Feminist dilemmas and the agency of veiled Muslim women: Analysing identities and social representations

Introduction

Two readings of the Muslim practice of veiling predominate in feminist scholarship: the veil as a symbol of subordination and the veil as a symbol of resistance (Bilge, 2010; Mahmood, 2005; Phipps, 2014). At the heart of these readings is the issue of agency and how it is understood and ascribed (Bilge, 2010; Mahmood, 2005). Within the tradition of liberal feminism, the equation of Muslim veiling with submission to patriarchy bluntly denies agency and ‘invalidate[s]’ the possibility of consent, ‘coupling the voluntary submission argument with a false consciousness’ (Bilge, 2010: 15). Opposing this reading and its construction of Muslim women as ciphers of culture are postcolonial feminist accounts of veiling as resistance – to Western hegemony, the sexualized culture of the West, and hostility towards and stereotyping of Islam (Bilge, 2010; Phipps, 2014). While an important corrective to the subordination thesis, such accounts arguably often ignore issues of power (Bilge, 2010; Phipps, 2014) and delimit agency by conflating it with resistance and neglecting frequently given reasons for veiling, such as those of piety and submission to religious duty (Bilge, 2010; Mahmood, 2005).

Researchers seeking to both respect different modes of reasoning in relation to veiling and critically explore gender-related dynamics face dilemmas about how to engage theory and methodology with complex issues of culture, gender and identity. In this article, I reflect on my own research within the field of social psychology to approach these issues from a new angle. Engaging first with Saba Mahmood’s account of gendered religious subjectivities in Politics of Piety (2005), I show how her work reinvigorates feminist theorization of agency through a Foucauldian approach to practices of self-formation but leaves open for further exploration important questions about how this approach might be folded into scholarship interested in critically examining socio-cultural imperatives and processes of transforming social relations.

Drawing on my research about veiling and identity among Muslim women in the UK and Denmark, I argue that analysis of the role of competing representations in the relational negotiation of identity has the potential to shed new light on matters of agency, engaging as it does a dialectical approach to how identities are challenged and contested in the meeting (and clash) of perspectives and systems of ideas, beliefs and values. Such an account requires attention to the social psychological effects, both threatening and transformative, of ‘knowledge encounters’ (Jovchelovitch, 2007). I show how a social conception of agency might be approached through examining the role in identity construction of different systems of social knowledge attached to social groups and milieus – what social psychologists term ‘social representations’ (Jovchelovitch, 2007; Marková, 2003; Moscovici, 2008).

I use the term identity to denote the social psychological concept of social identity as ‘that part of an individual’s self-concept which derives from his[/her] knowledge of his[/her] membership in a social group (or groups) together with the value and emotional significance attached to that membership’ (Tajfel, 1978: 63). While not excluding individual aspects of identity formation, research on social identity and representation highlights the importance for self-concept of social processes of group identification (Duveen, 2001; Howarth, 2002a). It guides my interest here in the social dynamics of gender identity among Muslim women living in northern European contexts and the implications for self-concept of negotiating competing representations of identity relating to the practice of veiling. My analysis of these dynamics shows that a social conception of agency must be distinguished and detached from the norms of agency that are central to particular representations of gender and be rooted instead in theory and research of the dialectical nature of identity as at once ‘contextual, imposed, inter-subjective and collaborative’ (Howarth, 2010: 8.2).

Dilemmas of agency for feminist research

Emphasis within feminism on women’s autonomy and choice arguably remains an important act of political recognition and a way of emphasizing the possibilities of challenging gender relations (Madhok et al., 2013). This is more clearly the case with respect to enduring constructions of the ‘Third World Woman’ (Mohanty, 1988) and specifically of veiled Muslim women as passive victims rather than competent agents (Duits and van Zoonen, 2006). At the same time, the turn to agency creates dilemmas for feminism. Critiques highlight the shift of attention from the psychological and material conditions of inequality (Hemmings and Treacher-Kabesh, 2013) and the evacuation of ideas of cultural influence in favour of ‘the discourses of individualized, postfeminist neoliberalism’ (Gill, 2007: 79).

