Abstract
Researching sexism is not only a controversial undertaking, but one that is rendered problematic due to the fact that many individuals are reluctant to name certain experiences or practices as ‘sexist’. In this article, I use a discursive psychology approach to transcend arguments as to whether certain experiences and practices should be understood as sexist, focusing instead on how, in the context of a research ‘conversation’, participants attempt to warrant their own interpretations of these processes. Using data from research conversations held with two policewomen, who present very different accounts of sexism, I argue that social facts, like sexism, possess an inherent interpretive duality: they can be understood, simultaneously, to be both objective and subjective experiences. The study illustrates that the resolution of competing reality claims (e.g. is sexism a ‘fact’ or is it ‘in the eye of the beholder’) depends upon the processes through which particular versions of reality acquire authority. This essentially political process is, I argue, critical for understanding the reproduction, resilience and endurance of social facts such as sexism.
Keywords
Introduction
An enduring problem with research into sexism in workplaces is that many individuals, including women themselves, are reluctant to name or label certain practices, actions and behaviours as sexist (Denissen, 2010; Gough, 1998). Of course, this problem is itself infused with a fundamental epistemological dilemma – whose perspective on the interpretation of social phenomena should be privileged? On the one hand, if we take an extreme position of relativism, whereby we accept that each perspective is as valid as any other, we are in danger of failing to expose or lay challenge to systems of oppression that operate to disadvantage and marginalize women in workplaces (Burr, 1998). On the other, if we insist on privileging our own perspective, arguing that certain phenomena have a universal meaning, we are in danger of alienating the very groups whose interests we are seeking to represent (Bourdieu, 1991). Additionally, we are failing to adequately grapple with the issue that social phenomena are inherently ambiguous, having slippery, contingent and partial meanings (Iedema et al., 2003).
In this article, I take the meaning and interpretation of social phenomena as a central analytic focus, but my interest is not in trying to show or establish whether some interpretations are more or less valid, but rather with how actors attempt to warrant (Gergen, 1990) their own versions of ‘reality’ within specific situated interpersonal encounters. I address two specific questions:
how do individuals construct accounts of sexism and what tactics do they use to convince their audiences that sexism does exist or, conversely, that it is ‘in the eye of the beholder’?
in what ways does the research interview influence the production of such accounts?
Using data from a study conducted more than 10 years ago within a UK police constabulary, I analyse two very different accounts of sexism: one in which it is constructed as an everyday reality for police women, and another in which it is constructed as ‘in the eye of the beholder’. I use a discursive psychology approach to draw attention to differences in how actors make use of the discursive resources that are available to them in constructing accounts of sexism. These differences, I argue, can be explained by identifying the interactional goals that participants are trying to achieve, and by examining the research interview as an instance of ‘institutional talk’ (Drew and Heritage, 1992; Thornborrow, 2002). Specifically, how interactional asymmetries in processes such as turn-taking and questioning rights impact on the authority of the warrants used in the interaction. The core aim of the article is to examine what happens when individuals make ‘reality claims’ that may be contestable either in and of themselves, or because of how particular versions of reality achieve authority as a consequence of the power relationship that characterizes the interaction.
The issues discussed in this article are, I suggest, critical for extending and deepening our understanding of phenomena such as sexism. Actors responsible for developing, implementing and using work place policies to deal with sexism, such as, for example, discrimination and harassment policies, have to rely on both written and verbal accounts of behaviours and practices that are labelled as sexist (Townsley and Geist, 2000). As I will illustrate in my analysis, because sexism is such a controversial category of experience, neither establishing nor disputing its ‘reality’ is a straightforward endeavour, which is, in itself, extremely problematic in policy terms. In addition, and as I will argue in the last section of the article, using the term ‘sexism’ to represent and categorize certain activities and behaviours has major implications for how individuals relate to each other at work and make sense of their experiences. In short, the term sexism (may) work not only as a representational form of discourse, but also as a regulative force, acting upon identities and practices. Thus, this issue extends beyond problems with policy and into the heart of how individuals come to understand and enact their relationships at work.
The article is organized as follows. First, I elaborate on the central issue outlined above – how do we theoretically resolve the problem that different people interpret the same or similar social phenomena in very different ways, without privileging particular versions of reality? I explore this issue as it pertains to studies of sexism, illustrating that the problem persists even in those studies that have embraced the polysemic nature of interpretive processes, such as discursive approaches. I then outline the theoretical framework adopted for this study, followed by the empirical material. The discussion and conclusions develop and discuss the implications of my analysis for the study of sexism at work.
The reality of sexism?
The precise origins of the term ‘sexism’ are difficult to trace, but it almost certainly dates back to the 1960s and the political activities of feminists in various educational establishments (Code, 1991). While the term has evolved to encompass a variety of forms of sexism (e.g. subtle versus overt [Swim and Cohen, 1997]; or ambivalent versus benevolent [Glick and Fiske, 1996]), definitions have in common the idea that sexism is ‘a system of oppression based on gender differences that involves cultural and institutional policies and practices as well as the beliefs and actions of individuals’ (Shorter-Gooden, 2004: 407).
Researching sexism is a difficult task, largely because the term is so difficult to define (Denissen, 2010) and capture using standard methods such as surveys or interviews. Typically, approaches to researching sexism involve directly asking individuals whether they have had certain experiences, such as the receipt of sexual comments (e.g. Klonoff and Landrine, 1995; Somvadee and Morash, 2008), or by inferring sexism from accounts of, say, organizational culture in specific professions (e.g. Watts, 2007). A major problem with these studies is, however, that while individuals may hold certain beliefs or may have experienced certain situations, they may not share the researcher’s interpretation that these beliefs or experiences are illustrative of sexism (e.g. Dellinger and Williams, 2002; Gough, 1998). This poses a fundamental epistemological problem. If actors do not interpret their attitudes and experiences in the same way as the observer, whose perspective should be privileged? Since discursive studies attend closely to the complexity of the processes that lead to multiple and conflicting interpretations of the same or similar practices, they may provide insights into this epistemological problem. Following Alvesson and Karreman (2000), these approaches can be broadly categorized as being either ‘big D’ or ‘small d’, according to how discourse is theorized. It is to these studies that I now turn.
Sexism as discourse
‘Big D’ approaches
‘Big D’ approaches (Bingham, 1994; Clair, 1993, 1994, 1998; Dougherty, 2006; Wood, 1994) are those in which discourse is understood in a broadly Foucauldian sense, as culturally and historically specific systems of knowledge that offer multiple and often conflicting representations of the world (Alvesson and Karreman, 2000). Notwithstanding this multiplicity, some representations achieve dominance as the accepted or commonsense version of a particular state of affairs, producing and reproducing privilege and advantages for some people in the social order and relative disadvantages for others (Mumby, 1997). While, therefore, practices and activities that can be discursively represented as sexism are subject to multiple interpretations, within organizations there tend to be dominant discursive practices (ways of talking and writing). Such practices, it is argued, encourage individuals to make sense of their experiences in particular ways, often in those that act to reproduce existing patriarchal relations of power (Bingham, 1994; Clair, 1998; Dougherty, 2006; Wood, 1994).
