Lone motherhoods in context

This paper investigates alternatives to studying the lives of lone mothers through the homogenizing category ‘lone mother’. This can be achieved by an analysis which combines the following three critical perspectives on the category: first, understanding the impact that contexts (both material and cultural) have on how lone motherhood is defined and experienced; second, acknowledging how the category ‘lone mother’ is but one aspect of a woman’s life in which also other categories such as gender, class and ethnicity intersect and interact; and third, examining the impact that the category ‘lone mother’ has on the lives of women. These three routes of analysis in combination allow an appreciation of the constraints placed by social inequalities on the lives of lone mothers, while not reducing these multi-dimensional lives to the one-dimensional category ‘lone mother’, and could form the basis of less stigmatizing and more effective policies. Real Life Methods Working Papers: Lone motherhoods in context July 2006 3 Introduction This paper examines alternatives to studying the lives of lone mothers through a homogenising, totalising and (at times) oppressive category, offering three avenues forward: first, focusing on the interaction between material and cultural contexts and the everyday lives of lone mothers; second, examining the multi-dimensional character of lone mothers’ lives; and third, studying how the category ‘lone mother’ comes alive in the everyday lives of women. These three elements in combination represent a research approach that appreciates the constraints that social inequalities bring to the lives of lone mothers, while at the same time not reducing the multidimensional everyday lives of women who parent on their own to the one-dimensional category ‘lone mother’. Such research could form the basis of less stigmatising and more effective policies that could attempt to grasp the complexity of lone mothers’ lives. Uniform policy that treats lone mothers as a homogeneous category and thus ignores the variety of experience among them risks being inefficient and oppressive (Duncan and Edwards, 1999). The present Government’s tactics towards lone mothers differ from the discursive battering offered by the previous Conservative Government (e.g., their ‘Back to basics’ campaign). With New Labour, there has been a shift away from a blaming and stigmatising discourse to an emphasis on finding solutions to the problems of child poverty and social exclusion. More prominence is given to providing support for lone mothers to enable them to become fully-fledged ‘citizens’, that is, to enter the labour market and become financially independent (e.g., the New Deal for Lone Parents). The issue of ‘social inclusion/exclusion’ is also given much attention (e.g., the Sure Start programme). But behind this benign language of ‘inclusion’ and ‘support’ lies a moral agenda and some harsh tactics for getting parents in line with ‘proper’ (that is, middle class) parenting (Gillies, 2005a; Gewirtz, 2001). In addition, New Labour’s policies are based on a particular view of the relationship between individual and society, one that lies close to some of the tenets of the individualization thesis. Social exclusion, for example, is in the policy context partly defined as an individual problem, and the proposed solution is to ‘train’ them on social and other skills. By emphasising individual agency and responsibility as a route out of social exclusion, the government is failing to acknowledge and deal with structural inequalities in society that act as barriers to inclusion (Gillies, 2005a: 86; Kidger, 2004). New Labour’s family policies are to a large extent based on the assumption that everybody in theory could have access to the same middle class privileges. Thus the government has been criticised for not paying enough attention to larger structural forces but rather micro-managing problems by focusing on area-based politics and on particular groups of people, which means that problems in an area or among a group of people are not connected with structural inequalities. As a result, social inequalities are rendered individual problems ‘faced by particular groups of women, such as lone mothers’ (Lister, 2001: 435). Lister argues that social exclusion cannot be eradicated simply by focusing on socially excluded individuals; what is required is policy that aims to eradicate social structures that create inequalities based on, for example, gender, class and ethnicity. The kind of research that would be required to support this kind of social policy would have to therefore look at both social processes and structures, and Real Life Methods Working Papers: Lone motherhoods in context July 2006 4 contextualised lives. The way forward in the area of lone motherhood that is suggested in this paper is to develop research that focuses on the following elements: context (e.g., time, place, and cultural and material context); the multidimensional nature of lives that cannot be reduced to one category (borrowing from theories of intersectionality); and the