The terms of the debate itself highlight the difficulty of disentangling theory and analysis of agency from feminism as a liberal political project. They tether agency to a ‘dichotomous frame’ of resistance versus submission to cultural forces of oppression (Bilge, 2010) and reveal an underlying dualism of individual versus society grounded in humanist ideas of pre-social autonomy and the free-willed individual (Davies, 2000).

A powerful challenge to these norms of agency exists in Mahmood’s Politics of Piety (2005). In her ethnographic account of women’s lives and experiences as members of the mosque movement in Egypt, Mahmood gives witness to the practices and sensibilities of a devout way of life and in so doing elucidates a form of religious agency that is oriented to piety and disinterested in subverting gendered norms and values. Through elaboration of her subjects’ practices of piety, Mahmood expounds a culturally situated moral agency that draws on Butler’s understanding of ‘becoming’ as ‘always under a cultural compulsion’ (Butler, 1990: 12) and is rooted in a Foucauldian model of self-formation that requires an account of ‘the scaffolding of practices’ that provide ‘the necessary conditions for both their subordination and their agency’ (Mahmood, 2005: 154).

An example is in her account of modesty as a virtue and quality of character that is achieved through the practice of veiling. Mahmood accepts that modesty is a heavily gendered virtue, like many Islamic virtues that in ‘their measure and standard vary when applied to men versus women’ (2005: 156), but argues that it should not be equated with submission and passivity. On the contrary, she concludes from close observation of women’s lives that Islamic virtues are ‘site[s] of considerable investment, struggle, and achievement’ and integral to a ‘constructive project’ of personal transformation in which veiling functions not as a sign of a virtuous self but as a bodily means of achieving it (2005: 174).

In this way, Mahmood decouples agency from resistance and elucidates it through the work that situated practices perform in the making of subjects. She thereby exposes the limitations of liberal feminist thought and the humanist ideas it is rooted in, including the belief that there is an innate drive to assert autonomy and individuality and that agency only resides in acts that defy socio-cultural norms. She challenges feminist scholarship to recognize ‘other modalities of agency whose meaning and effect are not captured within the logic of subversion and resignification of hegemonic terms of discourse’ (2005: 153).

In doing this, however, Mahmood leaves open questions about feminist analytical engagement with relations of power and dialogic processes of resignification and change (Bilge, 2010; Ismail, 2006). Mahmood asserts that she is not ‘indifferent to feminism’s interest in theorizing about the possibility of transforming relations of gender subordination’ (2005: 154) but her starting point is ‘that any discussion of the issue of transformation must begin with an analysis of the specific practices of subjectivation that make the subjects of a particular social imaginary possible’ (2005: 154). Politics of Piety illuminates the particularities of these ‘specific practices’ but not, as Mahmood acknowledges, the further discussion of ‘the issue of transformation’. For this, as scholars have reflected in response to Politics of Piety, attention to discursive processes and dialogic dynamics of resignification is required (Bilge, 2010; Ismail, 2006). Studies that expand on Mahmood’s work in this way hold the promise of shedding light on historico-cultural forces shaping modes of agency and intersubjective processes of contestation and change, particularly if combined with an understanding of how contemporary identities must negotiate ‘the flux of unstable representations around us’ (Howarth, 2002a: 145).

My research about veiling and identity among Muslim women in the UK and Denmark is a case in point. Analysis of discursive dynamics shows that, while reasons of religious obligation and modesty are significant across my respondents’ accounts of veiling, religiosity does not stand alone as a certain or uncontested grounding of reasoning and identity. In line with other studies of identity and belief among Muslim women in northern European contexts (for example Dwyer, 2000; Jacobsen, 2011), most of my respondents discuss the veil in ways that exhibit diverse and often apparently contradictory modes of reasoning, reflecting negotiations of intersecting socio-cultural contexts and competing representations relating to gender, Islam and the veil.

