Introduction

There is a long history of legal engagement with the problem of forced marriage in the UK. However, it is only recently that this issue has become the focus of media attention and policy debate, with the emergence of three high profile cases: the murder of Rukhsana Naz by her family after she left an arranged marriage and became pregnant by another man (Hall 1999); the plight of Jack and Zena Briggs, who were forced into hiding by bounty hunters employed by Zena’s family after she had refused to marry one of her cousins in Pakistan (Watt 1999); and the successful return to England of a young Sikh girl after her parents had abducted her and taken her to India to be married.Footnote 1 Following the publicity surrounding these cases, a Home Office report on forced marriage was published in 2000, which drew a distinction between arranged marriages (which operate by tradition in numerous countries worldwide and in certain minority communities in the UK) and forced marriages, on the basis of consent. Arranged marriage was defined as a marriage facilitated by family but which requires the consent of both partners, and a forced marriage as one to which either one or both parties do not give their consent, or do so only under duress (Uddin and Ahmed 2000, p. 10).

Despite the attempt to signal the acceptability of minority cultural practices by the adoption of the somewhat misleading binary distinction between arranged and forced marriage in policy documents, media and policy-related discourses continue to frame the problem of forced marriage in cultural terms, rather than as a specific manifestation of a wider problem of violence against women. In this debate, forced marriage is constructed as a culturally sanctioned aberration from the norm of freely contracted marriages, and little attention is given to the many ways in which all women located within a matrix of structural inequalities can face social expectations, pressure and constraint in matters of marriage. The UK media in particular treats coercion as a salient feature of particular communities, which they represent as “patriarchal and inherently uncivilised” (Razack 2004, p. 129). Women within these communities are re-presented as particularly oppressed and lacking in agency, unlike the putatively liberated women in the West who are seen “as educated, as modern, as having control over their own bodies and sexualities, and the freedom to make their own decisions” (Mohanty 1988, p. 65). These discourses on forced marriage also invoke specific cultural traits or religion (the two are often used interchangeably) to explain violence against Third World women, whereas culture is not similarly invoked to explain the forms of violence that affect mainstream Western women, a problem transnational feminist scholars have highlighted in other contexts (Narayan 1997, pp. 84–85; Abu-Odeh 1997; Volpp 2000).

There are many consequences of adopting this colonialist stance towards cultures of Black and minority ethnic (BME) communities, but for our purposes it is important to note that this approach conceives tradition as static and ahistorical, and thereby effaces all ideologies and politics that constitute a set of practices within a given context as ‘traditional’. Not only does this disregard the agency of women, whose resistance shapes the contours of what is constituted as culture, but by extricating gender from the social relations that produce particular constructions of femininities, the category of women in such scholarship is thereby homogenised: they become stereotyped as “Third World women being victimised by Traditional Patriarchal Cultural Practices” (Narayan 1997, p. 57), the always already formed victim-subject.

These ‘abuses of culture’ (Dustin and Phillips 2008) do not mean that we should ignore the specific socio-political contexts that underpin the practice of forced marriage. As Purna Sen argues, just as it is flawed to posit a cultural specificity that fails to see the linkages between particular manifestations of violence against women, “to deny specificity if it exists is also problematic” (Sen 2005, p. 50). Research indicates that within the UK, forced marriage is a problem that primarily affects women originating from the Indian sub-continent (FCO and Home Office 2005, p. 15) as well as smaller numbers of women originating from Iran, Afghanistan, Turkey, Armenia, Somalia, Eritrea and Sudan, and Irish traveller communities (Hester et al. 2008). Accurate figures on the incidence of forced marriage within the UK are hard to come by, but estimates range from 450 (Newham Asian Women’s Project 2007) to 1,000 (Southall Black Sisters 2001) cases a year. Recent media reports have spoken of the “500 missing girls” who are thought to have been removed from schools in the UK and forced into a marriage in the Indian subcontinent, and some reports even speculate that there are almost 4,000 cases of forced marriage a year (Taylor 2008). Reliable figures are hard to obtain, in part because of the difficulties of distinguishing between coercion and consent: research among South Asian communities in the UK indicates that while most people perceive a difference between arranged and forced marriages, they also recognise some overlap (Gangoli et al. 2006). Additionally, this uncertainty about the extent of forced marriage is also a consequence of the underreporting of forced marriage, as with any form of violence against women.

This article is specifically focused on the problem of forced marriage as it affects the South Asian communities, in part because these are the communities that we know best, and in part because statistics clearly indicate that the majority of women who seek help from services when faced with an unwanted marriage are of South Asian origin.Footnote 2 Clearly, however, many of the issues raised here are relevant to the many other communities in which women face the problem of coercion in marriage. Only by understanding and engaging with the particular forms that violence against women takes in different communities can we tailor support and resources to the specific needs of women in these differing contexts, and so this is in part a call for further work to be done to alleviate the suffering of women. As Dustin and Phillips (2008) point out, we should not exaggerate the cultural component of what is at root a form of domestic violence—the failure to recognise this could cost lives.

