Social Identity and Social Policy: Engagements with Postmodern Theory

Transcription

1 Jnl Soc. Pol., 27, 3, Printed in the United Kingdom Cambridge University Press Social Identity and Social Policy: Engagements with Postmodern Theory DAVID TAYLOR* ABSTRACT This article argues for the centrality of the concept of social identity in contemporary analyses of social policy. It attempts to transcend arguments for or against postmodernism and argues that debates about identity and difference, when combined with an analysis of social relations, need not undermine a focus on structural inequalities and should be at the heart of theoretical considerations in social policy. It suggests that the concept of social identity is still poorly understood in recent debates and goes on to outline a provisional theory which distinguishes between ontological and categorical identity. Such a discussion, it is argued, may help suggest ways in which the role of social policy in the process of identity formation may be better understood. INTRODUCTION: AGAINST A NEW BINARY Postmodern analyses of social policy have recently been described as a great leap backwards (Taylor-Gooby, 1994), a small step forwards (Penna and O Brien, 1996), and a contradiction in terms (Hillyard and Watson, 1996). Taylor-Gooby (1994, p. 385) believes they may be used as a smokescreen for recent trends to inequality, privatisation, retrenchment and the regulation of the poorest groups, whilst Ferge (1996, p. 3) has gone further, equating postmodernism with the neo-liberal values impregnated with the interests of national and international capital. For her postmodernism is part of the neo-liberal paradigm that undermines a commitment to welfare. It is argued here that it makes little sense to see modernism and postmodernism as either for or against welfare or to associate postmodernism with a neo-liberal paradigm. Instead, a consideration of recent developments in social theory, in particular the consideration of social * School of Policy Studies, Politics and Social Research, University of North London. The author would like to thank Norman Ginsburg for his helpful comments on an earlier draft of this article.

2 330 David Taylor identity as it arises in various forms of postmodernism, is seen as central to understanding contemporary issues in welfare. In order to understand its relevance, however, it is necessary to address the new false binary opposition of modernism and postmodermism and, as John Clarke (1996) has recently warned, to avoid a battle of travesties. The first travesty is to see modernism as either a unitary project embodying a commitment to welfare based upon universal principles dedicated to meeting human needs or, alternatively, the source of all falsehoods, empty promises, impossible dreams and mistaken epistemologies. The second is to associate postmodernism with either a simplistic Sunday supplement life-styles politics, celebrating consumer culture, or the ultimate critique of critiques which once-and-for-all puts an end to the mechanical clanking of 200 years of theoretical confusion. If modernism refers to the whole range of intellectual ideas elaborated within the context of modernity and postmodernism to all those critiques arising in late or postmodernity which seeks to question the paradigmatic assumptions and explanations of the former, then we must expect both to embrace the range of political positions found in over 200 years of changing social formations. Further, there is ample evidence that recent writings which explore issues associated with postmodernism in the area of social policy do not seek to sustain a neo-liberal antipathy to welfare. Squires (1990), for example, in a poststructuralist analysis of welfare, develops a trenchant critique of the anti-social aspects of neo-liberalism; Williams (1992, 1996) has sought to apply radical new models to the analysis of social policy which draw on postmodernism, feminism and political economy, and Penna and O Brien (1996) point to the enormous political diversity within the various post literatures. At the same time and the work of Bauman (1992) notwithstanding the collectivist, rationalist, welfare state project by no means typifies the history of modern capitalist societies, nor does it find complete support in the historical spectrum of modernist thought. Such support has certainly been characteristic of the social administration tradition in social policy, with its emphasis on state planning guided by informed expert opinion, but can hardly be said of other modernist critiques of welfare emanating from Marxism and feminism or, indeed, liberalism. Nor are those defending the importance of a focus on inequality unaware of the limitations of the discipline. Taylor-Gooby, the target of a recent postmodern critique by Hillyard and Watson (1996), described over fifteen years ago the limits of the empiricist and rationalist traditions in social policy, with their focus on expert knowledge, reflecting the perspective of the state (Taylor-Gooby, 1981). The great irony is that