For example, research has examined how workplace policies may construct sexual harassment in ways that ultimately absolve the perpetrators and their employing organizations of responsibility for this practice (Clair, 1993; Townsley and Geist, 2000). Policies, for instance, that encourage the victim to, in one way or another, confront the perpetrator, take for granted the idea that not only will assertiveness work to halt the harassment but also that harassment is somehow a consequence of not being assertive enough. Likewise, many organizational policies designed to address sexism fail to take into account the complexity and nuances that accompany an individual’s experiences of thinking about and deciding whether to report sexism, tending to assume ‘non-gendered, rational models of decision making’ (Townsley and Geist, 2000: 211).
The problem of interpretive privilege
The issue, however, of how to explain those accounts of events or experiences that participants implicitly or explicitly refuse to name or interpret as sexism is as relevant to discursive as it is to more mainstream studies. Clair (1993) addresses this issue by considering such accounts as illustrating the use of framing devices, such as trivialization (e.g. interpreting events as jokes). Such devices, she suggests, work by sequestering sexual harassment, acting to reproduce hegemonic patriarchal relations in organizations.
Useful though these approaches are, authors in the ‘Big D’ tradition do not seem to recognize that the category ‘sexism’ is as much constituted by their own discursive practices of researching, writing and theorizing as it is by the discursive practices of their research participants. As such, despite embracing the multiplicity of meaning that is invoked by using the notion of discourse to explore sexism, authors tend to become trapped by mundane reasoning, in which ‘the world is seen as a determinate and definite object’ (Pollner, 1987: 15). Thus, while acknowledging that sexism is a discursive achievement, sexism is nevertheless reified and assumed to be independent of the discursive practices through which it is constituted, as evidenced by the use of ideas such as the ‘sequestering’ or ‘pervasiveness of sexual harassment’ (Clair, 1993: 115). To put this another way, while these authors argue that sexism is the product of interpretive processes, they nevertheless assume that sexism is not an interpretation (see Hester, 1991 for a discussion of this issue as applied to ‘deviance’). Foss and Rogers (1994: 160) problematize this tendency as follows: The constitutive nature of all discourse − including academic texts like this one, produced by and/or for oppressed or marginalized groups – demands that we attend to the possibilities and consequences of our discourses. We believe that, as academics and feminists, we ought to be consistently self-reflexive and self-critical, recognizing the powers as well as the limits of the particular approach in which we are framing an issue such as sexual harassment.
Hence, it seems that in many studies, despite their political commitments, the researcher’s perspective is privileged. Scarduzio and Geist-Martin (2008), however, offer one explanation for why the interpretations of the researcher and participant may differ. In a study of male professors’ experiences of sexual harassment in academe, they note that their reluctance to name certain experiences as sexual harassment may stem from the subject position ‘victim’ in sexual harassment discourse. This subject position, they argue, is not easily taken up by men, especially high status men, who may experience high levels of internal conflict and ambiguity in using this term as they attempt to define and make sense of their experiences. Their study draws our attention to the fact that accounts of sexism have to be produced in specific interactional contexts that have distinct effects on what participants say. ‘Small d’ approaches, which I discuss next, treat the interactional context as a central analytic focus.
‘Small d’ approaches
The core concern of ‘small d’ approaches to discourse, such as discursive psychology, is an understanding of how experiences and events are constructed in everyday interactions, such that individuals are able to work up an account of, say, ‘sexism’ (Antaki, 1994; Antaki and Widdicombe, 1998; Potter and Wetherell, 1987; Wetherell and Potter, 1992). Discursive psychologists do not understand the vocabulary, labels and categories that individuals use to construct these accounts as reflecting the reality of either external events, or internal cognitive processes (Edwards, 1991). Instead, these linguistic elements are understood as discursive resources that can be used by individuals in creative and flexible ways, as they navigate the various interactions in which they are involved (Speer and Potter, 2000).
A difficulty with researching sexism from a ‘small d’ perspective, is how this is to be achieved methodologically. If, for example, researchers approach participants with questions such as ‘have you ever encountered sexual harassment?’ (Clair, 1993: 136), there is a danger that any account of ‘sexual harassment’ generated, reflects the researcher’s agenda and concerns. As Potter and Hepburn (2005: 293) argue: ‘At its most basic these issues present us with the possibility that a piece of interview research is chasing its own tail, offering up its own agendas and categories and getting those same agendas and categories back in a refined or filtered or inverted form.’
One way that discursive psychology has avoided this problem is to focus on categories, of which sexism could be one example, as participants’ not analysts’ categories. Taking this approach means that ‘sexism’ would be understood not as some form of external reality but as endogenously produced by individuals within specific interactional contexts (Schegloff, 1997; Speer, 2005). A potential problem with this approach, however, is that if we focus only on those categories that participants orient to as relevant, we may lose sight of how the macro social context constrains what is said and influences how participants account for their experiences (Wetherell, 1998). Clair (1993), for example, argues that women may account for their experiences without labelling them or orienting to them as instances of sexual harassment, due to the constraining influence of patriarchy. The danger then, in focusing on sexism only as a participant’s category, is that we may lose the opportunity to say something that is politically effective (Speer, 2005), a critical issue for studies of phenomena such as sexism.
In the attempt to respond to this problem without making discourse an ‘active agent’, critical discursive psychologists, such as Wetherell (1998), have drawn attention to some of the institutionalized and hegemonic aspects of the discursive resources utilized by individuals during specific interactions. The notion of ‘interpretive repertoire’ (Gilbert and Mulkay, 1982; Potter and Wetherell, 1987; Wetherell and Potter, 1992) has been employed to this end, which is defined as a ‘culturally familiar and habitual line of argument, comprised of recognizable themes, commonplaces and tropes’ (Wetherell, 1998: 400). Using this approach to research sexism, a consistent finding is that a discourse of ‘equality’ is frequently mobilized by participants to argue against the idea that men and women should be treated differently or afforded differential access to career or other life opportunities (Gough, 1998; Riley, 2002; Wetherell et al., 1987). At the same time, however, participants will frequently question the possibility of equality by drawing on discourses of, for example, natural differences, to construct men and women as essentially different (Gough, 1998). As Dixon and Wetherell (2004) point out, such processes illustrate the essentially dialogic nature of discourses such as equality, revealing how individuals’ understandings are driven not necessarily by underlying psychological processes, such as attitudes, but by the specific context in which such ideas are mobilized and how this may trigger attempts to defend or contest particular versions of reality.
The problem of interpretive privilege persists
‘Small d’ approaches are extremely useful for illustrating that discourse, in its cultural sense, cannot be simply read off individuals’ accounts of their experiences, as tends to be assumed in the more macro or ‘big D’ approaches to discourse reviewed above (Alvesson and Karreman, 2000; Fairhurst, 2007). Rather, they show that individuals are pushed and pulled by the discursive dilemmas that they encounter as they attempt to use these discourses to construct coherent and credible accounts of their beliefs and experiences (Wetherell, 1998).