Specialist domestic violence services, with whom we have worked for over 10 years now, have been speaking out about and tailoring service responses to the problem of forced marriage long before it came into the public eye. In doing so, they have been at the forefront of the debates over what constitutes ‘tradition’ within our communities. Their work has demonstrated that one way to go about understanding ‘cultural differences’—the specific contexts and meanings associated with forced marriage—without resorting to essentialist explanations, is to restore a sense of politics to the predominantly ahistorical picture of certain cultures, so that the dissensions and contestations through which particular values have come to be deemed as normative can be scrutinised (Narayan 1998, p. 93).

The purpose of this article is threefold. By drawing upon a combination of hypothetical scenarios, case law and existing research findings, the first objective is to interrogate the concrete distinction between coercion and consent in marriage that has been employed by courts in the UK as well as in Western liberal theorising on coercion. Building on this analysis, the second and third objectives are to uncover how existing conceptualisations of coercion shape both the discourse on forced and arranged marriage and practical approaches to its legal resolution in the UK, and to examine the implications of these conceptualisations of coercion for the theory and practice of gender justice. In doing so, this article attempts to understand forced marriage by drawing our attention to the myriad ways in which women, whatever their culture, ethnicity, religion or class, experience pressure to marry, thereby also culturalising the concept of coercion in marriage within mainstream communities. We argue that the specificity of forced marriages, the particular meanings and manifestations of it within diasporic communities from South Asia and the way in which the law fails to better understand consent and coercion must be unpacked within this wider context without descending into explanations that imply cultural deficit. In the last section, we consider how women exercise their agency in complex and contradictory ways within the intersections of various vectors of power, a process that eludes a binary conceptualisation of coercion and consent.

Forced Marriage and the Law in the UK: A Brief History

Though a marriage contracted on the basis of coercion has long been illegal in the UK, in response to the publicity surrounding the high-profile cases mentioned above, new legislation was proposed in 2005 to treat forced marriage as a specific criminal offence. Hastily composed (perhaps because of the media publicity) and superficial, the suggested legislation was focused on matters of law rather than on the subtleties of socio-cultural practice. Non-Governmental Organisations (NGOs) criticized the proposed legislation on the basis that it would be ineffective (Gill 2005; Mookherjee and Reddy 2005), would reinforce racist stereotypes (Imkaan 2006) and would fragment laws pertaining to violence against women. A civil Bill was then proposed, and although NGOs broadly welcomed the victim-centred process of civil law, they remained sceptical about the draft Bill because it failed to integrate the discourse of forced marriage within the wider framework of violence against women, thereby reinforcing essentialist understandings of forced marriage as a ‘cultural problem’. Women’s groups called for a civil response to be established within the Family Law Act 1996 (FLA).Footnote 3

The Bill received Royal Assent, as the Forced Marriage (Civil Protection) Act 2007, on 26 July 2007, and was incorporated as a new Part 4A into the FLA. ‘Force’ is defined in s 63A(6) as meaning to “coerce by threats or other psychological means”. Under the new law, forced marriage was rendered an actionable and specific civil wrong, with remedies (chiefly forced marriage protection orders) designed to protect those threatened with or subjected to forced marriages. The Act makes it easier for applications to be made before a marriage has taken place, and makes clear that it is unlawful for people inside and outside the immediate family to ‘aid and abet’ forced marriage. It expressly covers deception for the purpose of causing a forced marriage. Overall, the Act simplifies the process by which victims can seek help.

Before we can properly understand the contemporary legal approach to forced marriage, we have to know something about its history—not least because much of the current legislation is based upon presuppositions that have gone unchallenged for several decades. Before the mid-1980s, coercion or duress was interpreted narrowly by the courts and taken to mean either physical force or threat of physical force—“an immediate threat to life, limb or liberty” (Phillips and Dustin 2004) that undermined the free will of the coercee. In the case of Hussein (otherwise Blitz) v Hussein,Footnote 4 for example, the petitioner asked that her marriage to the respondent be annulled on the grounds that she had only contracted the marriage because he had repeatedly threatened to kill her if she did not. In granting a decree nisi, the court upheld that her free will had been undermined by the respondent’s conduct and that she had been coerced into the marriage. This case is quite representative: the most common claims of duress that have come up before courts have been duress imposed either by one of the parties to the marriage or by the family of one of the parties (McLarnon v McLarnon;Footnote 5 Harper v Harper;Footnote 6 Re KR Footnote 7).

The courts have also long considered the coercive power of physical force posed by third parties such as the state and by impersonal sources such as social and political circumstances. In the case of Buckland v Buckland,Footnote 8 the petitioner was charged under Maltese law with the corruption of a 15-year-old girl. He protested his innocence, but was advised that he would almost certainly be found guilty, imprisoned, and ordered to support the child he was believed to have fathered. The only alternative to this would be to marry the girl, which he did. After acquiring domicile in England, he petitioned to have the marriage annulled on the grounds that fear had vitiated his consent. In granting his petition, the court held that he had only agreed to the marriage due to his reasonably held fear of imprisonment. The coercive capacity of social and political circumstances was tested in the case of H v H,Footnote 9 in which the petitioner feared for her life, liberty and ‘virtue’ due to the dangerous political climate in Hungary where she lived. In order to leave, she went through a marriage ceremony with her cousin, a French citizen, and following her escape from Hungary, petitioned for the marriage to be annulled on grounds of duress. The court upheld the undefended petition on the basis that the petitioner had not genuinely consented to the marriage, but that her free will had been undermined by external circumstances which necessitated her escape from Hungary.