everyday experiences of lone mothers (e.g., what the category ‘lone motherhood’ looks like from the inside and issues of identity and agency). Ideally these three perspectives would be combined in any particular study, if not as objects of analysis then at least as theoretical underpinnings. A critical look at the category ‘lone mother’ A focal point that unites the three elements named above (context, multidimensionality, and everyday experience) is an understanding of lone motherhood as a socially constructed category. I argue that much could be gained by a critical look at the category ‘lone motherhood’ itself: how it is constructed and employed in lay, policy and academic contexts, as well as the impact this has on the lives of lone mothers. While some scholars have problematised the meanings accorded to lone motherhood by for example right-wing politicians in debates over family values (e.g., McIntosh, 1996; Phoenix, 1996; Roseneil & Mann, 1996), few have questioned the actual existence of the category, with the notable exception of Duncan & Edwards (1999). Creating categories is what we humans do in order to understand the complex world around us. Through language (words, concepts, theories) we order, make sense of, and provide labels for things, people and experiences. The act of placing phenomena into categories is, however, never a neutral exercise; categories are created in relation to underlying social relations of power (Taylor, 1998) and thus categorization can serve social purposes such as (re)creating social inequalities and boundaries between different groups of people, between ‘us’ and ‘them’. Relations of domination and subordination are reflected in dominant discourses which are employed to position groups of people through categories. Language is in other words never ‘just words’ but rather, the names and labels we give things have real consequences in the lives of individuals. To paraphrase W.I. Thomas (1967): if individuals define a category as real, it is real in its consequences. Take the category ‘lone motherhood’ as an example. This concept has historically come with negative connotations that are related to ideologies around gender, sexuality and family. Lone mothers have been thought to be breaking against social norms by being unmarried or divorced, that is, without a man. Thus the concept evokes a boundary between ‘proper’ women who are married or with a male partner and problematic ones whose sexuality is not under male control. There is furthermore a differentiation between ‘normal’ nuclear families and other, ‘inadequate’ families (VanEvery, 1999). As a consequence, lone-mother families are perceived as a threat to the stability of society because they are bringing up their children in such an ‘inadequate’ family environment. Lone mothers are also seen as a social problem not only because of the stereotypical view that they are dependent on benefits and but also because they are thought to be perpetuating a culture of dependency by generating a social underclass (Duncan & Edwards, 1997; Roseneil & Mann, 1996). There are, however, also counter-discourses that present lone mothers as strong, Real Life Methods Working Papers: Lone motherhoods in context July 2006 5 independent and autonomous women, but these discourses are in the minority and do not have the same impact or authority as the social problem and social threat discourses. Lone motherhood is a widely used category, and it is often employed in an oppressive and even derogatory manner, such as the stereotypes outlined above. One of the main sites of categorisation with great impact on the lives of lone mothers is social policy (cf. Anthias, 2005; Duncan & Edwards, 1999; May, 2001). Lone mothers as a group are seen as suitable for categorisation and intervention because they are regarded as ‘dependent’ and ‘alone’, as a potentially threatening ‘other’. Yet this rhetoric of dependency does not take into consideration that lone mothers are so much more than just ‘lone mothers’. In other words, ‘lone motherhood’ has become a totalising identity category that can be used in social policy discourses as a way of fixing in place or disciplining certain groups of women (Taylor, 1998). The identity category ‘lone mother’, when used in this way, attributes to lone mothers essentialised (im)moral characteristics that in turn are related to an evaluation of their ‘genuine’ needs and the legitimacy of their claims to welfare (Taylor, 1998). Thus, apart from creating boundaries between differently situated individuals, the act of categorizing – such as categorizing mothers living without a (male) partner as ‘lone mothers’ – tends to homogenize groups, that is, to create the illusion that members of a category share more in common than they in fact do (cf., Anthias, 1998: 564). Lone mothers can be and are viewed through this homogenising or totalizing lens irrespective of whether or not their circumstances fit the stereotype. Nevertheless, this homogeneity is merely a discursive illusion that hides the variety of social positions occupied by those who are ascribed the categorical