Nonetheless, despite offering a finer grained analysis of discourse ‘in situ’, some ‘small d’ studies share a problem in common with ‘big D’ studies of sexism, which is the tendency to privilege the researcher’s interpretation of the participants’ experiences/accounts. Dixon and Wetherell (2004), for instance, exploring how the gendered division of household labour can be understood from a discursive psychology perspective, argue that when couples assess their contributions to household labour using ideas of ‘give and take’, they are ‘employing a shared discourse of equality, a discourse that [operates] to conceal gender inequalities and to constrain possibilities for their resistance’ (p. 176). Despite, therefore, expressing a commitment to provide explanations for how women account for the division of household labour without evoking concepts such as ‘false consciousness’ or gender differentiated values, this division is interpreted ipso facto as an unequal division, which slides into the intellectual hegemony the study is seeking to subvert.
Speer (2005) has argued that it is possible to both avoid intellectual hegemony and maintain political sensitivity by more carefully abiding by Schegloff’s (1997) insistence that social scientific categories and concepts should be treated as participants’ and not analysts’ categories. Critiquing the idea of hegemonic masculinity as a macro discourse with constraining effects on the identities of men, Speer (2005) draws on a critical insight endemic to the discursive psychology tradition: discursive categories, such as masculinity, do not have a universal meaning. Rather, their meaning is both context free and indexical (context dependent). Speer uses this insight to argue that, in any given interaction, participants may invoke discursive categories such as ‘hegemonic masculinity’. This, however, does not mean that participants are being regulated by or resisting this discourse, but rather that invoking this discourse enables them to manage dilemmas that are being generated by the ebb and flow of the interaction.
To illustrate this point, we can apply Speer’s (2005) approach to Dixon and Wetherell’s (2004) account of the ‘equality discourse’ above. Rather than understanding the mobilization of this discourse as an attempt to disguise inequalities in the division of household labour, it could be argued, instead, that this discourse (its context-free meaning) was mobilized by couples as a means to defend their relationship from the implication that it contained inequalities (indexical meaning), an implication generated by the research interview itself. From this perspective, therefore, inequalities ought not to be understood as being the definitive ‘reality’ of how domestic labour is divided, but rather as a potential (and normative) interpretation of how these couples divide up their household tasks; an interpretation that they both orient to and seek to subvert. As Speer (2005: 150) observes, such an approach does not exclude politics, rather it reveals that politics inhere not in grand, macro-level processes or structures but in the micro, highly local level of everyday interaction (Thornborrow, 2002).
In this study, therefore, I utilize Speer’s (2005) approach applied to the study of sexism. I do not want to reject the idea that sexism exists as a discursive category with distinct meanings; but, to avoid investing the category with an ‘autonomy of its own’ (Al-Amoudi and Willmott, 2011), I want to focus instead on examining how participants themselves orient to, foreground and actively produce this category within an interaction. I want to look carefully at how my position in the research conversations reported here and detailed below, influenced the accounts of sexism that were constructed by research participants. In addition to understanding the research interview as an essentially dialogic process (Alvesson, 2003; Alvesson and Skoldberg, 2000; Tanggaard, 2009), in which the interests of both the researcher and the participant influence what gets said and why (Alvesson, 2003; Kauffman, 1992; Potter and Hepburn, 2005), I also view the research interview as a specific instance of institutional talk (Drew and Heritage, 1992; Thornborrow, 2002); that is, as an interaction in which there are asymmetries in processes such as turn-taking and questioning rights that fundamentally influence the construction of the account. I am particularly interested in understanding the role such asymmetries play in enabling certain representations of reality to achieve authority (Thornborrow, 2002) and with what consequences.
Method
The data for this article were drawn from a larger project conducted more than 10 years ago that was concerned with an examination of police culture and gendered processes. While these data are ‘old’ and while some of the meanings articulated in the data may well have changed, the processes that I want to explore are as relevant today as they were at the time the data were collected. This project, which was based on a social constructionist epistemology, used ‘conversations’ as one of the primary methods of data collection. Underpinning this method was a rejection of the idea that the interviewer is an objective expert, and an overt attempt to equalize the relationship of power existing between the individual and the interviewer (Hollway, 1989; Mama, 1995). Conversations were conducted with 16 individual police officers and two groups (comprising 10 and two individuals respectively) of police officers at three separate locations (hereafter referred to as locations 1, 2 and 3) in a large police force situated in the North East of England. Individuals were recruited via a process of snowballing. For the purposes of this article, I utilize material from only two of the interviews, for reasons that will be discussed below.
The conversations
Informed by the idea that the research interview is a specific instance of knowledge production within a situated relationship of power (Henriques et al., 1984), I attempted to equalize this relationship by:
providing a full explanation of the research project on meeting the participant;
explicitly positioning myself as a participant, albeit one with a specific agenda that I wished to pursue in conversation;
avoiding the term ‘interview’, so that the participant understood that a set format would not be followed.
However, following completion of the research, I came to understand that there were a number of problems with my approach. First, equalizing a power relationship is rendered especially problematic in research where, generally, participants tend to read the researcher as an expert. Thus, while I embraced the notion that researchers should, in the name of emancipation, attempt to engage in equal relationships with research participants, I came to believe that the extent to which this is possible is probably limited due to the cultural ideas that exist about academics; namely, that they are knowledgeable and middle class.
Second, as I embarked on the data analysis, I realized that I could not bracket myself off from this process. While I had intellectually engaged with the notion of the research interview as a specific instance of knowledge production, what I had not fully grasped, up until I came to analyse the data, was how the relationship subsisting between myself and the participant, and our mutual reading of each other, was critical in understanding what got said and how it was said. This came to be central to my analysis, as I outline below.
Nevertheless, in conducting the conversations, I followed the three steps outlined above. I introduced myself and explained fully the aims of the research and gave some background to it. This varied for each participant because I was purposefully avoiding following a set script, in order to subvert the idea that standardization will result in the production of ‘truth’. I asked the participants if they had any questions before moving on to explain the method. I told participants that I did not intend to ask any set questions, but that I hoped we could have a full and frank discussion around the questions the research was aimed at addressing. Typically, therefore, all conversations began with a discussion of the police organization and what it meant to them as an individual, what they understood by the term ‘organizational culture’ and what their views were about gender in the context of policing. I shared frank views and observations with the participants, disclosed lots of information about myself and shared confusions or gaps in my knowledge about the research topics.
I asked permission from each participant and each group to tape record the conversation, and explained that, in any written materials generated, their individual identity would be protected. The participants discussed below are referred to with pseudonyms. I provided additional assurances regarding confidentiality and anonymity by explaining that the material would be accessed only by me, would be securely stored at a location some 40 miles distant from the force and that the tapes would be destroyed following transcription. No participants objected to this approach. All conversations took place on police premises, usually when the officer was either at lunch, or at the end of a particular ‘shift’. Conversations lasted for a minimum of 90 minutes, with three conversations lasting for more than three hours. All conversations were fully transcribed by me, the author.