In the cases discussed above, the focus of the courts has been on the so-called ‘choice-prong’ (Wertheimer 1987). This refers to a bi-conditional proposal whereby the coerced person is presented with two choices: an unwanted marriage or violence to her person. Theoretically, the petitioner still has a choice, but the choice is between two unpalatable alternatives, and the petitioner has to give up one right—to a free choice in marriage—in order to protect another—her right to safety and/or liberty. The petitioner’s case seems to rest on whether the courts believe that the alternative to marriage is so undesirable that any reasonable person would seek to avoid it at all costs; if it is, then the marriage can be said to have taken place under duress and not through free will. It was on this basis that emotional pressure was for a long time considered to be insufficient to overcome the will of a petitioner, unlike physical violence or threat of such violence (Singh v Kaur Footnote 10). For this reason alone, cases involving coercion, as we now understand it, were ignored and claims rejected.

It was the case of Hirani v Hirani Footnote 11 which marked a new phase in the understanding of coercion. The application was made by a 19-year-old Hindu woman whose parents had arranged a marriage to prevent her from associating with a Muslim man. Having separated from her husband after only 6 weeks of marriage, she petitioned for a decree nisi on the grounds of duress exercised by her parents, upon whom she was financially dependent and who had threatened to turn her out of the home if she did not go through with the marriage. Although her petition was initially rejected, in allowing her the right to appeal and eventually granting her the decree nisi, the court held that: “The crucial question…is whether the threats, pressure, or whatever it is, is such as to destroy the reality of consent and overbear the will of the individual”. Phillips and Dustin (2004, p. 538) have pointed out that this judgment marks a definite shift in legal rhetoric, from a restrictive definition of duress centred upon threats of physical violence to one in which the key issue was “whether the mind of the applicant (the victim) has in fact been overborne, howsoever that was caused”. Subsequent cases (Mahmood v Mahmood;Footnote 12 Mahmud v Mahmud;Footnote 13 Sohrab v Khan Footnote 14) have involved acceptance by the courts that emotional pressure can take a variety of forms, from being made to feel responsible for bringing about a loved one’s death to threats of suicide made by the coercer. More recently, when forced marriage was feared in the case of an adult woman who went ‘missing’ after being taken to Bangladesh (Re SK Footnote 15), Justice Singer issued a summons to SK’s family to produce her at the British High Commission and granted injunctive relief restraining her family from threatening, intimidating or harassing her, using violence against her or proceeding with her wedding. Singer J argued:

that there is a spectrum of forced marriage, from physical force or fear of injury or death in their most literal form, through to the undue imposition of emotional pressure which is at the other end of the forced marriage range, and that a grey area then separates unacceptable forced marriage from marriages arranged traditionally which are in no way to be condemned, but rather supported as a conventional concept in many societies. Social expectations can of themselves impose emotional pressure, and the grey area to which I have referred is where one may slip into the other: arranged may become forced but forced is always different from arranged.Footnote 16

Despite the extension of the definitional ambit of coercion to include such hazily defined “grey areas” and the allusion to social expectations, the notion of ‘free will’ remains central to the legal discourse on forced marriage in the UK, where what is seen as the norm of self-constituting, free individuals entering into a consensual marriage is contrasted with a marriage contracted through coercion. Singer J’s “grey area” is still not murky enough to disturb the clear distinction between forced and arranged marriages.

The problem is that the preoccupation with ‘free will’ that informs the legal discourse ignores the fact that consent itself is constructed in the context of power imbalances and gendered norms, and—crucially—often in the absence of explicit threats: simply put, many coercive forces often go undetected. Another problem is that the conceptualisation of the legal subject as an autonomous agent who is able to choose and act freely is not a gender-less, race-less being; this notion of the free self is predicated on the normative experiences of a white man. In the following sections, we shall examine how the prevailing discourses on forced marriage do not capture the experiences of South Asian women in the UK who encounter the problem of coercion in a marriage.

Towards a Nuanced Approach to Forced Marriage: From Coercive Intent to Coercive Burden

The Home Office report on forced marriage (Uddin and Ahmed 2000) emphasised the distinction between arranged and forced marriage to allay community concerns that this initiative might cast suspicion on all arranged marriages. The courts, too, have been anxious to distinguish between the two types of marriage, with troubling consequences. In two cases where the petition for nullity was granted on grounds of emotional duress, the courts simultaneously and somewhat paradoxically also re-asserted the legitimacy of parental pressure. In Mahmood v Mahmood,Footnote 17 the judge commented, “I accept entirely that parental consent is perfectly legitimate and proper…I also accept that the consent which has to be given to marriage need not be enthusiastic consent, but even reluctant consent will suffice provided that the consent is genuine” (our emphasis). And in Mahmud v Mahmud,Footnote 18 the judge stressed that if “under pressure – and perhaps very considerable pressure—a party does indeed change his or her mind and consents to a marriage, with however ill a grace and however resentfully, then the marriage is in my opinion valid”. The problem here is a pressing one: how can “reluctant” or “resentful” consent on the basis of “very considerable pressure” not amount to marriage under coercion? In essence, although there was some recognition of the psychological pressures brought to bear in forced marriage cases, the patriarchal structures remained unexplored beneath this façade of an individual ultimately free to choose and to change her mind. Not only does this ignore the covert, subtle forms of coercion brought to bear upon women, but there is also little recognition of the structural constraints under which women exercise their agency in matters of marriage.