identity. Lone mothers are women of different class and ethnic backgrounds with varying understandings of what ‘good’ motherhood entails; some working some not; some bringing up their children alone while others doing so in co-operation with the father, grandparents or other family members; some with extensive social networks, others isolated. Just as there is no ‘standard British family’ (Duncan & Smith, 2002: 473) there is also no ‘standard lone mother family’ and how ‘lone motherhood’ is constructed and experienced varies. Duncan & Edwards (1999) have argued that because ‘lone mother’ is a taxonomic category that is not necessarily determinant of behaviour (cf. May, 2004b), it is possible to question the validity of using the category in social policy. It is therefore vital to understand that the welfare needs of certain groups cannot always be understood with the help of traditional categories such as ‘lone mother’ and social policy must be based on an understanding of the range of social positions that individuals inhabit in order to be effective (Taylor, 1998). Many might think that the solution to this problem of whether or not categories are useful, and to what extent they can even be harmful in the lives of individuals, is to draw on current theses on individualisation and reflexive modernisation, which propose that we live in an age of expanding freedom and ‘choice biographies’ (*Beck; Giddens). In other words, individuals are believed to be reflexively choosing their lifestyles and to have the scope and ability to place themselves in categories that are relevant to them. According to this view for example lone mothers would not have to passively accept being categorised as stereotypical ‘lone mothers’ but would be able to create their own meanings or ally themselves with different categories. I would however argue that this is a step too far. These theories have been criticized for downplaying the continued Real Life Methods Working Papers: Lone motherhoods in context July 2006 6 importance of structural forces in shaping the lives of individuals, as well as for overemphasizing the role of individual freedom and choice. ‘Reflexivity’ is not universal but rather can only be exercised by individuals who possess the requisite knowledge and material resources – in other words, as a result of social inequalities, not everyone is equally situated to become reflexive (Adkins, 2003; McNay, 1999; Skeggs, 2004b). The ability to be an ‘individual’ is the result of privilege (Skeggs, 1997: 163), something the New Labour rhetoric does not take into account. It has also been pointed out that the ‘reflexive self’, although a gender-neutral concept, is in fact a male self, and that in a world of gender inequalities women may not have the same access to autonomy and reflexive agency (Hoggett, 2001). In addition to these debates over to what extent individuals have the capacity to be reflexive there have been calls for social science theorizing to redirect its focus back to the constraints that create inequality (Adkins, 2003; McNay, 1999; Brannen & Nielsen, 2005). An increase in reflexivity does not necessarily entail the freedom to question and refashion all aspects of life as there continue to exist supra-individual structures and elements of subjectivity that remain difficult to avoid and are thus less conducive to self-fashioning (Adkins, 2003; Alexander, 1996; McNay, 1999). There are limits to what we can be – we cannot fail to make ourselves in particular ways (Skeggs, 1997: 162). Individuals are placed into or find themselves in a variety of externally defined categories such as ‘woman’ or ‘working class’, and these categories have real effects in their lives (Skeggs, 2004a). However, the issue of reflexivity and agency is not as simple as a question of either/or, as Hoggett (2001) notes. He argues for an understanding of human agency as multifaceted and at times contradictory; in some areas of life we may be reflexive agents, whereas in others we are non-reflexive objects: If [ . . . ] we construe the personality in terms of a number of selves each in relation to the other, then it is possible, for example, to see how one part of us may act while another resists or rebels. From this perspective we can be both responsible and innocent. (Hoggett, 2001: 53) Thus the best way of trying to capture simultaneous reflexivity and nonreflexivity (or agency and non-agency) is an approach that combines insights from postmodern theory while not losing sight of structural inequalities. Individual lives are in other words neither fully structurally determined nor freely chosen but are rather something in-between, with different groups having different access to the resources required by ‘agency’ and ‘reflexivity’ (Taylor, 1998). The construction of categories and identities happens in the complex interplay between material and discursive conditions. It is therefore not helpful to state that lone mothers can now choose to escape the stereotypes attached to the category ‘lone mother’, because to do so requires cultural, social and financial resources that not all lone mothers