Drawing upon Bilge (2010), I argue in this article for an approach that brings together poststructuralist insights on subjectivity as both ‘constituted and constituting’ with an intersectional analysis – that is, an analysis that anchors ‘the formation of subjectivities and agency within a nexus of social relations and structures’ such as race and gender (2010: 23). I contend further that a feminist account of agency demands theorization of identity that does justice to its dynamics in late modern societies characterized by a plurality of forms of social knowledge and the challenges this presents to the stability of social identities (Giddens, 1991; Moscovici, 2000). That is, I propose a multi-level analysis that addresses the ways in which identities are negotiated in relation to competing systems of values, ideas and practices – or social representations. This requires an account of both the content and force of representational systems and the ways in which these are challenged and re-evaluated in dialogue and social practice (Howarth, 2010). While identities are always being challenged and negotiated in the encounter with others, these dynamics are particularly clear among those with minority identities threatened by collective stigmatization (Hall, 1997; Howarth, 2002a). Research among those with threatened identities can therefore help illuminate the dialectical nature of identity and agency in context, as I discuss in the remainder of this article.

A social psychology of knowing

Central to the theory of social representations is its concern with the increasing diversification of systems of knowledge in late modern societies (Farr, 1996). Moscovici (2008) introduced the term cognitive polyphasia to describe this diversification of knowledge at the individual level and how ‘a plurality of modes of thought can often co-exist within the same individual’ (2008: 185). Further empirical accounts of cognitive polyphasia have emphasized the function of different modes of knowledge – such as forms of scientific knowledge and religious knowledge – in maintaining social identity and making sense of complex social issues (De-graft Aikins, 2004; Jovchelovitch and Gervais, 1999; Provencher, 2008).

Jovchelovitch (2007) writes that ‘the key theoretical proposition underlying the concept of cognitive polyphasia is to ground the social psychology of knowing in the dynamics of social interactions and cultural contexts’ (2007: 69). ‘[T]he social world has a fundamental impact on not only what we think but, crucially, how we think’, observe Voelklein and Howarth (2005: 437). The understanding of psychologies as characterized by the coexistence and interaction of different modes of thought complements the discursive turn in social psychology and its account of the dialogic nature of thinking (Billig, 1996). Grounded in this understanding, social representations work on identity seeks to address the ways in which identities are continually challenged and constructed through dialogical encounter (Duveen, 2001; Howarth, 2002a).

At the same time, social representations theory is concerned with the regulatory function of representations as ‘systems of values, ideas and practices’ that forge social identities and constrain other ways of being and thinking (Moscovici, 1973: xiii). Duveen (2001) draws on his research on the development of gender identities among children starting school to make the point that individuals come to have a sense of identity only through ‘a recognition of their situation within the symbolic space of their culture’ (2001: 258). As he puts it, ‘representations precede identities’ (2001: 258).

Lloyd and Duveen’s (1991, 1992) mixed-method studies of representations of gender in the classroom elucidate these dynamics by drawing attention to both the representational structuring of gender in discourse and practice and the dialectical processes of development through which children become social actors. They show how the development of gender identity is a function of the accumulative integration and reproduction of dominant representations of gender, but at the same time involves processes of resistance and contestation as individuals do gender and express a position in relation to a particular set of representations. Their research demonstrates in this way how systems of representation relating to identity are acquired as the condition for social action but are also continually negotiated and altered through social interaction and the encounter with difference. Rather than separating social and cultural forces from psychological processes, their research provides an account of the development of representation (of world and self) as a social and symbolic as well as epistemic activity (Castorina, 2010; Howarth, 2010).

These insights guide my approach to researching identity in a way that attends to both the core complexes of meaning that forge social identities and the ways in which these are variously affirmed, resisted and contested or ‘re-presented’ in the social encounter (Howarth, 2010). Further, they inform my development of a conceptual place for agency in the dynamics of representation that exist ‘in the in-between space we create in dialogue and negotiation with others’ (Howarth, 2006: 68). While individual and personal factors are involved in this social representational space, I contend that attention to core aspects of social identity, such as the practice of veiling, and related fields of representation can illuminate social psychological processes and further the development of a social conception of agency, as I now explore.

The role of representations in identity construction

A starting point for my research was a wish to interrogate norms of agency relating to veiling by listening directly to the perspectives and experiences of Muslim women living everyday lives in northern European contexts. Through individual and group interviews with 38 second-generation Muslim women in the UK and Denmark, ¹ I sought to explore the specificities of women’s negotiations of gender identity through their accounts and discussions of the veil – an item of dress that expresses social identification and serves as a visual cue for others’ representations, thereby becoming a site of struggle for identity (Wagner et al., 2012).