Feminists have long recognised the variety of pressures on women to marry, including poverty, pregnancy and social norms and expectations that are underpinned by the patriarchal structures of their culture, religion and state. Indeed, research on the marriage practices of South Asian women in the UK also indicates that the sanctions that underpin the dominant moral codes of their society are very apparent to the young women, and play a significant role in the exercise of their agency. As one woman said: “If a girl says no, it’s considered a bad thing”, and “if you didn’t [go along with the marriage] there would be hell to pay from your parents and all your relatives” (Bhopal 1999, p. 121; also see Gangoli et al. 2006, p. 10; Wilson 2006, pp. 18–19, 91–93). A woman’s family history and the pressure of social and cultural expectation may well convince her that if she withholds consent then the full force of physical or emotional threats will come to bear upon her, to the extent that she is likely to eventually succumb to this duress. It could be argued that it might be possible for a court in such a situation to test the strength and the reasonableness of her convictions empirically, on the basis of the past experiences of others in the family (for example, a sister who was disowned for refusing to consent to an arranged marriage). However, control is exercised in ways that are far more subtle than the dramas imagined in current definitions of forced marriage. It is within a range of constraints, both articulated and unstated, that particular groups of South Asian women exercise their agency in determining, to varying degrees, whether or not to marry, their choice of marriage partner or the timing of the marriage.

In this light, any test to determine whether or not coercion has taken place that focuses on the extent of the pressure (physical or emotional) that has actively been brought to bear on the victim might not reveal what Feinberg (1986) calls the “total burden of coercion” that the victim experiences. The “total burden” reflects the experiences of the individual, so that any decision about whether or not coercion has taken place is forced to acknowledge how pervasive, frightful, unwelcome and/or intense any pressure is, in order to determine whether or not the proposal coerces. Feinberg’s (1986) approach is useful precisely because, as we have already suggested, explicit threats do not exhaust the range of coercive techniques and structures experienced by South Asian women in the UK who face pressure to marry.

For courts in the UK, and indeed in most philosophical approaches to the analysis of coercion, what is critical is whether or not the bi-conditional proposal represents a false choice, because both the alternatives would leave the coercee worse off when compared to a baseline. This baseline has been variously defined by comparing the expected course of events, as determined by the subject’s past, with what is otherwise ‘normal’ in a society (Nozick 1969, p. 447), or by making reference to the subject’s rights or an external notion of justice (Wertheimer 1987; Airaksinen 1988). An offer is distinguished from a threat by considering whether the proposal, if accepted, would leave the individual better or worse off with respect to this baseline. We would argue that there are contexts in which the difference between a coercive proposal and an offer is not as clear-cut as is presumed in this literature.

Feminists have long argued that there is a problem with the very notion of marital consent under conditions of patriarchy, involving the construction of femininities and gender inequalities (Pateman 1988; Okin 1989). A recent study on the problem of forced marriage among South Asian communities in the UK reported the disturbing case of a young woman in the UK who was raped and then forced to accept her assailant’s proposal of marriage in order to preserve the family honour (Gangoli et al. 2006, p. 23). There is significant stigma attached to rape in South Asian communities, where the woman is often blamed for the crime committed against her and may find herself ‘unmarriageable’. Because the proposal maker causes the proposal receiver’s adversity, this proposal would be seen as coercive in most Western liberal accounts of coercion. However, where acceptance of a proposal causes an improvement over the pre-proposal situation, and that situation has not been caused by the proposal-maker, it has been argued that this constitutes an offer and is hence not coercive (Zimmerman 1981; Wertheimer 1987). And yet the crucial determinant of coerciveness cannot be whether the proposal-maker has caused the adverse circumstances; in the absence of entrenched gender inequalities these proposals could not have been endorsed by the woman’s family. In order to see the pervasive power of these inequalities, consider that the above proposal was made by a third party who was not responsible for the woman’s immediate circumstances; even so, such an offer would be one that many women in this situation would not ‘reasonably’ be able to refuse.

Not all forced marriages begin with a rape, of course, but there exist less dramatic (though still persistent) inequalities that add to the total coercive burden experienced by women in matters of marriage. African-American feminists have drawn attention to the ways in which various forms of social inequality, including race, immigration status and class, intersect to create a qualitatively distinct “matrix of domination” (Collins 2000) for women of colour (Crenshaw 1991). As touched upon above (footnote 2), for South Asian women in the UK, numerous influences (including state policies, particularly immigration policies, the place of their country of origin in the international order and their own diasporic experience) intersect to shape their experience of coercion in matters of marriage and their resistance to such pressures. South Asian women’s groups have long criticised the designation of BME communities as ‘faith communities’ and the state practice of privileging (male) figureheads of organised religious institutions as ‘community spokespersons’, a practice which reinforces prevailing hierarchies within these communities (Wilson 2006; Patel 2008).