The material from these conversations amounted to more than 500 pages of typed transcription covering a wide range of issues. Sexism was often referred to explicitly or implicitly by either myself or the research participant. Much of the time this amounted to only one or two comments. Following Speer (2005), I chose the extracts below for analysis because in each of these stretches of talk, sexism was foregrounded either by the participant or by me, the researcher, within the conversation. Thus, sexism as an interpretive category is not imposed on this data, it is oriented to and worked up by participants themselves. Choosing extracts from research interviews on this basis is typical of discursive psychology approaches (see e.g. Speer, 2005; Speer and Potter, 2000).
A second reason for choosing these particular passages was that they construct contrasting versions of sexism within the police service. Account 1, produced between myself and Sophie (see below for biographical details), constructs sexism as a reality. Account 2, produced between myself and Sally (see below for biographical details), sees Sally attempting to defend the police service from my (albeit ‘hedged’) accusations of sexism, and thus in this account, sexism is interpreted as being in the ‘eye of the beholder’. It was the contrast between these accounts that led to the first research question addressed by this study: ‘how do individuals construct accounts of sexism and what tactics do they use to convince their audiences that sexism does exist or, conversely, that it is “in the eye of the beholder”?’
Third, these extracts illustrate distinct interactional differences. For example, my interaction with Sophie, the participant who constructed an account of police sexism, was more typical of an informal conversation than a formal research interview, a fact illustrated by the level of personal disclosure, and the extent to which Sophie initiated and terminated topics. The dialogue also showed many overlaps, short sequences of turn-taking and much interjection mid-turn. Sophie, in short, orients to the interaction as a conversation with an equal, in contrast to Sally who, despite my attempts to equalize the power relationship, clearly saw me as an expert. I inferred this from the fact that she oriented to the conversation as an interviewee, generally waiting for me to ask questions rather than initiating topics. Additionally, as I will show, while Sally also defended the police from my implied accusations of sexism, she also showed a great deal of deference to my opinions. I also display far more confrontational tendencies in this dialogue than I did in those with other participants who did not defer to me in this way.
It was these differences that led to my second research question, ‘in what ways does the research interview influence the production of [such] accounts?’ This responds to the argument made by Potter and Hepburn (2005: 300) that ‘much more research is needed into the social and interactional nature of the research interview itself. Despite its ubiquity it has remained surprisingly under studied’.
To analyse these accounts, I draw on a discursive psychology approach in which I pay careful attention to the discursive devices (Antaki, 1994; Potter, 1996; Whittle and Mueller, 2011) used to construct the account, looking particularly at warrant (Gergen, 1990) – how does each party in the interaction attempt to persuade the other that their version of events or reality is the authoritative version? Here, I am utilizing the idea that accounts are often constructed with rhetorical intent, designed to anticipate and contest potential counter-arguments or positions, even when these are not expressed (Billig, 1989). In addition, I also understand the research interview as a ‘productive network within which people are both the agents of power and affected by it’ (Thornborrow, 2002: 137). Thus, I examine how the discursive resources used to construct these accounts are influenced by the interactional context, and by the relationship subsisting between myself and the participants. The accounts are neither coded, nor their content categorized. Rather, the analysis is oriented to understanding the fine grain of the interaction as it evolves between myself and the research participants. Again, this is typical of discursive psychology approaches, where the aim is not to identify categories, make inferences about the role of internal processes (e.g. cognition) or external processes (e.g. organizational culture), or to look for generalizable features of the content of the interaction. Instead, the aim is to understand how the demands of the interaction influence how each party responds to what the other is saying (see Potter, 1996).
Analysis
Account 1 – Sophie
Sophie was a full-time police sergeant at location 2, who had been in the service for five years. She was British, white, aged 35 and was in a long-term heterosexual relationship at the time of the study. She was a graduate entrant to the police service and had been ‘fast tracked’ to Sergeant as a consequence. Sophie and I established an immediate and strong rapport on meeting, possibly due to similarities in age and educational background. We had much in common and disclosed lots of very personal information during our conversation. After explaining the research and its aims, I expressed some scepticism to Sophie about how police officers had portrayed themselves as typically ‘androgynous’ in an earlier questionnaire I had used. Sophie supported my scepticism and moved on to tell me that she had been bullied. Thus, our orientation to the police organization is jointly established as critical from the start, and we share an interest in producing accounts in which the police are constructed negatively. Nonetheless, despite the establishment of this mutual interest, as we will see in this first extract (where Sophie is describing her reaction to some postcards on the wall of a CID office, showing women in scanty bikinis), she has much work to do in producing an account that persuades the hearer (me) that sexism is a fact that is ‘independent of her wishes’ (Smith, 1978: 25):
S: But I remember thinking, ‘This is a professional office!
I’ve never seen naughty post cards in a professional office before’.
And I was foolish enough in retrospect to say, ‘My Goodness me!
What are those doing on the wall? That’s a bit off isn’t it?
Having pictures like that in a CID office?’
Cos y’know, you wouldn’t walk into a solicitors’
and see that, or most places. I mean
I equated that kind of calendar girl
stuff with garages!
Me: Absolutely! That’s where I would have expected to see them!
Because it is possible for people who interpret situations and events as instances of sexism to be constructed as overly sensitive (Clair, 1993; Hinze, 2004; Townsley and Geist, 2000), in Sophie’s first turn, in lines 1 to 2, she prefaces her disapproval of the postcards by suggesting that it is not the postcards per se that she found troubling; rather, that their contextual location was the source of her concern. In doing this, she establishes and anticipates the idea she expresses in Extract 2 below, that she is not a ‘prude’ and, hence, unlikely to be someone who is prone to make interpretations of sexism. Having established that the postcards troubled her because of the situation and not because of her own sensitivities, she goes on to describe her reaction (I was foolish enough, in retrospect to say ‘My Goodness me! What are those doing on the wall. That’s a bit off isn’t it?’, lines 3−4). In imputing naivity to her reaction (I was foolish enough, in retrospect, to say), she uses stake inoculation (Potter, 1996) to signal that this was not the consequence of being an overly sensitive person reacting to the postcards with ‘all guns blazing’, but the consequence of being an ordinary person (Edwards, 2007; Wooffitt, 1992), perplexed by the incongruity of the display of the postcards.
In lines 5−9, she continues to build the credibility of her account about why she found the postcards troubling by constructing the idea that the display of such material in a public office is a sign of that office being unprofessional (Cos y’know, you wouldn’t walk into a solicitors’ and see that, line 6). Further, by using an extreme case formulation (or most places, line 7), a device that acts to legitimize a claim (Pomerantz, 1986), she implies that the display of these postcards is highly unusual and is, therefore, noteworthy whether the ‘office’ concerned is professional or not. In lines 8−9, she continues to particularize the ‘problem’ of the location of the postcards by stating where she would expect to see such material (garages), and where she would not (a CID office). This serves to further justify her earlier claim that professional locations are not places that such material should be displayed. This justification works, in this instance, due to my tacit acceptance that sexism may be something that could, perhaps, be expected in manual rather than professional occupations, indicating a shared agreement about the role of education and possibly, class, in explaining sexism. Note my enthusiastic receipt token in line 10. Here, I not only provide an explicit expression of approval for Sophie’s account, but in endorsing the idea that the source of her trouble lies in an external not internal circumstance, I also warrant her version of reality.