Participants were recruited in Copenhagen and London through diverse routes – including employment faith networks, places of education, mosques and community projects – in order to ‘maximize the variety of representations by extending the range of strata’ (Bauer and Aarts, 2000: 33). Profiles reflected a range of family backgrounds, ethnic origins and religious traditions. Most of the women wore a headscarf and a few wore a face-covering niqab. Styles of veiling and forms of dress were varied and reflected the diversity that may be seen in multicultural areas of Copenhagen and London. The intention was not to distinguish a priori between different practices but to remain open to the patterns of social identity work that might emerge.

The analysis was thematic in approach (Attride-Stirling, 2001) and assisted by the qualitative analysis software NVivo 10. It was guided by the two central related concerns of social representations theory: first, the substance of core complexes of meaning that forge particular social identities; and, second, the function of self–other encounters and the meeting of different systems of representation for the ongoing dialogical construction of meaning and identity. The analysis revealed patterns of representational dynamics across the two research fields and also to a large extent across a range of family backgrounds, ethnic origins, religious traditions and veiling practices. This suggests the significance of veiling for the construction of identity both in terms of ways of knowing that are rooted in belonging to a Muslim faith, family and community and in terms of shared experiences and challenges of living in northern European societies.

A majority of the women in my study started to veil under the age of 16 and grew up in households where it was customary. For most, gendered virtues relating to veiling have been a core part of the representational field of gender that has surrounded them from childhood. Their struggles over identity must therefore be located in this constituting context. As Duveen (2001) observes in critique of social identity theory’s emphasis on self-categorization and an innate drive for individual esteem: ‘Before it becomes thematized as a struggle for the individual, an identity is first a social location, a space made available within the representational structures of the social world’ (2001: 268). A story told by my respondent Zahira (Copenhagen) is illustrative. When asked about her reasons for wearing the headscarf, Zahira recalls that in the week before she started to wear it permanently, at the age of 15, she wore ‘for the first time in my life, a short blouse and tight jeans’ to school. She says: ‘I think I wanted to know if it was that way I wanted to follow but I got a very bad conscience.’ Zahira relates her bad conscience to ‘a Danish guy commenting’ on her appearance:

And nobody had ever commented on my clothes before and I thought okay, that means that people notice it. That means the things you say about the headscarf, that it’s covering and it’s protecting, that must be right then. I think it was that connection I made. It was not the reason I wore it. It was a decision I had made over the past two or three years.

Zahira’s identification of her fellow student as Danish and her description of veiling as ‘protecting’ suggest her sense of difference and of contrasting value systems over dress as a site of identity. This encounter implicates Zahira’s ‘conscience’ and requires that she take a position in relation to particular representations of gender and sexuality. To narrate her affirmation of veiling here in terms of resistance to a sexualized culture (or submission to patriarchal religious codes) would be to misconstrue processes of identity and reproduce what Mahmood refers to as the ‘flat narrative of succumbing to or resisting relations of domination’ (2005: 174). That is not to deny relations of power and resistance. Like Zahira, all my respondents are negotiating veiling and sexuality at the intersections of familial and religio-cultural contexts and wider cultural forces that exert particular influences on girls and women. Rather, it is to situate them as part of struggles for identity and meaning that cannot be conceived apart from the attachments, values and ‘consciences’ that particular ways of belonging and knowing forge.

Zahira’s reflections here serve to highlight the social location of identity as well as the dynamics of representation and identity in the intersubjective encounter. Representations may be re-evaluated or they may be socially enacted and affirmed as we interact with others and encounter social forces that sustain or threaten particular aspects of identity and meaning (Duveen, 2001). Social representations research is concerned with the social psychology of this encounter and how it ‘throws self, individual and collective into uncertain directions, offering new models for its identification project while by the same token hardening its defences’ (Jovchelovitch, 2007: 129). As work on identity and representation demonstrates, identity cannot escape but must always negotiate ‘the gaze of the other’ and the particular representations conveyed (Hall, 1997; Howarth, 2002a). These processes of representation and identity are at work in a research encounter like any other dialogical encounter involving relations of similarity and difference (Howarth, 2002b; Khawaja and Mørch, 2009). Analysis of these processes in my research interviews therefore sheds light on my respondents’ negotiations of identity, as the following extract from my interview with Ayesha illustrates.