The current emphasis on the uniqueness of forms of domestic violence such as forced marriage and so-called ‘honour crimes’, informs and is informed by racist and essentialist stereotypes and feeds into anti-immigration agendas (Dustin and Phillips 2008). BME women’s experience of violence within their families is not separable from the wider structures of racism. Research indicates that women from Black and other minority ethnic communities are less likely to access statutory services for help with the violence they are facing, for a variety of reasons, including fear of racism and the desire to avoid reinforcing essentialist stereotypes of their communities through the normalisation of abuse within minority ethnic communities (Batsleer et al. 2002; Rai and Thiara 1997). With the recent policy shift towards ‘community cohesion’, the domestic violence services that cater to the needs of BME women face closures, thereby removing the only accessible routes for many South Asian women who seek support to resist coercion and violence within the family.

It has also been argued that the diasporic experience creates an additional dimension to the meaning of forced marriage, as parents impose an unwanted marriage to stem the influence of Western culture over their daughters or end their daughters’ associations with ‘unsuitable partners’ (Samad and Eade 2002, p. 67; Gangoli et al. 2006, pp. 13–14). Similarly, it is in the context of the immigration policies of the state that consanguineous marriage may be re-asserted as a traditional cultural practice.Footnote 19 Successive legislation has closed immigration routes into the UK, and marriage remains one of the few means of settlement there. Research indicates that the factors that shape the decisions and choices of marriage arrangers in the particular context of Pakistani migration to Britain include a sense of obligation to kin and the need to maintain links with home communities through the use of marriage as an immigration strategy (Shaw 2001). This practice also varies widely according to region of origin, caste, socio-economic status and family history within the Pakistani communities in the UK, and coercion is a factor in some, though clearly not all, consanguineous marriages, and is a gendered experience (Shaw 2001, pp. 329–331).

In the absence of explicit threats, the coercive potential that arises from great imbalances in power or prior historical injustices between bargaining parties has been recognised in previous analyses of coercion (for example, Feinberg 1986; McGregor 1988–1989). Within these constraints, however, women exercise agency in myriad ways, a fact which undermines efforts to strictly polarise notions of coercion and consent. The myth of free choice, however, persists in Western liberal theory, and is even enshrined in one piece of legal rhetoric: ‘reasonable alternative’. In the next section, we will examine this concept, which underpins the legal discourse on forced marriage in the UK.

‘Reasonable Alternative’ and the Gendered Legal Subject

In the legal discourse on forced marriage in the UK, the test of ‘reasonable alternative’ is used to determine whether a marriage has taken place under duress. In the absence of such an alternative it is accepted that duress, and not free will, is the basis of the marriage. We would argue that this test, which is represented as universal, is in fact highly exclusive. In constructing a benchmark for determining the strength of any duress, the gender-neutral, race-less facade of the legal language in which the notion of ‘reasonable alternative’ is couched hides behind it a presumed norm of the legal subject as a white male, and thereby takes little account of how categories of women, and of South Asian women in particular, may experience and respond to duress in the matter of marriage.

Many feminists have critiqued presumptions of choice and consent exercised by a formal, neutral, legal subject (Davies 2002; Naffine 2003; Richardson 2004; Hunter and Cowan 2007), pointing to the fact that in early law relating to the regulation of marital authority and property the husband’s personhood subsumed that of the married woman, and drawing attention to the continuing patriarchal assumptions underlying this gendered construction of legal personhood. While it is not our intention to rehearse their arguments, the question that informs this critique of the current legal discourse on forced marriage is one that has engaged much of recent feminist theorising on law: how does one go about constructing a legal discourse that does not view women as an aberration from the norm, but recognises the bodies and lives of women (Richardson 2004)—and specifically, for our purposes, those of South Asian women?

When courts in the UK apply the test to see whether there is a reasonable alternative to a marriage, their assessment is implicitly based on what would be reasonable for a white male living in the UK. It is with these androcentric assumptions in mind that the courts have been willing to accept that physical threats from a personal source or from unattributable sources such as political circumstances and emotional pressure from a personal source are coercive, but they have been far more reluctant to accept the influence of factors that may be specific to some communities, and particularly onerous on women (such as fear of ostracism from the community or notions of shame). Thus, although ‘reasonable alternative’ is posited as an empirical test of coercion, it is far from being so, because it presupposes certain gendered, culturalist ways of imagining the self as neutral, universal and pre-social. In the case of women facing forced marriage, this means that the gendered, socio-cultural norms and expectations that may overwhelm their will are disregarded, presumably because they would not similarly coerce an atomised self easily differentiated from similarly disaggregated others.

Feminist legal scholars have long noted that the atomistic, unencumbered subject deemed to be in full command of self, personhood, labour and property bears little resemblance to women’s experiences as socially situated and heterogeneous subjects constrained by gendered norms (Walker 1999). The construction of personhood among women can be said to be constituted with reference to complex relations of dependency shaped by political and social hierarchies (Meyers 1989; Mackenzie and Stoljar 2000). Within communities where arranged marriage is practiced, determining whom a woman can and cannot marry in fact serves to delineate the boundaries between communities and castes (Wilson 2006), a conflation of women’s bodies with the boundaries of the nation and community that is not unique to (im)migrant cultures (Yuval-Davis 1997). Refuges for South Asian women in the UK have long dealt with the ostracism that many women face as a consequence of disclosing domestic abuse and choosing to leave the abusive relationship. Being exiled from social networks can involve more than the emotional loss of significant others: it may mean the loss of one’s own identity, what Reitman (2005) terms the “sociopsychological costs of exit”, the loss of a sense of belonging that defines the self (Weinstock 2005). In this context, a refusal to consent to a marriage that will result in a potential loss of family, and thus community, is surely perceived as inherently coercive by many women. Furthermore, for many women of South Asian origin, as indeed for any woman of colour, the option of exit from their family and (often) thereby, their community is to ‘escape’ into a racist society, one potentially hostile to (im)migrants, particularly since 9/11.

We can see the coercive power of socio-cultural norms and expectations in the case of Singh v Singh,Footnote 20 where the petitioner was a 17-year old girl who consented to marriage out of respect for her parents and religion, but did not wish to go through the civil ceremony. Her appeal was dismissed because of the absence of any explicit threat. Even today, if an individual acts out of deference to cultural, gendered notions such as respect, it would be hard for any court to see this as coercion or duress. However, this notion of respect is not a spontaneous feeling that arises within the subject; it is a gendered, socially and culturally constructed code of behaviour enforced through a range of notions, such as honour (izzat) and shame (sharam).Footnote 21 Women’s agency is exercised in the context of a number of coercive forces, including specific forms of socialisation (for example, being taught to value success in relationships). Not all South Asian women in the UK confront these norms, and not all of those who do, do so in the same way. However, research among South Asian women in the UK indicates that among particular communities from South Asia, women are assigned the role of the bearers of the family honour; there are sanctions that underpin this patriarchal code, and the women’s behaviour is policed so that they do not breach it (Gill 2004; Wilson 2006). Socially and culturally constructed notions like shame may well make a woman feel that she has no choice but to consent to a marriage, to avoid stigmatising her family and to preserve her own sense of self (Gangoli et al. 2006, p. 11), constraints which also prevent many South Asian women from articulating their experiences of violence and seeking help (Gill 2004). Women’s class, education and employment status, along with generational differences in outlook, differences in migration routes into the UK, regional variations in the subcontinent, the woman’s position within community networks in the UK, the perception and reality of racism, and access to appropriate support and services, all intersect to shape women’s agency in the face of these constraints. These norms do not go unchallenged by women living within their communities and those who have resisted these norms by taking the route of exit; conversely, even women who have left their family home continue to define themselves with reference to these norms in contradictory and conflicting ways, while simultaneously challenging them through their actions as they seek to rebuild their lives.

Rather than characterise such contexts as coercive, Wertheimer (1987) refers to the notion of ‘constrained volition’, where a person faces undesirable alternatives, yet in some sense is capable of making a ‘rational choice’. The courts in the UK have continued to apply a similar standard of individual rationality and volition, which presumes that the individual in question is a pre-social, ahistorical, self-constituting subject who does not belong to an identity-conferring community, nor values relational aspects of personhood. The patriarchal structures within which all women find themselves, and the culturally and historically specific forms in which these structures manifest themselves for particular groups of (im)migrant women from South Asia, continues to be elided from legal debates on forced marriage in the UK.

There remains a need to reconceptualise the legal subject such that South Asian women gain the status of legal persons. However, in doing so, there is also a danger of reiterating specific norms of femininities as all-encompassing and unchallenged and falling into the trap of culturalist essentialism. Carline (2005) examines the case of Zoora Shah, who was convicted of murder for poisoning Mohammad Azam after years of abuse, to argue that Zoora’s failure to conform to a number of essentialist gender and racial scripts led to the court labelling her as an ‘unusual woman’ and representing her as an unintelligible gender. This in turn prevented her from telling the ‘truth’ about her situation and so obtaining justice. It has been argued that when a cultural defence is invoked in courts, both by male perpetrators of culturally specific forms of violence against women or by women seeking justice following the murder of a perpetrator of domestic violence, pre-existing beliefs about culture, gendered violence and patriarchy have also been reified, and only occasionally dismantled (Volpp 1994; Phillips 2003). Such strategic uses of the law are also evident in the case of forced marriage, when solicitors acting on behalf of women who seek the annulment of their marriage often deploy similar arguments to present to the jury the circumstances in which a woman may have succumbed to pressure to marry (Newham Asian Women’s Project 2008). While these tactics might secure the rights of some women in the short term, there are inherent problems with the re-constitution of the victim-subject through these scripts.

This socio-cultural context has always been readily invoked by mainstream communities, though presented as a de-culturalised norm, when they draw upon this understanding to explain the meanings of their behaviour, as when men have long been able to argue that ‘their’ women’s sexual transgression or decision to end the relationship is enough to provoke a ‘reasonable man’ into retaliatory violence (Volpp 1994; Phillips 2003). There can be no inherent argument against extending the deployment of socio-cultural contexts to persons from BME communities, but when culture is invoked in an essentialist way, as something static, unchallenged and isolated from the politics and history which shape it, women from these communities are constructed as always already victim-subjects who are completely determined by their culture, or as ‘unusual’ women who cannot thereby invoke culture to contextualise their actions or explain its meanings. Such constructions of women as victim-subjects invite protectionist responses from the state, such as the recent changes in immigration law in the UK to increase the age of marriage to 21 where one of the partners is from overseas, a policy that has implications far beyond the protection of women. It is because of the historic inability of the law to capture the multiple ways in which women exercise agency within (and despite) constraints, the preoccupation of the law with the victim-subject, the protectionist responses that often erode women’s rights and the reinforcement of gendered stereotypes within legal discourse that feminists have argued that the law remains an essential but insufficient route to gender justice (Menon 2004; Kapur 2005). The way forward, we argue, must be to conceptualise the legal subject in ways that incorporate South Asian women’s experiences, but which are not constrained by essentialist assumptions about these attributes and so are able to respect women’s volition within these structural constraints.

Arranged Marriage, Forced Marriage and Somewhere in Between: Women’s Agency in Oppressive Contexts

What we are calling for, then, is a new discourse of personal freedom and constraint to encapsulate the experiences of particular groups of South Asian women, and to reflect upon the diverse forces that bear down upon women in relation to their marriage choices. In so doing, the hope is that we can begin to abandon the binary conception of consent and coercion. This is a difficult argument to make in the current context, where populist explanations of forced marriage take place within essentialist discourses that stigmatise Muslim communities and feed into racist responses. The dominant media discourse on forced marriage comes to what on the surface may seem like a similar conflation of arranged and forced marriage, and views them both through a culturalist lens which overwhelmingly portrays such women as passive victims, rather than unpacking the context of gendered inequalities within which women in any community cannot meaningfully exercise consent. According to this logic, some South Asian women are lauded as heroines for having resisted these patriarchal structures within their communities, having managed to escape their control and successfully reconfigured themselves as completely autonomous agents existing outside their communities. This can be seen in the media treatment of Jasvinder Sanghera, who in 2007 wrote a book, Shame, based on her experience of having survived a forced marriage. Others are portrayed as passive victim-subjects awaiting rescue by the British state (Harding 2000).

Perspectives like these also emerge in the scholarship on agency. Women’s apparent consent to marriage in the face of coercive socio-cultural and structural forces has often been broadly interpreted as acquiescence to patriarchal authority, whereas agency is equated with women’s declared resistance (Goddard 2000, p. 3), often through the strategy of exit. Others have argued that the very idea of choice in the context of more overarching systems and networks of power and domination is problematic (Wilson 2006). Research suggests that South Asian Women actively negotiate different aspects of their environments, and make deliberate strategic choices within structural constraints, in order to manoeuvre the grey areas between (relative) coercion and consent when it comes to marriage decisions (Samad and Eade 2002; Gangoli et al. 2006).

We earlier discussed the context of racist immigration policies of the UK government, within which women of South Asian origin may come under pressure to facilitate the entry of a relative to the UK through marriage. These very policies also create opportunities for women who attempt to deploy them strategically to escape living within a forced marriage. Unable to jeopardise their relationships with their families by rejecting the marriage or walking out, on return to the UK these ‘reluctant sponsors’ shop their husbands to the Home Office in the hope that this would be sufficient to prevent their husbands from obtaining a visa to come and join them (Wilson 2007, p. 33). Faced with pressure to marry, South Asian women’s strategies have also included negotiating a later marriage, prolonging a period of study, setting a limited ‘veto’ right over prospective partners selected by parents, and asserting a right to ‘choose’ a partner within certain acceptable categories of caste and class (Bradby 1999; Wilson 2006; Gangoli et al. 2006). The complexity of women’s actions in oppressive contexts can be understood by recognising the role of structural inequalities, while retaining a strong sense of respect for women’s agency in acting within the constraints and possibilities presented by their contexts (Wendell 1990; Chetkovich 2004).

How far these negotiations are actually a sign of the transformation of marriage patterns and a loosening of patriarchal controls over women’s sexuality, thus enhancing women’s capacity to be free from coercive marriages, has been a matter of debate in much of the literature on this subject. Wilson (2006, p. 18) frames this process as an acceptable compromise from the perspective of the family, because it enables families to retain ultimate control over their daughters’ sexuality; Wilson therefore sees it as far less assertive of women’s agency than it first appears. Nonetheless, in the case of marriage proposals, research indicates that South Asian women assess the emotional, financial, physical and cultural cost of saying ‘yes’ or ‘no’ in a situation of inequality, and do their best within these structural constraints, rather than blindly accept them or be able to reject them in their entirety (Samad and Eade 2002; Bredal 2005). Such accounts recognise women’s roles as cultural negotiators who may indeed accept and derive their identities from aspects of their culture, community and family life, while simultaneously seeking to actively negotiate and rework these norms rather than walk away from them. Women’s individual survival strategies, unmarked as they are, are not devoid of struggle and resistance to patriarchal family structures, and constitute the changes in what comes to be defined as traditional in particular contexts.

Furthermore, not all South Asian women seek to exercise their agency in order to enhance their autonomy. Within the context of patriarchy, women will often act to uphold gendered norms, such as those of beauty, or adopt disciplinary bodily technologies like elective cosmetic surgery (Frank 2006). Most feminist celebrations of women’s agency are in service to the politics of emancipation, and such accounts interpret women’s lack of autonomous impulses as acquiescence to patriarchal power structures, and see women’s desires as informed by oppressive norms of femininity (Morgan 1991; Wolf 1991). One argument that does not conceptualise agency as oppositional has been voiced in Mahmood’s (2005) account of the women’s piety movement in the mosques of Cairo, which uncouples agency from liberatory politics. In the context of discourses which presume that BME women must pit themselves against their communities in order to recover their agency, research on the marriage practices of BME women have also uncovered instances where young women have articulated their desire to uphold certain traditional norms by actively pursuing arranged marriages out of a positive need to assert their sense of belonging within a community under siege post 9/11 (Bredal 2005).

Such non-oppositional demonstrations of agency have implications for the exit-centred state response to the problem of forced marriage. The right to exit is, at heart, an extension of the idea of reasonable alternative—it suggests that women can only express their agency by fleeing from their own cultures, just as the normative white male might be expected to do. However, women are socially constrained from taking such a course of action, which is, to some of them, not a legitimate course of action at all. Agencies need to give more support to those women who wish to express their subjectivity within the framework of the communities of which they fundamentally perceive themselves to be a part.

Conclusion

We have argued that awareness of the socio-historical, political and cultural context of women’s lives, and their location at the intersection of several vectors of inequality, is crucial to understanding the relationship between consent and coercion in matters of marriage. Given this, the definitions of arranged and forced marriage that are currently being deployed need to be re-examined. Women’s experiences in matters of marriage choice form a continuum of attitudes, with consent and coercion standing at two opposing ends of this continuum. Although making a clear distinction between arranged and forced marriage seems relatively simple and straightforward, in between these two poles they shade into one another through varying degrees of social and cultural expectation: exercise of control, persuasion, pressure, threat and force in the context of gendered inequalities which create the potential for exploitation. Within such constraints and despite the extent of their subjugation, women exercise their agency in complex and often contradictory ways as they assess the options that are open to them, weigh the costs and benefits of their actions, and seek to balance their often competing needs with the expectations that weigh upon them.

Recognising that subordination and gendered power imbalances create conditions within which women’s consent is constrained in minority and majority communities involves moving away from a dichotomous understanding of culture as a marker of absolute difference. However, this does not erase the task of explicating the specific forms which coercion acquires in particular contexts. We have tried to do this by exploring the meanings and manifestations of coercion in matters of marriage for categories of diasporic South Asian women in the UK, and by looking at women’s decision-making processes in the context of the structural factors that constrain them and create specific opportunities for action. This article argues that it is possible to focus on one particular grouping, and indeed that it is necessary to do so where a specificity needs to be unpacked in order to understand and respond to particular forms of violence against women without minimising the connections and commonalities in women’s experiences of violence and their resistance to it. Specialist domestic violence services have long been able to challenge the patriarchal structures within their communities that constitute individual acts of abuse, without resorting to cultural deficit explanations. The response of the UK government to our calls for more attention to be given to the problem of violence within our communities has been heeded, but also derailed by current government attention to forms of domestic violence such as forced marriage and so-called ‘honour crimes’.

In the case of forced marriage, a binary understanding of consent and coercion in marriage would primarily generate a legal response aimed at preventing such marriages through injunctions and/or criminal proceedings, thereby enabling women to leave such marriages. However, this article has argued that such a law is unable to recognise the range of constraints that women face in matters of marriage, and that BME women are particularly marginalised by the legal discourse in the UK. Women’s groups have long argued that not all women seek to leave abusive families, and that an even smaller number want to initiate criminal or civil proceedings against family members in cases of forced marriage. Black feminists have thereby articulated the need for a diverse range of responses to the problem of violence against women, including funding specialist outreach services (which can create spaces where women can meet and form discussion groups), supporting education and awareness work in schools and in the community, securing welfare services and childcare structures and continuing the provision of refuge spaces.

The current state response is geared towards legal intervention, however, and the recent legislation on forced marriage has not been accompanied by any additional funds. Paradoxically, specialist domestic violence services face mergers and closures in the current shift in state policy towards ‘community cohesion’, an initiative which remains unquestioned and unquestionable in the current anti-immigration climate, as a consequence of which recent concerns about forms of violence within BME communities have gained such currency. BME women’s experiences of coercion in matters of marriage are not coherent, explicit, identifiable and distinct from their experiences within other structures of inequality. Only by addressing these structures can there be a lasting solution to the problem of violence against women.