have access to, but also because many of them continue to be categorised in such a way by others in various meaningful contexts such as schools and social policy. I therefore argue that it is necessary to examine the interplay between reflexivity/agency and supraindividual structures. For example, it is necessary to focus on how the category ‘lone motherhood’ is employed and in which contexts; who gets placed in it, how and by whom; and the effects that either being categorized or ‘escaping’ categorization has on the everyday lives of differently situated mothers. Real Life Methods Working Papers: Lone motherhoods in context July 2006 7 The approach outlined in this paper avoids problematising lone motherhood yet focuses on the social inequalities that do impinge on the lives of many lone mothers. It is thus not an approach that simply says that the experience of lone motherhood would be problem-free for women if only we stopped stigmatising lone mothers. I argue for the need to continue focusing on the material realities under which lone-mother families live. Even if the stigma attached to lone motherhood were eradicated, certain material and structural realities would continue to persist for many lone mothers, such as the fact that there tends to be only one earner and carer in the household, and that this earner/carer is a woman of a specific class and ethnic background in a gendered, classed and raced society. This paper thus advocates a critical look at the category ‘lone mother’. This means unpacking the category by examining first, the variety of lone motherhoods over time and place and second, the multi-dimensionality of lone mothers lives (that is, that ‘lone mother’ is only one category among others impacting their lives). Third, it is necessary to also focus on the impact the category has on the lives of lone mothers. What all three elements have in common is that the analytical focus is shifted away from lone motherhood per se and onto issues of social differences and inequalities in general; the multidimensional nature of social reality; and the lived experiences of lone mothers. In the following, I examine each of these three facets more closely. Variety of ‘lone motherhoods’ – historically and geographically anchored definitions The first element under consideration concerns the contextual nature of ‘lone motherhood’, that is, the way in which the construction of lone motherhood depends on the when, where and who. In the section above I discussed how the individualization thesis tends to gloss over inequalities of, for example, gender, class and ethnicity (as discussed above), but it can also be criticized for being decontextualized because of a failure to focus on the specifics of where and when (e.g., Brannen & Nielsen, 2005). Individuals are all situated beings who act within specific contexts, taking advantage of opportunities and negotiating constraints, which are never universal but rather context-bound. Research that has focused on context has shown that definitions of lone motherhood vary over time and from place to place (e.g., Duncan & Edwards, 1999; May, 2003). For example, discourses on the national level tend to represent lone motherhood as a social problem or a social threat, whereas discourses on the local level can be more positive and supportive (Duncan & Edwards, 1999; cf. Holloway, 1998). It has also been documented how general attitudes towards divorce, ‘out-of-wedlock’ births and lone motherhood have become more accepting in the past 50 years (Kiernan, Land & Lewis, 1998). It is therefore vital to understand the context of individual action in order to make sense of this action. The importance of both material and cultural context to how lone motherhood is defined and experienced becomes visible when lone motherhood is examined over time. A study of life stories written by Finnish lone mothers of different generations investigated the contextual nature of these women’s diverse experiences (May, 2003; May, 2004b). It was clear that the different material Real Life Methods Working Papers: Lone motherhoods in context July 2006 8 circumstances and ideological climates had shaped the life story narrators’ experiences. The older lone mothers, who brought their children up in the harsh realities of post-World War II Finland in the 1940s and 1950s, focused on the material aspects of parenting. They recounted how they had managed to clothe, feed and educate their children – a feat not to be taken lightly. Their experiences of lone motherhood also occurred at a time when lone motherhood was highly stigmatised and the narrators recounted instances of harsh treatment they had received at the hands of family, neighbours and communities. The younger life story writers, who had been lone mothers from the 1980s onwards, had experienced very different circumstances. Bringing up their children in a post-industrialised country with a comprehensive welfare state, they could take a certain standard of living for granted. Thus providing their children with the basic necessities of life did not occupy space in their narratives, but rather