In response to Ayesha’s account of her experience of social pressure to wear a niqab during a visit to Saudi Arabia, I ask her how that made her ‘feel about veiling’. Ayesha replies:

Every place will make you feel like you need to dress a certain way … I remember going out, years ago, to Bluewater with my sister and we were in a lift and me and my sister were wearing long dresses like a jilbab and these two white ladies started talking about how horrible it was that in this heat we had to cover … And my sister at that point used to wear a niqab and that’s what brought about the woman’s comments. Then she went, ‘Why do you do it? Why do you let them do it to you?’ That’s what ticked me off and I just turned around and I went, ‘What made you think anybody told her to do it?’ … I became a bit, I was a bit stroppy … I was resentful of the way she attacked my sister and said ‘Why do you let them do that to you?’ Those were her words. Because I thought, my sister’s done this, that’s her choice.

Attention to the role of representations here necessarily involves critical reflexivity in relation to my identity as a white non-Muslim researcher and the content and impact of what I say. In asking Ayesha how social pressure to wear a niqab made her ‘feel about veiling’, I convey a norm of valuing that suggests the idea of veiling as inherently oppressive. In telling her story in response, Ayesha is able to articulate explicitly the stigmatizing representation of her identity and contest it. First, by observing that ‘every place will make you feel you need to dress a certain way’, Ayesha draws a parallel between sartorial norms in different contexts and thereby challenges assumptions of coercion and oppression in relation to veiling. Second, she contests the representation of Muslim women as subjugated and passive by presenting herself in this encounter as confident and assertive (‘stroppy’). Third, in describing her defence of her sister’s niqab, Ayesha uses the language of individual choice and autonomy that counters ideas of coercion and subordination to men (‘my sister’s done this, that’s her choice’). Significantly, Ayesha does not discuss veiling here by elaborating ideas of submission to religious duty and protection for women as she does elsewhere in the interview. Ayesha confronts stigma through amplifying ideas of individual autonomy, choice and assertiveness – ideas that are central to liberal norms of agency and used against Muslim women and values (Bilge, 2010; Mahmood, 2005).

Analysis of the role of representations across the interview with Ayesha reveals that she draws on different and sometimes apparently contradictory modes of reasoning in relation to the veil depending on the intersubjective context. It thereby draws attention to the relations of power that exist for Ayesha socially as a veiled Muslim woman and how this identity demands negotiation of the content of others’ representations. It sheds light on how identities are challenged in the self–other encounter and constructed through and against particular representations. This is both a threatening and productive process for identity, ‘hardening its defences’ while at the same time ‘offering new models’ for identification (Jovchelovitch, 2007: 129). A few of my respondents’ accounts suggest that, for them, this process entails forces of polarization. For example, Isma (Copenhagen) describes a troubled sense of ‘double identity’ and the experience of dressing and behaving in very different ways in different contexts. This highlights the possibility on the spectrum of Muslim women’s experiences of a non-dialogical conflict between competing social contexts and representational systems that may end up hardening defences in particular ways.

That being said, most of my respondents discuss veiling in ways that suggest both the functional use of different modes of reasoning to negotiate specific challenges, as illustrated in the interview with Ayesha; and, in parallel, ongoing evaluation of beliefs, values and practices that involve negotiations of relations with family and community as well as social relations in contexts such as education and employment. That is, their accounts suggest that, in both drawing on and resisting different forms of reasoning, new perspectives and social identity positions emerge. As Provencher (2008) observes in her account of cognitive polyphasia: ‘The transformation of social representations may therefore happen through the use of different types of knowledge that bring to light different dimensions and perspectives about an issue or social object previously taken for granted’ (2008: 268).

Such dynamics of perspective and re-evaluation in relation to veiling are revealed in my interview with Miriam (London). Miriam tells me that she first became conscious of outsiders’ perspectives when she attended school in her headscarf for the first time at the age of seven and became ‘very aware’ of it through the gaze and comment of others. She describes years of struggle over identity and feelings of resistance to some of the gender values she was brought up with in relation to veiling:

I’m quite loud and, you know, I just laugh at everything and I’m not really a kind of prim and proper girl, and I was always told by my cousins to sort of constrain myself because that’s not the way a girl is supposed to act, especially a girl who is wearing a headscarf. So when you’re a teenager and you’re just being yourself, that just stifles you in a sense, you feel a bit rebellious and don’t like it.