In this next extract, we see more evidence of how Sophie establishes her identity as someone who is not overly sensitive or prudish, and how she continues to work up the facticity of sexism. Sophie is responding to some comments I have been making, in which I have expressed surprise that a female officer had told me she didn’t mind officers making comments about her body:
S: I think I can appreciate that. It didn’t used
to bother me and when I realized that it did
bother me and I was just pretending it didn’t . . .
and now it does. And I’m beginning to confront it.
Somebody sent me an obscene email
and it was somebody I get on with very well.
And I said to him . . . ‘Actually, you offended me’.
And you do get people making comments.
They make comments about your (pause) sexual activities . . .
and expect you not to mind . . . y’know . . . or other people’s sexual
activities . . . and it’s very difficult to confront
that sometimes without looking like you’re
a fucking prude . . . y’know?
And you can’t say anything without it being
misinterpreted . . . and I was joking with somebody
that I wanted my belly button pierced . . . and
immediately they started saying – ‘Oh well,
you could have your clitoris pierced’
. . . And I was thinking – ‘Oh, for God’s sake’ . . .
In lines 1−3, Sophie is doing two particular pieces of work. First, she is adding further credibility to her identity as an ‘ordinary’ person, by using a contrast device, ‘avowal of prior scepticism’ (Lamont, 2007; Wooffitt, 1992). First, she says that ‘it [comments about her body] didn’t used to bother me’ (lines 1−2), then she realized that it did bother her (lines 2−3) and that she ‘was just pretending that it didn’t’, and finally that it does now bother her (lines 3−4). By suggesting a gradual recognition of the ‘problem’, she is deflecting a potential charge of being someone who is predisposed, or overly quick, to make presumptions about events and situations. Such predispositions and evaluative processes can be seen as signs that peoples’ judgements are suspect (Wooffitt, 1992). Second, she is further establishing our rapport by reinforcing my position (that people should be offended by such comments) and repairing a potential disagreement with me (‘I think I can appreciate that’ line 1).
Nonetheless, despite the harmonious conditions that are a feature of our interaction, Sophie continues to work on the facticity of her account in a number of ways. In line 4, after establishing that she now finds sexual comments offensive, she says ‘and I’m beginning to confront it’, a device that works to persuade the hearer that she does genuinely believe such comments to be offensive, and is not simply making that claim for some other, more interested purpose such as, for instance, to agree with me. In lines 5−6, she uses another contrast device ‘Someone sent me an obscene email . . . and it was somebody I get on with very well’, in which, by establishing a friendship connection with the emailer, she deals with the potential accusation that her interpretation of the email as obscene is due to some form of bias, such as dislike (Smith, 1978). In line 7, by changing footing with the reported speech, ‘actually you offended me’ (line 7), which provides an illustration of how she is now confronting sexism, Sophie can be understood as adding to the facticity of her account by rendering her claim about finding the email offensive more robust or vivid (Wooffitt, 1992). Moreover, the reported tone of the confrontation is calm and rational, acting to provide further instructions for me to read Sophie as an ordinary and not an overly-sensitive and, perhaps, reactionary individual.
In lines 9−11, the pause between ‘your’ and ‘sexual activities’ indicates that Sophie has detected some troublesome aspect of her account, and we see the source of this trouble in the repair ‘other people’s’. Should Sophie stick with the claim that people make comments about ‘your’ (meaning ‘my’) sexual activities, there is a danger that her reading of the situation (as a comment about sexual activity) may be seen as subjective and, hence, unreliable. By indicating that this also happens to other people, she establishes these comments as facts that happen independently of her (Smith, 1978). Moreover, by acknowledging that confronting such comments can mean that a person is seen as a ‘fucking prude’ (line 13), she adds to the facticity of her account by using this stake inoculation (Potter, 1996) to imply that the cause of confronting people who make sexual comments does not proceed from her own sensitivities, but from the offensive nature of the comments themselves.
This facticity is further reinforced in the remainder of the extract. One common device used to naturalize comments that may be interpreted as ‘sexist’ is to suggest that these are actually jokes or ad hoc comments, that are being misinterpreted (Clair, 1993; Townsley and Geist, 2000). Sophie exploits this device by particularizing a comment she made, ‘wanting my belly button pierced’, as being a joke (line 15), which is then used to turn the interaction explicitly sexual by ‘somebody’ who suggests Sophie could also have her ‘clitoris pierced’ (line 18). By rendering the interaction casual and not intimate ‘I was just joking with somebody’ (i.e. nobody special or particular), the hearer (me) is helped to interpret Sophie’s comment as a joke, because the term ‘belly button’ would not generally be considered offensive in a casual interaction. In contrast, I am helped to read the reply as offensive because the term ‘clitoris’ is not one that would be routinely used in a casual interaction and Sophie’s claim that this was an ‘immediate’ response to her comment, implies that it was not made as a consequence of the conversation heading in a more intimate direction. Thus, Sophie is working to avoid the possible interpretation that the offensive comment was caused by the nature of her own behaviour.
Overall, therefore, although I provide an auspicious environment (Wooffitt, 1992) for this account, through the explicit aligning of my interests with Sophie’s and by providing hearable warrants for her version of events, in order to establish sexism as a fact, Sophie has to use a number of devices to produce the idea that sexism exists in an objective sense. She does this by using counter-dispositional devices (Edwards, 2007) to construct her identity as an ordinary and not an overly sensitive person; by making appeals to norms about professional conduct, which locate the cause of her offense in external and not internal (i.e. subjective) circumstances; and by establishing a lack of interest and stake in constructing the object ‘sexism’.
Account 2 – Sally
Sally was a full-time probationer constable. She was British, white, 36 years of age, in a heterosexual marriage with two children and had been in the police service for 18 months. Prior to joining the police service she had worked part time in Social Services. Our dialogue began by me asking Sally why she had wanted to become a police officer, and from there we moved on to talk about her husband and family. Sally used a variety of discursive resources to convey her satisfaction with the job, the organization and her colleagues, and spent a good deal of our dialogue trying to convince me that men and women are basically the same, largely as a consequence of my repeated insinuations, through the questions I asked, that women officers are treated as, or are seen to be different from, men. In the following extract, I make a direct assertion about gender differentiated experiences in the police by labelling them as sexism.
Me: One of the things that I often wonder about is,
as a woman, if I was considering a career, I wouldn’t
consider a career in the police because, among many things
I’m not suited to being a police officer but among other things
. . . one of the things that I think would put me off,
is what you hear about it being such a sexist
organization and yet you say you’ve never
had any experiences . . .
S: If you’re talking about sexual innuendoes and things like that . . .
Me: Does that happen?
S: It happens in every job you’re in
Me: Does it? (Laughs)
Every job you’re in? Did you get it in Social Services?
S: Yeah! It’s every job you’re in and I think
it’s how you deal with it. If something was said
that offended me I would say . . .
Me: What sorts of things do you mean? I actually
have never experienced that I have to say . . .