they focused on the psychological aspects of parenting in reference to wide-spread discourses such as to what extent their children might be harmed by not growing up with a father figure in their everyday lives. These life stories reflected changing notions of child rearing, with increasing attention being paid to children’s emotional well-being. These younger narrators also did not recount specific instances of stigma, indicating a relaxing of attitudes towards lone motherhood, but they did refer to general notions that somehow one-parent families are ‘not as good as’ families with two parents, thus reflecting the fact that lone motherhood is not fully accepted in today’s Finnish society. Definitions and experiences of lone motherhood vary not only across time and space but also for differently situated lone mothers and upon a closer examination of categories their internal boundaries and hierarchies are revealed. Lone mothers, for example, are ranked according their age, class, ethnicity and route into lone motherhood. The stereotypical ‘problematic’ lone mother is not the white middle-class career woman, but the white working-class ‘feckless’ ‘Tracy’ or ‘Sharon’ living on an estate who becomes pregnant at 17, has several children by different fathers and is ‘dependent’ on benefits. Another lone mother who figures in both public and academic discourse is the Afro-Caribbean lone mother who is perceived to be ‘normal’ within her ethnic community. This category of lone mother has slightly more positive overtones as she is perceived as a strong woman embedded in a culture of matriarchy. But nevertheless this depiction does have patronising and racist overtones when discussion turns to the ‘culture of absent fathers’ among Afro-Caribbean families, and the perceived lack of a ‘healthy’ nuclear family norm (cf. Reynolds, 2001). I propose that the first element of any study of lone motherhood has to be a focus on the contextual nature of how lone motherhood is defined and how women experience lone motherhood. A woman’s experience of living in/with and being placed in the category ‘lone mother’ is embodied and context-bound, depending the historical time period and on her class and ethnic background, and locality (cf. Duncan & Edwards, 1999; Rowlingson & McKay, 2005; May, 2004a). In other words, lone motherhood is not a homogeneous category but rather encompasses a range of definitions and experiences. Thus where, when and who are crucial questions to pose in order to grasp the variety of experiences of lone mothers. The following element looks at this question of ‘who’ a lone mother is in more detail. Multidimensional lives in a field of cross-cutting structures The second route out of viewing all lone mothers as part of a homogeneous category is to understand lone motherhood not only as a complex category, but Real Life Methods Working Papers: Lone motherhoods in context July 2006 9 also as only one category among many others. This can be achieved with the help of the concept of ‘intersectionality’, which entails focusing on the multidimensional nature of social reality. Individuals do not simply inhabit a single category or social structure such as gender, ethnicity and class but rather their lives incorporate aspects of several categories or develop in-between or on the boundary of these categories, resulting in complexity, ambiguity and multidimensionality of identity (Soja, 1996). Theories of individualization and reflexive modernity argue that traditional categories such as class have lost their meaning and have become ‘zombie categories’ (Beck & Beck-Gernsheim, 2002). Yet empirical research shows that there continue to exist structurally shaped inequalities along for example class lines (Savage, 2000; Skeggs, 1997). Furthermore, these ‘old’ categories are still meaningful in the lives of lone mothers as well as other parents (Duncan & Edwards, 1999; Rowlingson & McKay, 2005; Gillies, 2005b). Thus it is not enough to merely state that we have entered a new age of choice and reflexivity which allows us to some extent to disregard ‘traditional’ structures. It remains important to focus on how even in these times of increased choice lone mothers from different socio-economic and ethnic backgrounds are viewed differently (as discussed in the previous section), but also how their everyday lives differ. Rowlingson and McKay’s (2005) study of lone motherhood and socio-economic disadvantage shows the variety that the category ‘lone mother’ often hides. They maintain that social class has a significant effect on the experiences of lone mothers. They found that lone mothers from a working-class background are more likely to become socially and economically disadvantaged than middleclass lone mothers are. However, Duncan (2005) argues that class is too simple and narrow a concept for explaining the differences in attitudes, behaviours and living conditions of different mothers, because even within one class attitudes to parenting differ according to locality. He argues for the development of a more complex understanding of class-based mothering, taking also, for