills, S. (forthcoming,) Gender and Politeness. Cambridge, Cambridge University Press.

Mills, S. (2002) `Rethinking politeness, impoliteness and gender identity,' in eds. Litosseliti, L. and Sunderland, J. Gender Identity and Discourse Analysis, Amsterdam/Philadelphia, John Benjamins

Mills, S. (1999) `Discourse competence: or how to theorise strong women speakers,' in eds. Hendricks, C. and Oliver, K. Language and Liberation: Feminism, Philosophy and Language, Albany, NY, State University of New York Press [cited]

Pauwels, A. (1998) Women Changing Language, New York, Addison Wesley Longman. [cited] [cited] [cited]

Potter, J. (1996) Representing Reality: Discourse, Rhetoric and Social Construction,  London, Sage [cited] [cited]

Queen, R. (1997)  `I don't speak Spritch: locating lesbian language', pp. 233-242, in eds. Livia, A. and Hall, K. eds. Queerly Phrased: Language Gender and Sexuality, London, Routledge [cited]

Sadiqi, F. (2003) Gender and Language in Morocco, Leiden, Brill [cited]

Schegloff, E. (1997) `Whose text? Whose context?' pp.165-185, in Discourse and Society, 8/2 [cited] [cited]

Shaw, S. (2002) Language and Gender in the House of Commons, PhD thesis, University of London

Spender, D. (1980) Man Made Language, London, Routledge [cited]

Stryker, S. (2002) `Plenary paper; transgendered identity and feminism', Third Wave Feminism Conference, Exeter University

Swann, J. (2002) `Yes, but is it gender?', in eds. Litosseliti, L. and Sunderland, J. Gender Identity and Discourse Analysis, Amsterdam/Philadelphia, John Benjamins [cited] [cited] [cited] [cited] [cited]

Tannen, D. (1991) You Just Don?t Understand: Women and Men in Conversation, London, Virago [cited]

Thornborrow, J. (2002) Power Talk: Language and Interaction in Institutional Discourse, Harlow, Longman [cited] [cited] [cited]

Toolan, M. (1996) Total Speech: An Integrational Linguistic Approach to Language,  Duke University Press, Durham and London [cited]

Troemel-Ploetz, S. (1998) `Selling the apolitical?, pp.446-458, in ed. Coates, J. Language and Gender: A Reader, Oxford, Blackwell [cited]

Vetterling-Braggin, M. ed. (1981) Sexist Language: A Modern Philosophical Analysis, US, Littlefield Adams. [cited] [cited]

Walsh, C. (2001) Gender and Discourse: Language and Power in Politics, the Church and Organisations, London, Routledge.

Webster, W. (1990) Not a Man to Match Her, London, Women's Press

Wodak, R. and Meyer, M. eds. (2001) Methods of Critical Discourse Analysis, London, Sage


[1] Earlier versions of this article were presented as papers at the International Gender and Language Association Conference, Lancaster University, 2002, and at the Third Wave Feminism Conference, Exeter University, 2002.

[2] The term Third Wave feminism is not one which the theorists I discuss necessarily adopt themselves, but they seem to share certain similarities of focus and theoretical position which can be summed up as indicating a shared critique of certain aspects of Second Wave feminist work.

[3] Whilst the term Second Wave feminism is fairly uncontentious, referring to the largely liberal and radical feminism of the 1960s onwards which argued for the equality of women, the term Third Wave feminism is more contentious.  A recent conference at Exeter University (2002) on the subject of Third Wave feminism together with the work  of Mary Bucholtz (1999) and Janine Liladhar (2000) have convinced me that Third Wave feminism is a preferable term to postfeminism (which assumes implicitly that the aims of feminism have been achieved and that therefore feminism is largely irrelevant) and postmodern feminism  (which has difficulty formulating any notion of a political programme).