As evident here, Miriam’s account of herself emphasizes and embraces ideas of individual personality and assertiveness (‘I’m quite loud … I’m not really a kind of prim and proper girl’) in ways that challenge aspects of the gendered religious values of her upbringing and draw on ideas of individualized agency (S Mørch et al., 2013). Intersubjective dynamics of representation and resistance in the interview suggest the dialogical processes involved in Miriam’s negotiation of identity. For example, when I ask Miriam if ‘there’s a part of it [veiling] that’s about modesty’, she responds emphatically: ‘To me, it’s not about covering up my sexuality or anything. I like to wear it to show that you can be like that and be completely just like anybody else.’ In responding in this way to my use of the term modesty, Miriam confronts and contests a representation of her identity that positions it in opposition to liberal sexual mores – mores that are associated with autonomy and liberation (Gill and Donaghue, 2013) and central to the ‘gender-equality-and-sexual-freedoms frame’ that constructs Muslim otherness (Bilge, 2010: 10). In contesting this representation, Miriam elaborates identity in ways that integrate liberal-secular values and norms of agency.

At the same time, Miriam identifies as religious, dresses in ways congruent with traditional ideas of modesty, and describes feeling ‘indebted’ to the headscarf because of how it has ‘protected’ her from ‘wrong situations’ growing up. She describes its function in both ensuring that others ‘respected me’ and ‘calming’ her personality down. As Miriam talks with me, therefore, she variously draws on and resists ideas and values of both permissive and modest sexuality that are part of competing fields of representations relating to gender and her threatened social identity as a veiled Muslim woman. Attention to the content of representations and the intersubjective dynamics of the interview therefore suggest ongoing evaluation of the veil’s meaning and function for identity that is tied to social negotiation of specific contexts, challenges and pressures.

These dynamics of identity are evident across the two research fields. However, one apparent difference concerns how these dynamics relate to familial and community contexts. For the women in Copenhagen, community pressures in relation to veiling appear to be more pronounced than for the women in London. ² For example, while Miriam (London) relates struggles over veiling to familial expectations, Sana (Copenhagen) discusses them in terms of the pressures and expectations of a close community context. Sana describes how ‘guys … are allowed to go out, be with girls, do whatever, but if a girl does the same, it’s, it’s haram [forbidden]’. She gives the following example:

There was this concert, some Egyptian group that came to Denmark and she [a friend] … went to this place called Rust and when she entered, she said everyone was looking at her because she was wearing the scarf. And she was like, I am allowed to be here, why are you looking? Her [male Muslim] friends, they came over and said, in a loving tone, but like, you shouldn’t be, it’s odd that you’re standing here in front of me. And she was like, why? You’re here as well. They link it to something, she is more religious than me, she’s not supposed to be here, but they think it’s fine that they’re in the club themselves.

While Sana emphasizes that her own upbringing was not restrictive or ‘unequal’, she reflects how many of her female Muslim friends struggle with veiling and related gender norms, particularly in light of ‘Danish values’:

It’s like an indication of what kind of gender differences we are raised with here in Denmark, from our home countries and the culture. There’s a huge gender difference and that’s the only reason I am stressing to you that I was not brought up unequal because the normal in my head is that girls are not allowed to do the same as boys are. And that’s not what you see in the Danish culture and the Danish values.

To sum up, my respondents’ accounts of veiling tell how their identities are continually challenged in the intersubjective encounter and negotiated at complex intersections of traditional-religious and liberal-secular representational structures. Attention to interrelated dynamics of representation and identity in this way yields an intersectional perspective: it reveals the significance of identification as both Muslim and female and the nexus of power relations that exists for Muslim women (Bilge, 2010; Brah and Phoenix, 2004). It also thereby provides an empirically and theoretically instructive contrast with Mahmood’s (2005) study. Mahmood describes how the women in her study cultivated subjectivities through practices and ideas of piety and modesty that run counter to liberal norms of autonomy, agency and equality. By contrast, my interviews with mostly second-generation Muslim women in northern European contexts show how ideas, values and practices relating to Islam and veiling interact with other ideas, values and practices in the social field. In her discussion and critique of feminist norms of agency that obliterate religious subjectivities, Mahmood observes:

It is clear that certain values (such as humility, modesty, and shyness) have lost their value in the liberal imagination and are considered emblematic of passivity and inaction, especially if they don’t uphold the autonomy of the individual. (2005: 174)

Analysis of the interplays of representation and identity in my interviews suggests that my respondents construct identity through and against both values of modesty, humility and shyness and the ‘liberal imagination’ that diminishes these values and upholds opposing norms of autonomy and liberation.