example, biography, sexuality, social networks and local culture into consideration (Duncan, 2005; Duncan, 2006). The work by Duncan & Edwards (1999) has been important in unpacking the category ‘lone mother’ to reflect some of these cross-cutting social structures. They have demonstrated how lone mothers’ approaches to paid work are partly shaped by their class and ethnic backgrounds (see also Duncan, 2005 on how the meanings of these show regional variation). Lone mothers make their employment decisions based on their views of what ‘good’ motherhood entails, that is, whether a ‘good’ mother stays at home to look after her children or whether she enters the labour market in order to support her children. These views in turn are derived from national, local and neighbourhood cultures, norms and discourses. As a result, lone mothers from different class and ethnic backgrounds living in different localities do not employ the same criteria when making employment decisions. In other words, these women do not make decisions simply as lone mothers and consequently lone motherhood is not enough to understand these women’s lives. To begin to grasp the complexity of their experiences it is necessary to employ a multidimensional lens that encompasses several aspects of their lives. The best way to do so is to understand that lone mothers are not simply lone mothers, but that they are situated in a variety of different social structures that impact upon their lives in complex – at times mutually strengthening, at times contradictory – ways. Real Life Methods Working Papers: Lone motherhoods in context July 2006 10 Further unpacking the category ‘lone motherhood’ is a study by Levitas, Head and Finch (2006) examining the impact that marital status and gender have on living conditions. Their work raises the question to what extent lone motherhood can be used as an explanatory variable when discussing lone mothers’ poverty and social exclusion. They found that single women and lone mothers tend to be worse off than partnered mothers on a number of measures because they are single (sole earner/carer with less available money and time) and because they are women with an on average weaker earning power than men. Thus again, it is not enough to simply focus on lone motherhood when trying to explain for example the poverty that many lone mothers face, but rather a combination of factors or variables need to be taken into account. This is where theories of intersectionality, which represent social structures as complex and interacting with each other (e.g., Browne & Misra, 2003; Crenshaw, 1991), come in useful. The concept of social structure is not as simple as it at first appears: there is not one uniform structure in which individuals of say the same class of ethnic background live, but rather each individual inhabits a multitude of intersecting structures. Consider for example the likely differences between a White and an Afro-Caribbean lone mother, or a middle-class and a working-class lone mother. Although all of these women inhabit the same ‘lone mother’ category, their different backgrounds and contexts lead to a different constellation of experiences. Some of the structures they inhabit can potentially lead to discrimination (e.g., woman, working-class, Afro-Caribbean), while others can convey social privilege (e.g., white, middle-class). These various categories or social structures are not independent of each other: where they intersect, they interact with each other to create for example disadvantages that cannot be distilled or disaggregated to the component categories or structures. Because the lives of lone mothers are shaped by a variety of structural inequalities, any effort to improve the average standard of living for lone mothers has to be based on an understanding of social inequalities in general. This understanding appears to be lacking in the current policy climate. The Government’s activities in the area of parenting are couched in a language of ‘universal fact’ and as a result, the target of policy measures, that is, workingclass and ‘deprived’ families, remains veiled (Gillies, 2005a). Furthermore, policy defines ‘inadequate parenting’ as the root cause of a ‘cycle of deprivation’, thus ignoring material and social inequalities (Gillies, 2005a). A more constructive approach would involve viewing parenting as an embedded, situated process, amenable to change only through social and material circumstances [ . . . ]. From this perspective, policymakers would have to recognize and address the crucial significance of power relationships and inequality in sustaining ‘cycles of deprivation’. This would require an end to the longstanding persecution of disadvantaged families and a greater focus on the principles of social justice to counter privilege, advantage and the self-exclusion of the rich. (Gillies, 2005a: 87) An approach that considers intersecting structures or categories brings such inequalities and differences to the fore and enables addressing how different groups of lone mothers are impacted differently by for example ‘deprivation’. Real Life Methods