[4] We need to question the homogeneity of our current characterisation of Second Wave feminism. In a recent paper Susan Stryker (2002) argues that Second Wave feminism was a more diverse than most feminists acknowledge ; there was a great deal of dissent and alternative accounts of gender - for example, see work by Angela Davis and Chela Sandoval (both in Lewis and Mills eds., forthcoming)

[5]   Tag questions are questions such as `It's very hot in here, isn't it?' or `You don't want to go yet, do you?' which were characterised by Lakoff as less assertive and potentially more manipulative than direct commands or requests. Back channel behaviour consists of the use of `mmm', `hmm', and `uhuh'  by the hearer to signal support for the speaker during a turn at talk and to signal also that the hearer does not intend to take a turn. Lakoff and others claimed the women used more tag questions and more supportive back-channel behaviour.

[6] These can be seen to be analogous to Jennifer Coates' (1998) co-operative and competitive strategies.

[7] This is rather curious because many of the linguists who draw on Butler's work would be critical of the use of Speech Act Theory from which the notion of the performative is drawn.

[8] For example, in talking to your grandmother, because of an assessment of her judgement of the meaning of swearing, you may decide to swear less than you would in the company of your female friends (this depends on the grandmother and the friends).

[9]   This concern with process means that Third Wave feminist do not stress the elements within an individual's identity which do seem to be fairly consistent.  Since each individual does seem to develop linguistic habits and verbal tics which give a certain predictability to individual linguistic behaviour, this overemphasis on the variability of the individual might be seen as a disadvantage. Again it is a question of emphasis, and there are ways of integrating a concern with process without entirely dispensing with the notion that certain elements are relatively stable.

[10] However, in other contexts of intimate conversation, research has shown that males may in fact engage in combative displays of aggression and verbal play (see the essays in Johnson and Meinhoff, eds. (1997) especially  Cameron, 1997)

[11] It could however be argued that Tony Blair's speech style is a mixture of assertive conventional masculinity and a more informal feminine style, and this may be part of the planned projection of Blair's persona by his media advisors (Fairclough, 2000)

[12]  Those in positions of institutional power in Diamond's study can `afford' in a way to use indirectness because others interpret their indirectness in relation to their status; thus, their indirect suggestions may well be interpreted by others as having the function of commands, as Manke (1997) has shown to be the case when teachers use indirectness in their instructions to children in schools.  Thus, if we are truly to analyse the local context, we cannot simply ignore the force of institutional status as Schegloff and other conversation analysts insist that we do, paying attention only to the way that participants orient to status within the interaction (Schegloff, 1997)

[13] Cameron (1998a) argued that the elements which feminists identified as sexist were in fact of very different types and drawn from a range of different linguistic levels; some were semantic, some grammatical and some morphological, and therefore a single global anti-sexist reforming policy could not possibly be effective.

[14]  For a fuller discussion of this issue see my (2003) article ` Caught between political correctness, sexism and anti-sexism: feminist negotiations with naming practices'.

[15] For an exemplary policy, see the Greater Manchester Police's equal opportunities language policy (2000; 2001)

[16] However, it should be noted that `chairperson' is unerringly used for female convenors and not for males.

[17] For those languages with a gender system, where gender is a morphological feature of the language, such as in  French, German and Arabic, sexism is much more sedimented than it is in English; thus in French it is much more difficult to say `la Ministre' if you want to refer to a female minister, since the word for Minister is masculine.  Furthermore, the rule in these languages that you use a masculine pronoun and noun ending for plural nouns if there is a masculine and a feminine referent together  is one which causes great difficulty for feminist speakers.  There are similar problems with highly gender-inflected languages such as Arabic and Berber, as Sadiqi has shown (Sadiqi, 2003).  However, as Pauwels argues (1998) changes are taking place in all Western European languages at a morphological level rather than just at the level of semantics.  This type of sedimented sexism can only be contested using a Second Wave feminist analysis, and contrary to some Third Wave feminist assertions that reform of sexism is impossible, although change is difficult and slow, it is possible.

[18] These seemingly more sedimented forms of overt sexism are changing rapidly, so that although one does hear the use of the generic pronoun, terms such as `hostess', `spinster', `aviatrix', and so on seem anachronistic and many of my undergraduate students simply do not recognise these as examples of sexism which they might use or hear.