This analysis helps highlight the distinction between particular ideas of agency and the dialectical dynamics of representation and identity that provides a conceptual location for a social and contextual account of agency. Agency is not reduced to a particular idea of agency but located in the social representational processes of gender identity work that engage particular norms of agency. Approached this way, feminist research can be troubled less by what agency is and can focus on the work that agency does. From a social representations perspective, this work is identified in the specific ways people affirm, resist and contest ideas, beliefs and values in the threatening but also productive encounter with difference (Jovchelovitch, 2007). For the women in my study, identities are co-constructed at the intersections of being Muslim, veiled and female in a European context, revealing patterns of dialogical engagement and the cultural forces and relations of power that engage processes of resistance and re-presentation.

Conclusion

As these reflections on my research have shown, analysing the role of representations in the construction of identity helps advance a social and contextual conception of agency. In a number of ways, it can help address tensions within feminist research between narratives of coercion and the turn to individualistic notions of agency.

First, research on representational dynamics in the self–other encounter reveals the continual dialectic between processes of representation and identification. It thereby overcomes the tensions between ideas of the culturally dislocated unified individual and rather limiting analyses of subjectivity as a product of power. A social representations account of how regularities of social knowledge and social identity are variously reproduced in processes of social interaction counters the dualisms of individual versus social and mind versus world that continue to limit research in psychology and other fields (Ewing, 2006; Farr, 1996; Jovchelovitch, 2007).

Second, and relatedly, the theory of social representations addresses, at different levels, the productive plurality and diversification of knowledge in increasingly multicultural and interconnected societies. The concept of cognitive polyphasia underpins an approach that respects different rationalities and their attachments to identity rather than treats them as subject to processes of replacement (Jovchelovitch, 2002). This resonates with Mahmood’s (2005) ethnographic embrace of ‘forms of human flourishing outside the bounds of a liberal progressive imaginary’ (2005: 155). However, while Mahmood’s focus is on giving an account of self-formation through particular modes of reasoning and practice, the focus of social representations research on identity is on the implications of the self-other encounter and the possibilities of dialogical exchange and transformation that are generated (Voelklein and Howarth, 2005). Jovchelovitch (2007) refers to the ambivalent, contradictory and potentially destructive aspects of knowledge encounters but also to ‘the redemptive potential of the intersubjective’ (2007: 138), the articulation of which she views as missing in otherwise powerful poststructuralist accounts of subjectivity.

These key aspects of social representations theory offer a basis for a social conception of agency that is rooted in the dynamics of representation and identity in the relational encounter. It is an account that highlights the significance of the gaze of the other and how we must continually negotiate others’ representations of our identity (Hall, 1997; Howarth, 2002a). It thereby draws attention to the significance of social categorizations of identity and assists in the important project of providing an intersectional account of identity and agency as tied to specific historico-cultural contexts (Bilge, 2010; Brah and Phoenix, 2004).

In my research, attention to the role of representations in dialogue provides a way to approach the complexities of my respondents’ cultural negotiations. It enables an understanding of how my respondents continually negotiate representations of gender in ways that reflect both their attachments to Islam, family and community and their belonging within liberal-secular societies. It becomes possible to recognize that identities are struggled for at these intersections and constructed as positions in relation to particular representations and the powerful norms of agency they engage. My analysis does not therefore deny dynamics of coercive influence and resistance but shows that they work in multiple directions as women confront and negotiate stigmatization, a sexualized culture, and specific religious, familial and community contexts.

In this way, a social representations approach draws attention to the cultural constraints and challenges of gender identity as well as to dialogical processes of re-presentation and resistance that defy blunt political judgement. Feminist research seeking to engage with the complexities of agency and gender-related dynamics can, therefore, benefit from examining the role of representations in identity construction and the multiple gazes that are significant in that process.