Working Papers: Lone motherhoods in context July 2006 11 The importance of everyday experiences The category ‘lone mother’ is however not simply a ‘structural location’ in which lone mothers are placed from the outside (McNay, 2004) but rather comes alive in the daily lives of lone mothers. It is therefore important to study experience, which offers not only a window into the subjective but also into how social structures and processes operate on the individual level: It is precisely the everyday negotiations of the mundane that do matter, that are formative, that do count and [ . . . ] these mundane experiences are a product of systematic inequality. They are not free-floating emotional experiences. They are profoundly located in structural organization. They may not be authorized, often un-eventful (and rarely spectacular) but as this study shows they do matter and they are constitutive. This means that we should not abandon the study of experience. Rather we need to rework it to explore how subjects are produced and produce themselves through their different experiences, exploring how different processes produce experience, which ones matter, which are authorized and how interpretation is central to productions. Experience relates to theory, not only in terms of how it can be encapsulated, but by showing how the subjectivity and the experiences of the knowledge producers inform the knowledge productions. (Skeggs, 1997: 166) A focus on the experiences of lone mothers highlights agency as an important part of how lone mothers negotiate with the category ‘lone mother’. Any analysis of the ‘institutional forms by which positively valued subjectivities are sustained and negatively evaluated subjectivities attributed’ (Taylor, 1998) is best combined with a perspective ‘from the inside’, that is, an examination of women’s experiences of lone motherhood. Such a combined focus on structure and ‘agency’ is necessary because one does not make sense without the other. Categories are not simply imposed from the outside or from above – rather, individuals actively relate to these categories and make them meaningful in their lives (May, 2004a). In the case of negative identity categories, such as lone motherhood can be, individuals who are placed in them may struggle for redefinition and challenge the fixity of the boundaries set by the category (Taylor, 1998). In other words, identity categories are not deterministic. ‘Lone motherhood’, though often viewed as a relevant and self-explanatory category from the outside, does not necessarily appear as meaningful to lone mothers themselves. This becomes apparent when adopting an approach that does not take lone motherhood as the starting point or lens through which a woman’s life is examined, but rather, where it is the woman’s life as a gendered, classed and raced person that is used to understand her experiences of lone motherhood (May, 2004a). How lone mothers themselves negotiate the mainly negative category ‘lone motherhood’ and the meanings they accord it within their identity construction is a vital component in any attempt to understand the impact that the category ‘lone motherhood’ has on the lives of women. Identity is not fixed but is instead relational and contextual, that is, forged in relation to others and within particular social contexts (Somers, 1994). In relation to particular categories or structures that could be used to define a person, such as ‘lone motherhood’, this means that where a lone mother is (in geographical, social and ideological terms) will affect what ‘lone motherhood’ means to her Real Life Methods Working Papers: Lone motherhoods in context July 2006 12 and to what extent she will opt for a ‘lone mother’ identity or resist it if she feels the category is being ‘imposed’ on her. Furthermore, identity is not one thing to an individual, but changes from setting to setting, depending on whom they are interacting with. This view of identity is based on a non-unitary view of self: we have many facets to our selves, some of them contradictory (Hoggett, 2001; Bhavnani & Phoenix, 1994: 9).Thus, because the categorical identity ‘lone mother’ can never fully define a woman, some lone mothers use other aspects of their identities, such as underlining other social positions such as class, ‘good’ motherhood or respectability, to counterbalance any negative classification (May, 2004a) or as a strategy for avoiding classification as a ‘lone mother’ (cf. Skeggs, 1997). At this point it becomes useful to distinguish between a person’s identity as viewed from the outside and as experienced by themselves (cf. May, 2004a). Taylor (1998) distinguishes between categorical identity and ontological identity. Categorical identity results from the process of categorisation, of classifying individuals as the same, and thus as belonging to the same social category. As discussed above, such categories can become totalising, that is, to be seen as all the person is. Categorical identity can be either imposed or chosen – it is more likely to be chosen if it comes laden with mainly positive definitions. But even if