3 Social Identity and Social Policy 331 setting-up unitary conceptions of modernism and postmodernism as binary opposites is itself a procedure which is thought to be at the heart of the postmodern deconstructionist critique, and does nothing to clarify that which is useful in recent theoretical debates. Rather than enter the debate on one side or the other, this article seeks to foreground one particular aspect of the encounter between modernism and postmodernism the theorisation of social identity. Unlike Taylor- Gooby (1981, 403) a consideration of this topic is not seen as a way of deflecting attention away from increased inequality. Indeed, understanding the sort of life-experiences and self-realisations made possible through welfare may demonstrate the way structural differences between individuals and groups are signified in concrete lives, and instead of individualising the social, as Ferge (1996) warns, this engagement with postmodernism may socialise the individual. A common concern of critics of the social administration tradition in social policy has always been its lack of attention to theory, and, if nothing else, the current postmodern turn in social policy forces us to focus on some of the implicit theoretical assumptions within the discipline about the subjects of welfare. If some of the insights of postmodern theory concerning the nature of identity and difference are combined with attention to social institutions and social relations (Taylor, 1996) there can be, in Nicholson and Seidman s (1995) words, a social postmodernism which is neither paradigmatically closed nor opposed to explanation. In recent years a number of writers have begun to apply aspects of postmodernism to the study of social policy, especially at the level of general theory. Hewitt (1994, 1996), for example, is concerned with the nature of change, philosophies of history, political agency and human need; Spicker (1996) and Thompson and Hogget (1996) elaborate the philosophical underpinnings of universalism and particularism in social policy; Penna and O Brien (1996) and Hillyard and Watson (1996) set out the broad range of postmodern theory suggesting a number of insights to move the discipline in new directions (p. 321), and Fitzpatrick (1996) attempts to establish a postmodern politics of welfare in defence of Citizen s Income. At the level of political economy, Burrows and Loader (1994), in their edited collection, draw together the implications of postfordism for the welfare state. An increasing literature which spans urban theory and social policy applies postmodern ideas to the analysis of spatial locations (cf. Watson and Gibson, 1995), and analyses drawing on postmodernism have been applied to managing and delivering welfare at the local level (Clarke, 1996; Clarke, Cochrane and McLaughlin, 1994). It is to a wider

4 332 David Taylor literature, however, that we need to look to find analyses of social identity and to draw out the implications for social policy. The work of authors such as Calhoun (1994), Nicholson and Seidman (1995), Benhabib (1990), Barrett and Phillips (1992), Brah (1996), Donald and Rattansi, (1992), Rattansi and Westwood (1994), Hall (1992), Law (1996) and Malik (1996) can all be said to have implications for understanding the positioning of gendered, racialised and sexualised subjects in relation to social policy, but it is only recently that writers within the discipline of social policy have come to deal with these issues. Fiona Williams (1992, 1994 and 1996) has gone furthest in making these ideas relevant to social policy. In particular, she has explored the concept of difference in relation to feminist and postmodern theory in an attempt to investigate the way in which welfare discourses shape the materiality of people s lives (1996, p. 68). Building on this approach this article seeks to investigate the notion of identity further and argues that it has a central role to play in the development of social policy IDENTITY AND SOCIAL POLICY Allusions to particular types of persons and categories of identity are frequently articulated in political debates about social policy and are equally at the heart of counter demands for recognition by user-groups and their advocates a politics of identity can frequently be found in campaigns for empowerment, for example. They are often present in the campaigns of new social movements (NSMs) and form the basis upon which interests are attributed to particular groups constructed as communities (for example, rights-based campaigns waged in the interest of the gay community ). Some writers have gone on to try to identify what they see as New Social Welfare Movements (Croft and Beresford, 1996) groups which assert particular welfare needs on the basis of empowered identities. Most of these uses of identity are framed within the context of debates about difference the deconstruction of the (false) universal subject of most modernist thought and the recognition of a range of diverse subjectivities within late modernity. Increasingly, however, such an approach to empowerment has had to come to terms with a central dilemma: in order to be recognised, specific groups need to call attention to their differences and focus on what is particular about their social identity. Yet to make claims for a (re)distribution of resources they need to deny their specificity by appealing to a wider criterion, such as justice or equality. This has been described by Fraser (1996, p. 69) as the redistribution recognition dilemma. Other writers have sought to go beyond the politics of identity (Nicholson and Seidman, 1995), pointing

5 Social Identity and Social Policy 333 out how some contemporary assertions of difference nevertheless fall back on an unsustainable essentialism. Such debates only go to show how concepts of identity and difference have to be used carefully in discussions of social policy. The broad question to be raised here is: what role do categories of social identity play within welfare? This is not simply an abstract or theoretical question. Appeals to particular identity categories function either as legitimating or disciplinary within discourses of entitlement and disentitlement. They may form the discursive backdrop for the inclusion and exclusion of particular groups and individuals from the social rights of citizenship and may constrain participation in the economic, political and cultural spheres. This approach to the analysis of social policy moves us away from a purely empirical focus on policy as provision to a focus on the social and the way in which individuals are positioned within it. Put another way, it might be seen as putting the social back into policy. Rather than concealing inequalities, as Taylor-Gooby warns, it can help us understand the way in which social groups and their members come to occupy a range of positions within the social relations of welfare and thus how their needs may be understood in ways not always satisfied by categories inherent in the empiricist traditions of social administration (or, indeed, many other modernist meta-narratives ). It draws us to a focus on how categories of identity are articulated in welfare discourses and how they are inscribed in the material practices and institutional forms of welfare. In this respect, social policy, whether as a set of benefits and services, a form of governance or a discursive system of power/knowledge, is deeply implicated in setting the ideological and material conditions for the realisation or foreclosure of particular identities. These issues are felt, then, at several levels: (i) general ideological constructions of legitimacy and illegitimacy attached to particular groups of actual or potential users typically inscribed with essentialist attributions of morality/immorality for example, single mothers ; (ii) specific identity constructions which act as conditions attached to entitlement for example, assumptions of dependency inscribed in various social relations of care which act as a pre-requisite to entitlement; (iii) attributions of interests to welfare users based upon abstract universal models of rational action, as in for example the attribution of consumer interest in expanded choice or the assumption of utilitarian calculations in decisions about benefit levels versus work incentives; (iv) institutional and material forms of provison which, through their physical structure, limit participation and assume particular types of users.

6 334 David Taylor It is not possible here to examine all the levels on which discourses of identity operate. What I hope to show, however, is that a clearer understanding of identity is required before we can map its use. Towards the end of the article I begin to analyse the way in which the first level of general ideological constructions of legitimacy/illegitimacy is played out in relation to one particular group. What follows is an attempt to examine the concepts of identity and difference more closely in order to understand the role of categories of social identity as they operate upon their incumbents and as they impact on policy making and welfare outcomes. THE POLITICS OF DIFFERENCE Discussions of identity in social policy so far have either been linked with or subsumed under discussions of difference. Williams (1996), distinguishes between political identities as the heart of anti-discriminatory practice; those which operate via the notions of choice (through consumerism); and those imposed from above to manage or manipulate the needs of different groups. Her typology is elaborated in the context of a discussion of the concept of difference and is aimed at clarifying the changing bases and ends of political agency. Are all differences the same? Are all differences to be celebrated? To what end is difference claimed?, she asks (p. 70). Fraser (1996), in elaborating what she calls a politics of recognition goes some way in answering these questions when she suggests such a politics would defend only those versions of the cultural politics of difference that can be coherently combined with the social politics of equality (p. 69). Difference, in these contexts is usually tied to a critique of universalism and normally takes one of two related forms: that universalism cannot deal with difference (Hilyard and Watson, 1996), or that universalism was never universal in the first place (Williams, 1992) employing socalled universal criteria which, nevertheless, discriminated against particular groups. These two views can lead to very different conclusions: the first that universalism in any form is a false project, the second, that there may yet be a universal process by which acknowledged differences can legitimately be judged (as in the case of Fraser, 1996). For either of these critiques the concept of difference is central. In this context, difference is seen in terms of the particular welfare needs of different social groups. Difference here is a difference from other social groups and their experiences, interests or needs a difference which may require particularistic forms of welfare rather than universalistic ones. As an example of the second version of difference attention has been focused on the analysis of New Social Movements as advocates of particular interests and

7 Social Identity and Social Policy 335 whether they can go beyond their specific concerns to generate a universal and inclusive politics (Hewitt, 1996). Either version of this approach to difference draws at some level on early postmodern writings such as those of Derrida (1978) and Lyotard (1984) which view the subject as multiple or fragmented, and seek to reveal the differences hidden within the unitary conception of the individual associated with many forms of modernism. This has led some writers to see the very concept of identity as part of the modernist grand narrative they wish to reject. Thus where identity has been discussed it is seen to flow from an understanding of difference in some ways as a denial of difference. For Derrida, difference is said to have been represented traditionally as other, as a negation of the self by the logocentric voice which speaks for a universal identity. Thus modernist identities have been deconstructed to reveal the negations of the various others upon which they depend for their spurious integrity. This deconstruction of the unitary subject into multiple or different selves has led to a view of identity as composed of, at best, fragments of difference. At its most extreme this results in a view of the disunified self that seems to defy classification, elusive and always in transition, engaged in Derrida s endless dance of difference. All that appears left is the assertion of particularity based upon the most transitory contextual aspect of experience which, if we are asked to believe in the fluidity of poststructuralist social relations, seems to cast the disidentified subject adrift in a sea of difference! Brah (1996), in a much more helpful approach to the question, distinguishes between four uses of the concept of difference: difference as experience; difference as social relations; difference as subjectivity; and difference as identity. Brah s overall concern is to understand the contested identities of ethnicity and gender, but what is particularly helpful here is her account of difference as identity. In the following formulation she allows us to reimpose some coherence on the fragmented and differentiated self: identity may be understood as the very process by which multiplicity, contradiction and instability of subjectivity is signified as having coherence, continuity, stability; as having a core a continually changing core but the sense of a core nonetheless that at any given moment is enunciated as the I. (pp , emphasis in original) What she shows is the importance of seeing identity as a sense of coherence that there is something at the core of each individual which unifies the fragmentation of experience. This does not imply natural or cultural essentialism but points, none the less, to the unity underlying multiplicity or difference.

8 336 David Taylor In this sense, then, difference as identity forms the basis of identification that which the subject asserts as I, has meaning for the subject and is recognisable within the context of systems of signification. However, identification does not take place in a neutral context. This form of difference is not passive. Criticising ethnicist discourses which seek to impose stereotypical notions of common cultural need upon heterogeneous groups, Brah points out, they often fail to address the relationship between difference and the social relations of power in which it may be inscribed (p. 100). Her point that we need to be attentive to the ways in which needs are socially constructed and represented in various discourses (ibid.) can be translated into the need to understand the way differences and identities with their attendant ideologies of legitimacy and entitlement and their imputation of needs are constructed in welfare discourses and how these relate to underlying social relations of power. Identity as difference, approached in this way, raises, then, the problem of coherence and its relation to social categories of difference and identification. What is at issue here is the categorisation of individuals into groups within social relations on the basis of common characteristics. These may be chosen or imposed, depending upon the position of such constructions within the social relations of power. It is not, however, Machiaevellian, but an inescapable aspect of social relations and processes of identification which are always signified in discourses of power/knowledge. For this categorisation to become the positive basis for social identity a process of identification must occur a process already well described in modernist theory in terms of class or gender identification. A difference here with many modernist accounts, however, is that this is not seen as a structurally determined process in which all that is necessary is the political realisation by unitary subjects of interests determined a priori by production relations, biology or gender. The construction and meaning of social categories is an intensely historical process yet one which, in the context of power, naturalises changing social relations and essentialises subjectivities. The historicity of this process, however, does not make it arbitrary or voluntaristic. It is clear that the history of capitalist social formations marks out key subjectivities as central in social relations, and that not all social categories operate at the same level of social relations, be they economic, political or cultural. Each may be sustained in particular ways through production relations, gender relations, political formations or cultural significations. For positive identification to occur, a sense of self identity must be realised through the various social categories inscribed in social relations of power. At the same time, of course, difference cate-

9 Social Identity and Social Policy 337 gories may be constructed from above to manage or manipulate subjects and to enforce negative identity ascriptions rather than positive identifications. So how is a sense of self sustained in this context? A common approach in contemporary theory is to see this issue as one of boundary maintenance between the positively evaluated universal self and the negated, particularistic other, or between the included, integrated citizen and the excluded alien. Postcolonial theory has sought to give voice to the other who speaks from the boundary (Homi Bhaba, 1992). Such a focus challenges the positioning of the negatively constructed other in social relations. Such a focus would lead to a consideration of the way in which negativised identities struggle for redefinition and challenge the fixity of inclusive/exclusive social boundaries. Whilst this approach may be necessary, it often neglects to address the institutional forms by which positively valued subjectivities are sustained and negatively evaluated subjectivities attributed. This is a process in which the institutional and discursive forms of social policy are deeply implicated. Hickman (1993), for example, has shown in a study of Roman Catholic voluntary aided schools in England, that separate religious schools were a means of ensuring that sections of the population, who differed in class composition and national identity, as well as religious denomination, were segregated from each other (p. 286) an example of the role of institutions in ethnic identification. Cornell (1996), also shifts attention away from boundary maintenance in ethnic identification suggesting that the content of ethnicities is sustained along three dimensions: shared interests, shared institutions, (and) shared culture (p. 265). This approach allows a consideration of the way in which the content of identities is given more fixity and less fluidity than in some postmodern accounts. Weak identifications are produced by communities of interest communities which arise out of circumstance alone: communities of interest are typically the consequences of a set of economic and political circumstances that place persons in common positions in the social order, provide for interaction among them, and provide as well a set of terms that highlight a particular boundary as the appropriate boundary by which to identify group situations. (p. 275) Italian Americans coming to the United States are given as an example of immigrants who came with very diverse identifications at the regional level but were encouraged through common economic and political circumstances to adopt a notion of common interests. Circumstance alone, however, is not sufficient to sustain a positive ethnic identification over time: it must be consolidated through institution building and elaborated

10 338 David Taylor through a shared culture as became the case with Italian Americans who built a more stable sense of identity than could have been achieved through common interest alone. Cornell is clear that this is always a changing and ongoing project, but it may produce relatively stable identities which are more or less able to resist boundary pressures. In this process, then, institutions play a central part they help maintain a stable conception of identity. Cornell is referring to voluntaristic institutions which ethnic communities establish themselves. Such considerations have another relevance to social policy, however, in that they may help us address the way in which social institutions inscribe cultural constructions of identity into the practices of welfare. There remains a problem, however, with Cornell s typology in that it appears to rest on a unitary notion of shared economic and political circumstances which generate ethnic interests. Do all members of ethnic groups share such circumstances are they not differentiated along gendered lines, for example? Iris Marion Young (1995) confronts a similar problem in her analysis of the concept of woman in feminist political theory. Acknowledging the heterogeneity of women in terms of class, colour, ethnicity, sexuality, etc., she nevertheless rejects the alternative voluntaristic redefinition of woman as simply no more than a marker of self-conscious political affiliation. On this view, feminism would have nothing to say for women who did not identify with it: The logical and political difficulties inherent in an attempt to conceptualise women as a single group with a common set of attributes and shared identity appear to be insurmountable. Yet if we cannot conceptualise women as a group, feminist politics appears to lose any meaning. Is there a way out of this dilemma? (p. 193) She suggests there is. She adapts the Sartrean notion of seriality in relation to the category woman. This approach retains a structural account of women as a group: the exclusions, oppressions and disadvantages that women often suffer can hardly be thought of at all without a structural conception of women as a collective social position (p. 193), but sees this located in women as a series a collectivity which has a passive relationship to its conditions of existence one which does not generate a unitary set of interests. The serial category of gender is, however, structured in the social body through enforced heterosexuality and the gendered division of labour, thus social relations and institutions set a context for political agency by the collective position occupied by women : Conceptualising gender as seriality does not claim to identify specific attributes that all women have. There is a unity to the series of women, but it is a passive unity, one that does not arise from individuals called women but rather positions them through the material organisation of social relations. (p. 208)

11 It is not that we necessarily need to adopt Young s use of seriality but that her account illustrates the general issue of how to understand the way in which we can theorise the nature of social categories as collectivities yet not assume that these automatically generate a set of shared interests which are seen as the basis of identity. Such a view allows for the assertion of difference without the abandonment of political agency or its reduction to a unitary set of interests. We can return to the work of Williams (1996) in search for the basis of agency at the heart of difference. She approaches the issue of difference by breaking up the analytical categories of feminism into, identity, subjectivity, subject position and political agency (p. 68) and considers the way these relate to welfare discourses and how, in their turn, these discourses shape the materiality of people s lives (ibid.). In so doing she distinguishes between diversity, difference and division: By diversity I mean difference claimed upon a shared collective experience which is specific and not necessarily associated with a subordinated or unequal subject position Difference denotes a situation where a shared collective experience/identity forms the basis of resistance against the positioning of that identity as subordinate. By division I mean the translation of the expression of a shared experience into a form of domination. (p. 70) Particularly useful is her analysis of the way difference may become division i.e., when the assertion of difference by the dominated turns into the consolidation of divisions and domination over other groups subsumed within the initial subordinate group. An example she uses is the women s movement in Britain in the 1970s, where the assertion of common interests amongst women consolidated divisions between different groups of women. Of more interest here, however, is the assumption that difference as collective experience is equal to identity. This presents a problem which the author is well aware of: does the process of asserting a common identity as one which is forged in its specific difference mean that other facets of a group s/person s identity are obscured and that such identities become frozen into an essentialist category of difference? (p. 71) This is particularly well illustrated in the Social Identity and Social Policy 339 tendency of equal opportunities policies to organise around discrete categories of oppression to freeze differences which, within the context of financial constraints, has slipped all too easily into competing claims of need. (p. 76) A result of this fixing of difference is to force members of the differentiated social category to identify with that category in order to press claims of

12 340 David Taylor recognition and entitlement. This may require such a deep identification that it violates all other aspects of a person s identity. The conceptual and political problems here stems, I believe, from the incorrect assumption that difference and identity are in some way the same thing. Discussions of difference are essential for an understanding of the ways in which particular claims to welfare can be asserted, but analysing claims to welfare based on difference is not enough to help us understand how categories of identity are mobilised in welfare discourse. This requires a consideration of social identity in relation to social difference. SOCIAL IDENTITY Rather than focus on identity as a function of difference we need to understand identity, not as unitary, but, nevertheless, as coherent, as possessing unity the I at the core to which Brah (1996) refers. Seen through the prism of difference alone, unity seems chimeric. Looked at as a condition of identity, however, it may be possible to talk about concrete subjectivities which are neither simplisticly unitary nor infinitely fragmented but possessing a coherent unity embodied in concrete individuals. The first step is to distinguish between two uses of identity. The first is the use of identity as difference described above, where identity is related to the social categories and common experiences of difference. The second is the use of identity as a coherent sense of self. We might call the first, categorical identity and the second, ontological identity. These different uses should not be seen as binary oppositions, however. They are not discrete entities, nor do they stand opposed to each other. Ontological identity is only realised through categorical identity and categorical identity is the expression of ontological identity within social relations. Young (1995, p. 208) distinguishes between these two levels of identity, invoking Nancy Chodorow s deep psychological sense of identity on the one hand (ontological identity), and identity as, self-ascription as belonging to a group with others who similarly identify themselves (categorical identity) on the other. Both these aspects of identity are discussed in relation to gender, and both rejected as adequate explanations of the series woman. Whilst I believe Young is right to reject them in relation to specific categories, such as gender, this conceptualisation can, nevertheless, be helpful in understanding the different processes at work in the construction of a coherent identity. The point is that subjects are created in complex social relations inscribed with a multiplicity of social categories of difference class, gender, race, ethnicity, sexuality, disability etc., but individual subjectivities are forged out of these social relations

13 Social Identity and Social Policy 341 into a coherent sense of self-identity, which may be multiple but none the less has unity. Whilst gender, for example, may not be an adequate label to describe the basis on which all those positioned as women in social relations construct their identities and become social agents, all those positioned as women must nevertheless construct a unified identity building coherence out of the multifaceted experience of diverse social relations which will have gender as one of its components. At the same time those who do identify primarily with a social categorisation based on gender, may well see this through the prism of other social categories, for example, class, race, ethnicity or sexuality, depending upon how such categories are signified and positioned in social relations. This is the process of construction of a coherent unity out of the fragments of difference inscribed in multiple social relations. The fact that a woman asserts the oppression most meaningful to her at any one time does not mean that is the totality of what she asserts. The different identifications she makes bear differently upon the coherence of her subjectivity depending on the context within which they are asserted and the meaning they have for her. Both aspects of identity exist within power relations which offer opportunities to express identities and identifications and constraints which seek to ascribe identities and attribute characteristics. These social categories of identity, then, are neither structurally determined subjectivities which can be read-off from biology or production as some versions of modernism would have us believe, nor free-floating and freely chosen through consumer life-styles as some interpretations of postmodernism would have it. They are constructed through processes which operate at the levels of production and consumption, but also of politics, cultural significations and institutional practices, and are shaped by macro and micro relations of power involving old and new political forces and movements. Some of these social categories are more resistant to change and some are more fluid and open. Some of them seek to ascribe essentialised characteristics to their incumbents and some may suggest themselves as social constructions as fusions, cross-overs and other forms of cultural hybridity. If we examine the relationship between these two elements of identity we need to understand the way in which, on the one hand, individuals build a sense of coherence through their multiple social identifications, and, on the other, the way in which categories of identity act back upon their incumbents, often ascribing ontological characteristics to their members. The former process may help us to understand the way in which individuals form attachments to social movements and enter into

14 342 David Taylor political agency in this case around struggles for welfare, and the latter, the way in which identity categories become inscribed in welfare discourse, positioning their subjects with ascribed characteristics. The meaning and content of social categories of identity is, then, central to understanding the process of identification of individuals with these social categories. This meaning is social and individual at the same time. The process of construction is also its articulation in practices, traditions, rituals, ceremonies and institutional forms. These are the processes by which identity categories are given social meaning it is within these practices that individuals come to recognise themselves as participants in social relations. Two key questions arise: to what extent and in what way do these categories through which the self is recognised constitute an adequate description of identity? and, to what extent do these categories claim to represent the totality of an individual identity? This is not a closed process. Different categories operate differently upon their subjects depending upon their position within relations of power. Some claim a greater fixity of content, and some seek to provide more or less exhaustive descriptions of their incumbents. They operate within discursive contexts which seek to give accounts of the nature of persons, their moral characteristics, their real or genuine needs, the legitimacy of their claims to welfare, and circumscribe their legal and physical access to benefits and services. Approached from above, this becomes the question of how dominant welfare discourses seek to totalise social categories of difference into an exhaustive representation of subjectivity in order to position or fix subjects through an ascribed identity with typically essentialised characteristics. Such constructions violate the complex unity of identity forged out of difference by characterising, in a totalistic fashion, one particular attributed essence or a constellation or related essences as accounting for the complete ontological identity of the subject. Such a process can clearly be seen in constructions which fix upon a partial social categorisation and impute to it an inescapable and limiting set of moral characteristics. This is most readily demonstrated in relation to the positioning and totalising effect of discourses of single motherhood, welfare scroungers, or the underclass, but may also be present, for example, in equal opportunity discourses which, for resource allocation purposes, squeeze individuals into discrete categories violating the rounded nature of their identities and needs. Approached from below, this becomes a question of how social welfare movements might assert their specificity or difference in the context of what must be a universal claim for recognition a claim which inter alia

15 Social Identity and Social Policy 343 recognises difference and the right to assert a particular need yet one which does not seek to subsume the complexity of individual identity within the category through which it is made present. Both these top-down and bottom-up approaches to welfare rights and welfare needs are clearly caught in a dilemma. In order to make any claim for a particular group of users, a specificity must be asserted which always runs the risk of overwhelming other aspects of identity. Yet without such a specificity the voice may not be heard above the uniformity of universalism. Without some universal criteria, such as equality or justice, and without a universal right to make claims, to what is specific voice appealing? And without asserting the aspect of identity which is the most salient, though partial, feature of the group s positioning in social relations, what distinguishes its needs from those of other groups? A consideration of the related issues of universalism and particularism as principles of welfare may help us overcome this dilemma. Notwithstanding those postmodernist voices which see all universal claims as terroristic (Lyotard, 1984), I believe the retention of both universalism and particularism may help us clarify the relationship between identity and difference. In the same way as identity and difference are analytically distinct elements of the same social condition, universalism and particularism may be seen as different aspects of a single philosophical principle. Theorised in this way they may provide a backdrop against which to understand the relationship between the simultaneous assertion of identity and of difference. This issue has been explored by Thompson and Hogget (1996) in a discussion of universalism, selectivism and particularism in social policy. Having set out different interpretations of universalism and particularism, they suggest, at the logical level, each type of principle depends upon the other; more strongly, such principles actually interpenetrate each other (p. 33). At the abstract level, universalism entails the idea of an independent measure to adjudicate between different cases, and particularism implies a normative universal criteria which demands that differences be recognised. They argue, that although the two points just made start from different positions they arrive at the same place together. On the one hand, universalism seeks to provide a fair standard by which to treat particular cases, and, on the other, particularism derives its moral force from an underlying universalism (p. 34). Universal welfare principles are always in practice limited to particular groups, most typically on a territorial basis, but further, within national communities they are limited to particular categories of claimant there are always eligibility criteria. Perhaps, then, universalism

16 344 David Taylor applies only to the degree of impartiality applied to the process (p. 27) rather than to unrestricted eligibility. The authors demonstrate how such a principle is tied to an institutional model of welfare which, in applying universal criteria impartially, frequently discriminates in terms of outcomes against those groups with particular needs. Both extreme universalism and extreme particularism are logically flawed and run the risk of negatively discriminating against individuals and groups. This does not, of course, mean that there is no difference between them, as justificatory principles of welfare quite different discriminations may be practised, intentionally or unintentionally in their names. Universalism may institutionalise discrimination in a passive sense by its refusal to acknowledge difference and the social relations of power within which it is inscribed. Particularism may draw attention to difference and mark it out as inferior, problematic or stigmatised most notably in the form of negative selectivism embodied in the idea of means-testing. Thompson and Hogget s aim is to combine universalism, particularism and positive selectivism (positive discrimination) in such a way as to balance guarantees of equality, fairness and impartiality on the one hand, and allowance for diversity and difference on the other (p. 39). They advocate a version of the associationist democracy outlined by Hirst as a vehicle for welfare provision which can accommodate both. Of less concern here, is whether associationalism can deliver than the authors view that universalism and particularism must incorporate each other. It is an approach which is close to Benhabib s interactive universalism, outlined by Hewitt (1994), which acknowledges the plurality of modes of being human, and differences among humans, without endorsing all these pluralities and differences as morally and politically valid (p. 53). This generates an ethic of caring and responsibility to others that is grounded in the human state of contingency, but founded none the less on the possibility of universal reasoning (ibid.). This theorisation of universalism and particularism as different aspects of the same principle is helpful in understanding the relationship between difference and identity. There is a tension inherent in the very notion of identity itself which parallels that found in universalism and particularism. It is, at the same time, a description of the unique and a description of the same. My identity is something which is only mine and I am different from others, but at the same time my identity is only possible because I exist in relation to others and can conceive of myself as part of social relations which I share. I can recognise myself as both different and the same. There is no difference without sameness and no sameness without difference, as each is the measure of the other. As Calhoun (1994) remarks:

17 Social Identity and Social Policy 345 Each dimension of distinction is apt at least tacitly also to establish commonality with a set of others similarly distinguished. There is no simple sameness unmarked by difference, but likewise no distinction not dependent on some background of common recognition. (p. 9). Identity as sameness what I have referred to as categorical identity results from the recognition and classification of others as the same, i.e. as a social category. Identity as uniqueness what I have referred to as ontological identity results from the recognition of individual selves as different from other selves. Identity therefore is a project or an achievement one of awareness, recognition and identification. The paradox of uniqueness and sameness is easily resolved, of course, by recognising that individuals as unique selves are only formed within social relations between others and through the participation in and construction of social categories. There is no pre-social individual squatting outside social relations. I only recognise my sameness by virtue of my difference and only my difference by reference to a common standard I share with others. This recognition or identification can be at the most general level as a member of the human race, or at more particular levels in terms of specific social categories within particular social relations class, gender, ethnicity, sexuality etc., which usually form the basis for political agency, whether of the old or new social movement variety. It is both to the particular and to the universal that social movements appeal. They assert their difference but typically against a universal recognition criterion and a universal standard or principle such as equality, based in rights or needs. These differences upon which claims are made must always be understood within the complexity of social relations. Social difference does not exhaust individual identity. In complex societies, with multiple social relations between groups, identities are constructed in and through a myriad of mutually constituting social relations. As we have already seen, however, this totalising of categorical identity (making one particular marker of social difference stand for the entirety of an identity) may be used from above to ascribe moral characteristics of disentitlement or from below to limit differentiation in the name of communal interest. We need to understand, therefore, discourses of power which seek to construct their subjects in total subordination and attribute to them negative essentialist characteristics, expressed through totalising categories of identity. When these become institutionalised in social and cultural practices a simplistic deconstruction which ignores the power which sustains them will not be sufficient. Long-standing historical notions of identity are not rendered irrelevant for all the arguments that they may be mythical or imagined. Imagined essences of identity are potent social forces, and

18 346 David Taylor appeals to them have been very powerful. This can be illustrated in certain ethnicist discourses. The confrontation between essentialised mythical others, is typically the case with ethnic conflict in which the other must be abolished from within. The absolute self, which is essentially fixed (in tradition, community, religion, ritual, territory etc.) comes under threat from the mere possibility of the other. In its most extreme case this takes the form of ethnic cleansing in which any form of other must be expelled from the social body. To admit the other to the territory of the absolute self is to raise the possibility of the uncertainty of self absolutism. Ethnic cleansing is as much a confirmation of the self as the denial of the other. To admit the other, in no matter how small a degree, is to admit the possibility of relation, and relation is always subject to change. It is to admit the possibility that we may become what we currently are not, because in us there is the potentiality of the other what Julia Kristeva has called, the foreigner within. In such cases, particularist discourses may be at least as terroristic as the universal discourse which Lyotard derides. The key, then, is the recognition that difference categories do not represent the totality of identity and that the formation of identity is both an historical process and an individual project. It is one which takes place, none the less, within relations of power which construct categories of identity as dominant and subordinate. These are important issues for a welfare politics of identity. A politics of identity which seeks to found political interests on the basis of social differences can be extremely dangerous. The categories of identity behind these differences may themselves become essentialised and ossified and lead to a particular reading of interest which flows from only one particular identification. An example of this is discussed by advocates of queer theory and politics (Seidman, 1995) who argue that gay politics assumes certain identifications with a social category the gay community which does not hold for those who do not identify with unitary sexualities or who see sexuality as a partial aspect of identity. This issue is indicative of a problem with a politics of identity which takes the categories of social difference as the basis for total individual identity. This may not be an example of terrorism but it does point to the inherent problems of allowing a position based on difference within social relations to become an assertion of identity which leads, in Fiona Williams words, to difference becoming division (1996). It arises because of an under-theorisation of identity and its relation to difference. IDENTITY IN WELFARE DISCOURSE An example of the way a totalistic ontological identity has been attributed to a particular group can be found in welfare discourse around the