S: Right. Well . . . I can’t really explain it . . . I mean
there’s things like . . . the nick names you get . . . things like that.
D’you know what I mean?
Me: What sorts of nick names?
S: Well I mean − I get called all sorts − y’know . . .
I mean like . . . not nasty names but . . .
I mean my surname’s [gives surname and a nick name]
I mean, to me that’s not derogatory.
Now somebody might find that offensive,
but it’s how a person sees what is said to
them. And to me if it’s said in a . . . a way
that is not derogatory . . .
Me: What would you call derogatory
S: If anything was said to undermine or
dismiss me as a person or if something
was said about me as a person or any sexual innuendoes
or ’owt like that, I’d find that derogatory
and I’d say ‘Excuse me!’
Me: So they can’t make sexual innuendoes to you then?
S: No. No. Not at all. But the thing is as well,
I’m on a shift where I think the majority of us are married
Me: That makes a difference?
S: Well I think it does . . . I mean you hear about
people having affairs and people going off . . .
The hedging that characterizes my opening turn in lines 1−5 is a sign that I am building up to introducing a controversial topic. Hedging can be epistemic or interpersonal (Holmes, 1988). Interpersonal hedging is often used to facilitate interpersonal solidarity. In contrast, epistemic hedging can signify a speaker’s lack of certainty about the validity of the claims they are making and, therefore, also signals a relative lack of authority. The hedging I use can be understood as epistemic and, as we will see, the tear in the fabric of the power relationship illustrated in this interaction has some interesting effects on Sally’s subsequent turns.
I attempt to construct my position on sexism as neutral by acting as an animator (Goffman, 1981), a tactic in which I distance myself from the accusation that the police force is a sexist organization, and project it onto others: ‘what you hear . . .’ (line 6). This tactic can also be understood as an attempt to build credibility for my claim about sexism, as it suggests that there is a widespread belief that this is the case. I do not qualify what I mean by sexism, but I initiate Sally’s formulation, ‘If you’re talking about sexual innuendos and things like that’ (line 9), by ‘fishing’ for information, ‘and yet you say you haven’t had any experiences . . .’ (lines 7−8). By qualifying the meaning of sexism in this way, Sally is able to particularize my claim and thus concretize it. This makes the claim easier to (potentially) contest because it reduces the ambiguity of the term ‘sexism’. Having concretized my claim of sexism, Sally moves on to contest its credibility by using an extreme case formulation (Pomerantz, 1986), ‘it happens in every job you’re in’ (line 11). As already explained, such formulations act as legitimizing devices. Here, she is able to deflect the accusation that the police force is unique in being a particularly sexist organization, by constructing sexual innuendo as a normal, everyday occurrence in all organizations.
A formulation typically acts as the first part of an adjacency pair (Antaki, 1994) because it will usually give rise to a confirmation or disconfirmation of the meaning of the formulated utterance and acts as a mechanism through which interactants establish the gist of what is being said. However, in more institutional interactions, it is typical for the more powerful party to adopt the position of formulator and for the less powerful party to adopt the position of responder (Thornborrow, 2002). Thus, Sally’s formulation at line 9 disrupts our relative power positions. I reassert my authority in line 10 by neither confirming nor disconfirming Sally’s formulation but by evaluating it, ‘Does that happen?’, thereby re-establishing Sally’s role as responder. Her emphatic confirmation of my formulation, ‘It happens in every job you’re in’, results in a further evaluative formulation from me, which is inferentially elaborative: ‘Every job you’re in? Did you get it in Social Services?’ (line 13). Here, I am challenging Sally’s extreme case formulation (every job) by insinuating that the police force is a more likely site for sexual innuendo than Social Services, which is where Sally worked before joining the police service. In Sally’s subsequent turn, she continues with her claim that sexual innuendo is a normal part of working life, by not only refusing to acknowledge the intended irony in my challenge ‘Does it? Every job you’re in? . . .’ and my attempt to single out the police force as a site for sexism, ‘did it happen in Social Services?’, but also by both reasserting her claim, ‘Yeah! It’s every job’ (line 14) and by elaborating her meaning, ‘and I think it’s how you deal with it’ (line 15). Here she provides a warrant for the claim by disaggregating sexual innuendo into that which may be offensive and that which, by implication, may be less offensive or inoffensive. She also elaborates on her comment ‘it’s how you deal with it’ by stating that she would confront a speaker if anything offensive was said to her (line 16). Thus, in this turn, Sally implies that the offensive or inoffensive features of sexual innuendo reside more in the hearer than the speaker, an idea that is legitimized by her claim that sexual innuendo is a normal and ubiquitous aspect of working life.
In my next turn, starting on line 17, I ask for clarification from Sally regarding what would be considered an offensive versus an inoffensive comment, and also contest her claim for the ubiquity of sexual innuendo, by stating that this is something that I myself have not experienced. There is an interesting shift in footing illustrated in Sally’s response to my question starting at line 18. Sally’s turns in the interaction, from lines 1−17, can be seen not only to challenge my version of reality but, more broadly, to challenge my power as the speaker with authority (an opportunity provided by my epistemic hedging, and illustrated by her take up of the formulator position in line 9). However, in lines 18−21, Sally’s response contains hedges, which include the tag question: ‘Do you know what I mean?’ This appears to be interpersonal hedging, whereby Sally is trying to show interpersonal solidarity with me. This could be understood as an attempt by Sally to repair the challenge to my institutional position that has been brought about by her (from my perspective), dispreferred responses to my questions (Heath, 1992).
In line 22, I probe her use of the term ‘nick names’, to which she responds with more hedges, before moving on to justify why she does not find the nick name that she describes, offensive. Here we see that she runs into some trouble, because, having implied in previous turns that what is offensive is a subjective evaluation, she nevertheless implies that ‘nasty’ or ‘derogatory’ comments are generally found offensive, though maintains her claim regarding the subjectivity of such evaluations when she says ‘I mean, to me, that’s not derogatory’ (line 26) and ‘but it’s how a person sees what is said to them’ (lines 28−29). She then qualifies her position by claiming that, for her, an inoffensive nickname might be one that is said in a ‘way that is not derogatory’ (line 30). I continue to probe her response in line 31, which draws from her the more elaborate justification in lines 32−36. Here, she makes use of a three part-list (Jefferson, 1991) to emphasize what she would see as derogatory, which includes: being undermined or dismissed; something personal; or sexual innuendo. In completing the list with this last item, however, Sally has effectively contradicted her earlier position on sexual innuendo. Despite my (sceptical) remark in line 37, ‘So they can’t make sexual innuendoes to you then?’, Sally does not appear to orient to this contradiction. Instead, she again uses a three part structure (‘No. No. Not at all’ line 38), to emphasize that she would not let people make sexual innuendos to her, but she also implies that this does not happen on her ‘shift’ due to her belief that ‘the majority of us are married’ (line 39). Note my evaluative formulation of this presupposition, ‘That makes a difference?’ (line 40), and her response, which implies that sexual innuendo is most likely to happen between single people perhaps looking for relationships (lines 41−42).