a category is imposed from the outside, this does not mean that the individuals who find themselves placed in that category are powerless to resist it. Thus lone mothers who find that the category ‘lone motherhood’ entails a negative identification may try to distance themselves from this definition (May, 2004a; Bock, 2000; Duncan & Edwards, 1999). As already noted, categorical identities define the inhabitants of a category as the same. Admittedly inhabitants of a category will tend to share at least one characteristic – in the case of lone mothers the fact that they are the sole residential parent, though as studies have shown even this definition can be problematic – but categorical identity is not the totality of a person: we are all the same but also different. For example the category ‘lone mother’ includes women with a multitude of different experiences and identities. It is this difference based on our unique characteristics as individuals that ontological identity can grasp (Taylor, 1998). These two aspects of identity, the categorical and the ontological, are not mutually exclusive; we create a sense of unity out of an interplay between the two. This helps to explain the perhaps contradictory way in which some lone mothers identify with the category ‘lone mother’ yet at the same time distance themselves from it (May, 2004a; Bock, 2000). Skeggs (1997: 106) argues that the working-class women in her study did not feel similarly positioned by femininity as they did by class. Femininity was in other words not a part of them, or, in Taylor’s (1998) terms, it was not fundamental to their ontological identity. I have argued elsewhere (May, 2004a) that lone motherhood can function similarly; it can be a classification that is externally imposed or one that is adopted for tactical purposes, but not necessarily one used to construct one’s ontological identity. Yet lone motherhood may be a categorical identity that lone mothers cannot completely avoid in their identity construction. Some categories, such as ‘woman’ are so powerful that all those positioned within it construct their identity with ‘woman’ as one of its components (Taylor, 1998). Thus even a refusal to inhabit a category does not automatically mean that it can be abandoned (Skeggs, 1997: 166). Lone motherhood appears in many cases to operate in the same way (May, 2004a). Whatever their views on the category ‘lone mother’, women who are Real Life Methods Working Papers: Lone motherhoods in context July 2006 13 placed within it tend to involve a dialogue with the totalising aspects of the category in their identity construction (cf. Bock, 2000). This means that some categories, such as lone motherhood, appear to be so powerful that individuals cannot avoid them completely. Although lone mothers do show agency when engaging in a complex dialogue and negotiation with the category ‘lone motherhood’, thus resisting some of the negative definitions that come with it, it appears to be difficult for them to ‘escape’ this category altogether even if they hope to do so (Bock, 2000; May, 2004a). But I wish to emphasise the agency involved in living with a negative classification, which means that individuals cannot simply be portrayed as helpless victims of categorisation. I also challenge social scientists to think beyond categories and to be aware of the effects that the categories they help to (re)create have on the lives of individuals. In order to better grasp the ability that individuals have to resist negative categories or embrace positive ones, we need to understand how individuals are agents but also the limits that are set on this agency by material, cultural and social structures (McNay, 1999). There is no absolute ‘choice’ over subject positions but rather circumscribed access and movement between subject positions. Categories pre-exist our agency and frame our responses but we also contribute to their reproduction and reformulation (Skeggs, 1997: 94-95). Greener (2002) argues for a complex, multidimensional understanding of agency that takes into account structural restrictions: Agents, through structural, habitual or informational constraints, may be severely restricted in their range of possible actions, but be unaware of this as their actions have become so ingrained in practice that they are considered normal, and a lack of opportunity is not only the most likely outcome, but also the ‘common sense’ one. (Greener, 2002: 696) It is also important that policy is based on an understanding of the multidimensional nature of agency. To pretend that we have empowered actors allowed to behave in a reflexive way may serve to conceal the deep-seated structural power relationships that exist. (Greener, 2002: 697) Greener (2002: 703) further argues that the ability of social policy to be effective relies on the underlying assumptions that policy makers hold about agency. Accepting structural constraints is commonly seen as accepting dependence, but it remains important to understand that not all subjects can behave in a reflexive manner and to find the causes for this in order to combat inequalities.