Overall, therefore, compared to Sophie, Sally has a much more difficult time in warranting her version of reality due to the way I am positioned in the interaction. Even though, throughout this dialogue, I attempt to position myself as neutral with respect to police sexism – I never explicitly say that I think the police is a sexist organization − my questions and responses signal that I do not have a neutral position on this issue. This non-neutrality is expressed in my sceptical evaluations of Sally’s responses and in my attempts at fishing. Sally’s aim in this dialogue is to establish that sexism is a subjective experience, which she achieves by constructing sexual innuendo as a normal and ubiquitous aspect of life in all organizations and by using this ground to render sexism a necessarily perceptual figure. Nonetheless, because I am positioned in most of the interaction, as the evaluator of Sally’s responses, she is required to defend her claims. Even when this asymmetrical relation is disrupted, we work collaboratively to re-establish it.
Discussion and conclusions
Sexism is a difficult area to research, both because it is a controversial idea and because different people interpret similar events in different ways (Welsh, 1999). I have argued that one of the major problems with research in this area is the researcher’s tendency not to analyse the role that their own discursive practices play in the construction of the category ‘sexism’ (see also Kauffman, 1992; Potter and Hepburn, 2005). In this article, I have illustrated that producing an account in which sexism is constructed either as a reality or as a subjective individual perception, is a narrative achievement that can be made more or less difficult, depending upon whether the interactional context produces power asymmetries. I have argued that the research interview can produce power asymmetries, in as much as the interviewer may adopt an evaluatory position in the interaction, a position that is tacitly legitimized by the interviewee and which concedes interpretive privilege to the interviewer (Heath, 1992). When interpretive privilege is conceded in this way, the interviewee is placed in the position of needing to justify their particular version of reality. Nonetheless, interpretive privilege is not an automatic concession and, as the extract with Sally illustrates, there were subtle changes in footing that enabled Sally to assume the more powerful position in the interaction at different times. However, even where the power relationship is more symmetrical, as in my interaction with Sophie, the interviewee does not avoid the necessity for warranting the claims made in the account. Thus, despite the harmonious and cooperative nature of the dialogue in this instance, Sophie employed a number of devices to convince me of the ‘facticity’ of sexism.
It is because sexism is a contested social fact, that participants were faced with the various dilemmas illustrated in the analysis. Sexism can be understood, simultaneously, as an objective reality and as a subjective interpretation. Both interviewees were oriented to this duality in different ways. Sophie’s account was aimed at establishing the facticity (or objectivity) of sexism while, in contrast, Sally’s account was aimed at establishing it as a subjective experience. Both participants encountered dilemmas as they attempted to achieve these goals, as I illustrated in the analysis. The analysis, in particular, illustrates that this dilemmatic aspect of sexism may be foregrounded due to the nature of the interaction, and that one or other interpretation of sexism (objective/subjective) may be privileged. In both interactions presented here, the dilemmatic aspect of sexism was foregrounded because sexism was topicalized within these discussions by the questions I was asking, and by my responses to the answers and comments offered by the participants (Alvesson, 2003; Potter and Hepburn, 2005). That is, my interests as a researcher infused the interaction with a meaning that it may not have had (Kauffman, 1992), had the participants and I been discussing, say, the impact of policing on crime rates. Second, the interpretation that was privileged in both interactions was sexism as an objective fact. This interpretive privilege facilitated Sophie’s account while troubling Sally’s. Moreover, it appeared to work, at least for me, as a moral imperative within the interaction, in as much as I treated Sophie’s account with tacit and overt approval, while my disapproval of Sally’s account was illustrated in my sceptical evaluations of her claims, which she made efforts to justify and defend, as well as repair.
There are a number of implications of the analysis presented here. The first concerns the issue of neutrality. A lot of traditional social scientific research is based on the assumption that the researcher is a neutral observer who collects data in order to uncover or reveal some underlying truth. This research was explicitly designed to subvert that idea, but some might argue that the accounts presented here reflect the research design and that better results might have been obtained had I actually behaved in a more neutral manner, or even used surveys or questionnaires to avoid personal contact with the interviewees. I would suggest that it is not possible for researchers to adopt a neutral position in the social sciences (and some might say this extends to the natural sciences as well) (Penrose, 1989), whatever the subject of the research and however that subject is researched, though, admittedly, this problem is revealed most starkly when researching contested issues, such as sexism. The vast majority of management research, for instance, is underpinned by an implicit morality in which it is quite apparent which attitudes, behaviours and activities are understood to be appropriate and inappropriate, be that employee participation, employee misbehaviour, strategic management or instrumental workplace attachment. Whether researchers collect data via surveys, or via interviews in which they demonstrate no reaction to participant responses, the questions asked in and of themselves reveal this morality, and undoubtedly influence how participants respond.
The attempt to achieve neutrality, and the valorization of this position in the social sciences and management scholarship especially, reveals the competing institutional logics through which such scholarship is constituted (see Bergman, 1992 for similar arguments regarding psychiatry). On the one hand, neutrality is seen to be the signifier of a discipline that can be taken seriously, in the way that chemistry and physics are understood. On the other hand, management scholarship is inherently interest-laden – it seeks either to help organizations ‘do better’ in whatever way that is understood, or, for scholars like myself, to reveal how power and politics are inherent in management as both a practice and discipline. This contradictory structure is present no matter how research is designed. So, while it would undoubtedly have been the case that, had I performed this research using a standard sexism questionnaire, I would not have been able to reach the conclusions I am presenting here, I would also not have data that were any less ambiguous about the apparent status of sexism. I may, however, have been able to appease some readers by being able to state, for example, how many police women reported having been on the receiving end of sexual innuendo. Nonetheless, the problem would have remained – how do we account for the fact that some people see this as unacceptable while others see it as harmless, without either privileging the researcher’s construction of reality or by suggesting that sexism really is ‘in the eye of the beholder’?
A second implication derives from the central claim from the analysis, that constructing an account of sexism as either an objective fact or a subjective perception, is rendered dilemmatic because social facts, like sexism, are inherently dualistic – they can be understood simultaneously as both objective and subjective. It is thus difficult, in some contexts (e.g. western culture and research interviews with feminist researchers), to refute the idea of sexism as a broader reality, without encountering problems caused by the idea that certain practices (e.g. particular words or particular behaviours) are to be considered as universally offensive. Conversely, because in other contexts particular words and particular behaviours are understood as normal aspects of everyday experiences, individuals who do take offense may be seen as overly sensitive or as misinterpreting the actor’s intentions (Hinze, 2004). In many western organizations, this dilemmatic or dualistic aspect of sexism is often glossed by tactility privileging the subjective interpretation in workplace policy on sexism. This is illustrated in the fact that it is individual victims who are positioned as responsible for reporting and acting on behaviour that they perceive as offensive (Clair, 1993). Perhaps one of the reasons why many individuals shy away from making use of these policies inheres in the dualistic nature of sexism (or harassment) (i.e. fears about whether it is possible to prove one has been the victim of sexism and the danger that, in naming certain behaviours as sexist, individuals may run the risk of being seen as overly sensitive or improperly motivated) (Hinze, 2004). Furthermore, given that should individuals seek to formalize a complaint of sexism, they will be required to provide both a verbal and written account of their experiences, and given the claim that many victims of sexism, particularly sexual harassment, are younger females (Scarduzio and Geist-Martin, 2008), the difficulties in producing convincing accounts may, as illustrated in my analysis, be exacerbated by needing to do so in inauspicious environments (Townsley and Geist, 2000).
Nonetheless, it is the objectivist interpretation of sexism that also operates as a self-regulatory mechanism, encouraging people to edit their off-the-cuff remarks in lieu of the fact that these may be interpreted as offensive, and sometimes to reflect on the motivational basis of their evaluations of other women and men. Sexism seems then, in western culture, to be a ‘dispreferred’ category (Speer and Potter, 2000) – individuals generally prefer to avoid being labelled as sexist and may wish to desist from practices that may be interpreted as sexist (Gough, 1998; Tracy and Rivera, 2010). Thus, as I stated at the beginning of the article, the question of whether sexism is a reality is not simply an epistemological puzzle but is fundamentally implicated in how people make sense of themselves and of their experiences and relationships at work and in other social environments.
Finally, therefore, I want to conclude with some comments about what this analysis implies for research into sexism and whether, in problematizing ‘sexism’, this analysis may be undermining other analyses that have explicitly feminist and political aims (e.g. Clair, 1998). Given the points I have made about how difficult it is to research sexism using qualitative interviews, an important area for future research is to examine how claims (or refutations) of sexism are built within naturalistic settings, such as between managers and ‘victims’ or ‘perpetrators’, or in more formal, institutionalized settings, such as industrial tribunals. Given the extreme sensitivity of this topic, the former setting may be ‘off limits’ in that it would seem highly unlikely that parties in such settings would agree to have their conversation tape recorded and transcribed. The latter setting, where open to the public, is, therefore, more realistic for such work and may yield important insights into how more institutionalized interactions influence the building of facticity and the nature and efficacy of the warrants mobilized to legitimize competing reality claims. Moreover, we need to appreciate that the social category of gender is related in complex ways to other social categories such as race and heterosexuality. Future research could, therefore, usefully explore how participants orient to this complexity or intersectionality when constructing accounts in which they are making claims about differential and unfair treatment in organizations, be that on the basis of sex, race or sexual orientation.
These points notwithstanding, it is critically important for us to understand more about how categories, like sexism, may be oriented to and constructed differently in different interactional contexts. The points made about qualitative interviews by Potter and Hepburn (2005) are important, but they should not prevent us from pursuing our research with this particular method. We need not only to build an awareness of our own role in the production of knowledge about categories like sexism into our analyses (Alvesson, 2003; Potter and Hepburn, 2005), but to also think more carefully about what the various sensitivities (or lack of them) that our participants demonstrate when constructing these categories may be telling us in political terms (Speer, 2005).
Where participants are able to relatively easily construct the category ‘sexism’ as either a definite fact or a definite subjective experience, this may tell us something about the nature of the power relationships within which such accounts are constructed. When there are shared normative associations created − perhaps, by similarities in institutionally ascribed status or roles − it is generally easier for participants to construct certain categories such that they possess a ‘taken-for-granted’ quality (Hester, 1991). Such taken-for-grantedness can be, and is, however, sometimes disrupted. An important manifestation of power relations in an interaction, like the research interview, as illustrated in my analysis, can be gauged by examining whose representations of events and phenomena ‘get to be the dominant categories’ (Thornborrow, 2002: 57), and how interactants orient to this. When women refuse to label certain discursive and material practices as sexist, this may be less because they are denying some underlying reality or being duped by dominant discursive practices, but because they are defending their own identities from the implication generated by the research questions, that they are somehow unwitting victims (Scarduzio and Geist-Martin, 2008). This in itself tells us something very important – to be an unwitting victim of sexism may, for many women (and men), be a dispreferred idea, especially in interactions with academic researchers. In refusing to name or label their experiences as such they are actually demonstrating resistance to the disempowering idea that they cannot or do not know how to interpret appropriately and deal with their own experiences at work.
In short, the regulatory or power effects of discourse are manifest only in specific instances of interaction, demonstrating Foucault’s (1980) argument about how power is located not in any grand or over-arching structures, but in the minutiae of local practices, such as talk (Thornborrow, 2002). It is at this level that we see power and resistance as individuals struggle to privilege their versions of ‘reality’ in circumstances that may be more or less conducive. The interpretive status of social facts like sexism is, therefore, never guaranteed nor settled, no matter how auspicious the interaction for any given, preferred interpretation.
Thus, what I hope I have illustrated is that while a particular power relationship can explain why certain warrants for a given claim may be more or less successful, at the same time, power relationships, including those that are argued to oppress and subordinate women, are neither fixed nor immutable. Such relationships are always in negotiation within specific situated contexts (see Foss and Rogers, 1994, for a discussion of this issue as it pertains to sexual harassment). It is these very dynamics that give rise to hegemonic struggle and, indeed, are those that give rise to categories like sexism in the first instance (Wood, 1994). From this perspective, as social science researchers, we need to recognize that not only do the discursive practices of organizations privilege certain understandings, such that employees may find it difficult to interpret certain actions/behaviours or words as sexist, but also that our own discursive practices privilege other understandings, such that these same actions/behaviours or words are seen to have a universal meaning. As Melvin Pollner (1987: 70) argues: Resolution [of different reality claims] often involves a ‘politics of experience’ in which a group’s or individual’s experiences (or claims) about reality are dismissed or discounted in favour of what will be regarded as the official or accredited version of reality . . . [M]any social scientific concepts are the product of the politics of experience: products ranging from false consciousness, to repression to interpretation are constructed through the treatment of the social scientific version as definitive of reality by reference to which the lay member is found to be a deficient or subjective observer.
We must remember, therefore, that politics and acting in politically effective ways are not the sole provinces of the researcher. Indeed, there is a danger that if we, as researchers decide what is and is not to be considered politically effective, we undermine, discount and underestimate the power of those everyday, subversive actions that emerge when women and men engage in mundane struggles to warrant, assert or defend their own versions of organizational reality, or to contest and resist those of others. One way that we can demonstrate such sensitivities is to pay closer attention to how the categories we deploy to explain certain phenomena are actually produced as facts by both ourselves as researchers and by our participants. In doing so, we may gain insights into how even highly institutionalized ideas come to be contested and even replaced through the politics of experience.
Footnotes
Acknowledgements
The author would like to thank Associate Editor Professor Karen Ashcraft and the three anonymous reviewers for their insightful comments on earlier drafts of this article.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Penny Dick is Professor of Organizational Psychology in the Institute of Work Psychology at Sheffield University Management School, UK. Her research interests include identity, resistance and power; the impact of family-friendly policies on organizations and individuals, and the role of institutional work in understanding processes of organizational stability and change. She has published in journals such as Journal of Management Studies, Work, Employment & Society, Journal of Organizational Behavior, Journal of Organizational and Occupational Psychology and British Journal of Management. [Email: