Scholarly article on topic 'Social transformation, collective categories, and identity change'

Social transformation, collective categories, and identity change Academic research paper on "Sociology"

0
0
Share paper
Academic journal
Theory and Society
OECD Field of science
Keywords
{""}

Academic research paper on topic "Social transformation, collective categories, and identity change"

Theory and Society (2005) 34: 429-463

© Springer 2005

Social transformation, collective categories, and identity change

JENNIFER TODD

University College Dublin

Abstract. Changes in collective categories of identity are at the core of social transformation. The causal linkages among identity change, institutional change, and change in modes of practice are, however, complex. Developing and adapting ideas from Pierre Bourdieu's work, this article shows the coexistence in tension of a plurality of elements within each collective identity category. On this basis, it proposes a typology of responses at the level of identity to socio-political change. This allows an explanation of patterns of identity change in terms of wider social processes and resource distribution, while remaining open to the sense and complexity of the individual's experience and the moments of intentionality that arise when individuals face choices as to the direction of change. The worth of the model is shown by analysis of modes of identity change in a society now experiencing radical change in socio-political structures, namely post-1998 Northern Ireland.

A succession of recent works has focused on processes of social and political transformation. In all of these analyses, the process of change in collective categories of identity is central. McAdam, Tarrow, and Tilly see identity change and cognitive shift as core elements common to all forms of political contestation.1 Changes in identity and aspiration play a key analytic role in a number of studies of the new European regional-nationalisms.2 In other studies of national, ethnic, and social mobilization, processes of self-categorization have been core elements of explanation.3 It is clear why identity has thus come back into analysis: it is only when institutional changes are accompanied by changing self-perceptions that new institutions begin to create new dynamics of interaction; otherwise new institutions and practices become assimilated within older meanings and oppositions.

That institutional and identity change are interrelated has its classic statement in Marx's third thesis on Feuerbach.4 The question is how best to analyze and grasp the interrelations. McAdam, Tarrow, and Tilly argue that identity, institutional, and interactive aspects are interrelated

in each phase of mobilization and contestation.5 They see social interactions and relations as core aspects of social life, institutionally and environmentally structured and mediated through individual perceptions, self-categorizations, interests, and strategic calculations.6 Within this nexus, variation in identity categories is an important link in the causal chain. On the one hand, changes in the meaning of identity categories may be key variables in the explanation of change in political and social behavior.7 In ethnic interface situations, for example, subtle shifts of meaning may make a difference in who is included or excluded, killed or let live, and whether peace or conflict prevails.8 On the other hand, change in identity categories is itself provoked by and responsive to changes in institutional structure and social practice.

The causal links between category change and interactional and institutional change are often complex, with time lags. There may be gradual disruptions of cognitive categories over time, underlying concepts that are put in question, inchoate cultural unease, and new practices that allow old concepts to fade into irrelevance, and these subtle changes provide the underlying conditions in which new categories suddenly become fore-grounded in practice, new self-definitions are crystallized and major institutional change occurs. Gradual changes in the "cultural substratum" then become threshold conditions for sudden radical category change and social transformation. In other cases, identity categories have their own inertia: imposed socio-political changes with correlative changes in social practices and incentive structures fail - at least at first - to impinge on collective categories of identity. This has been the case with European integration and with many peace settlements in situations of ethnic conflict.9 Short of identity change, these institutional changes remain less than transformatory; in some cases, older conflicts are simply transposed within the new structures.10 The key political question in such cases is whether and when identity change will follow and how it may be promoted.

Identity change is, however, complex to recognize and explain. McAdam, Tarrow, and Tilly treat identity shift, in its subjective, meaningful dimension, in very broad strokes.11 They are concerned with shifts in macro-categories, rather than with more subtle shifts in modes of identification, categorical content, and relations among categories.12 Yet the changes in contemporary national identity posited in the theoretical literature, to take one example, involve not just change in the bare category of nationality, but in the mode in which it is held and

in its associated meanings.13 In practice, the changes that make major impact are not, or not solely, the giving up or adoption of a Spanish, or a European, identity, but becoming a different sort of Spanish or European.

If the fact or failure of change in collective identity categories crucially affects social behavior, what factors favor such change? A typical explanation in the literature appeals to the activity of political or "identity" entrepreneurs - for example, rhetorical category formation, brokerage, definition of opportunities, or manipulation of fear.14 These are elements in any explanation of category change, but they are not the only elements. Political entrepreneurs are always with us and most often they fail to meet any public response. Then, suddenly, they mobilize the masses.15 Why do some entrepreneurs succeed and others fail, and why does success come when it does, rather than earlier or later? Part of the explanation lies in social structure and changing interests, but part lies in perceptions, in peoples' sense that these are demands that they would be happy to stand behind, or that the discourse simply does not speak to their key concerns. Once again we are thrust back to the "cultural substratum," the changing ways individuals and groups understand and interrelate categories.

To identify subtle changes in collective identity categories requires attention to the cultural meanings of these categories as well as to behavior and boundaries, and analysis of how these meanings are constituted and changed in different micro-contexts, as well as at the macro level.16 Bringing meaning and variation back in makes explanation considerably more complex. I argue in this article that it is possible, at least when we focus on a few politically crucial meanings of identity categories (in this article, on the interrelations of oppositional meanings and substantive values within the same national identity categories). It is also worthwhile: it allows us to discover patterns in the ways in which collective categories are used and reshuffled in times of social change, the instability of some meaning-configurations and their evolution into others, and dissonances between seemingly authoritative official meanings and on-the-groundpopular meanings: in short it allows more subtle typologies and robust explanations of change in collective identity categories. It is also of political importance to recognize patterns of identity change and the conditions that favor or disfavor particular directions of change while change is ongoing, before an old stability of categorical opposition evolves into a new stability of categorical opposition. Just

such a period of categorical flux is discussed in the final section of this article.

In the next section, I argue that the work of Pierre Bourdieu, which incorporates social interaction, institutional structure, and subjective meaning, can usefully be developed to deal with identity categories and identity change. On the basis of this discussion, I outline a typology of logically possible responses, at the level of identity, to socio-political change. This is intended at once to allow the explanation of patterns of change in terms of wider social processes and resource distribution, while remaining open to the sense and complexity of experience, and the moments of choice (and the constraints on choice) that arise in social transformations. I then show how these responses are exemplified in one contemporary case of politically-engineered change - Northern Ireland after the Good Friday Agreement. Most of the examples in this article are about changes in national categorizations, although the argument is intended as a more general one.

Collective categories and the formation of identity

Collective categories of identity

As has been shown in a number of important recent articles, a range of contemporary scholarship has converged in highlighting personal variability and new combinations of meanings in identity categories and moments of choice or intentionality in identity formation, as exemplified in individual life stories.17 This "soft constructivist" analysis decisively breaks with the notion of identity-categories as fixed and defined, as concepts into which individuals fit. Just as there is no longer any hegemonic cultural order, so there are no longer any hegemonically defined identity-categories, national or otherwise.18 It emphasizes the fluctuating, relational, and situational quality of self-definitions that are constructed in social practice and interaction, not in depth psychology.19

Brubaker and Cooper show, however, that this approach can easily become conceptually incoherent.20 Identity becomes plural, identities proliferate, varying in each situation where a new aspect of self is performed. "Identity-language" classically referred to the stability of the self through a succession of roles; once "identity" is unmoored

from such individual stability, it loses its raison d'etre. It also loses its usefulness for the analysis of social transformations. If we think of identity in terms of multiple, free-floating macro-categories that individuals may choose to emphasize or ignore, depending on charismatic orators, calculations of strategic self-interest, the needs of particular situations, their degree of suggestibility at any point in time, identity change loses any claim to be a significant part of the causal patterning of social change. It becomes no more than interactional change, epiphenomenal.21

If, however, we take seriously the view that collective categories of identity play a significant causal role in framing action, and if we posit a slowly changing "cultural sub-stratum" that may underlie more radical category change, we need a different model of how identity-categories function. We need to recognize not just the complex and varying meanings of these categories and their lack of fixed or foundational status, but also their social "embeddedness" and their personal "anchorage," which allow change or stasis to occur out of phase with other variables, and to affect them in turn. Pierre Bourdieu's work provides just such a model.22 In what follows, I outline the importance of his perspective for the task set out in this article, and then show that it must be adapted to cope with social transformation (rather than the social reproduction that he analyzed), to allow for the "moment of intentionality" and to analyze identity formation and change (rather than the broader concept of the habitus). This falls well short of a theory of identity change, but it allows the identification of conditions that are likely to provoke change in particular identity categories, the development of a typology of ways that this change may proceed and of hypotheses about the conditions that make it likely that it proceeds in one direction rather than another.23

Habitus, social practice, and identity

Pierre Bourdieu has shown how the social patterning of individual subjectivity and categorization takes place. He has given a powerful theory of social reproduction whereby individuals internalize the distinctions and values that structure the social world, and then, acting spontaneously in the light of these distinctions, reproduce the social structure whose meanings they have internalized.24 He has analyzed how the habitus - dispositions that give bodily form to collective categories and distinctions - is formed from earliest childhood, as the child situates itself in structured, gendered and differentially "pathed" social

space, and internalizes its objective life-chances in its own subjective expectations and interests.25 Such processes produce a set of cumulative, superimposed meanings, dispositions, and modes of perception embodied in the individual. Some of these may be laid down in infancy but they are social products, even if they sometimes appear to individuals as "primordial" givens.26

Social practice is at the core of his analysis; it is where individuals at once encounter and internalize distinction, and where, even in new situations, the habitus reproduces socially structured distinction and ensures historical continuity as if spontaneously.27 Through this process, collective categories are interrelated in understanding and habitus just as they are institutionally. The precise set of interrelated categories from any particular standpoint in any particular society gives the subjective possibility of shared experience. This is differentiated in national, class, gender, regional, and local terms, so that "being English" differs from different class standpoints, and also from different gender, regional and local perspectives, just as class belonging takes specific national, ethnic, gender, regional and local forms.28 These are not, on Bourdieu's view, multiple "identities" that one wears and changes, but rather the overlapping categories merge into particular and distinctive forms of habitus.29

Bourdieu moves decisively beyond many of the dichotomies that characterize contemporary writing on identity. Individual and social identity, instrumental and affective modes of identification are intrinsically interrelated in his writing.30 Collective categories are embodied in the individual habitus and expressed not simply in collective action but also in the subtler signs of individual distinction, down to the choice of jewelry, make-up, and clothing.31 This reproduces collective distinction at the micro-level, ensuring that spontaneous attraction and friendship, marriage and family relations tend to fall within collective categories. All such embodied collective categories bring with them a felt immediacy. All give the potential for immediate contact with others whom we recognize as "like ourselves," and a related sense of belonging with those who share our categories and who recognize and respond positively to our immediate intuitive distinctions. But while, in practice, the embodied categories carry this emotional charge, analytically the formation of shared experiences and the "warm" sense of group belonging and solidarity may be explained coolly, in terms of core social variables.

Bourdieu's theory, in these respects, provides an invaluable startingpoint for analysis of identity. It has, however, been criticized for its failure to explain change and to account for variation.32 Whether these criticisms are accurate for Bourdieu's work as a whole is a question beyond the scope of this article.33 They do, however, show that the preliminary schema presented above requires refinement. That schema appears to explain social reproduction and social evolution but not radical social change.34 It appears that such change can only come from outside intervention or the impingement of wider processes, since internal feedback patterns ensure the reproduction of both the habitus and the social order. Radical cultural change can, it seems, only be the delayed result of such imposed social change.35 However, if this is the only source of cultural or categorical change, it may be signally ineffective, since new elements may be "indigenized" by social practices determined by older distinctions. This practice, well theorized by Bourdieu, is a common phenomenon in deeply divided societies where actors are able to incorporate new resources brought by potentially radical social transformations such as industrialization or democratization or European integration within the divided social structure.36

Part of the problem lies in the positing of a single dominant social and symbolic order. In complex modern societies, there may be several competing symbolic orders coexisting in tension within specific institutions or "fields," not just one authoritative system.37 Studies of the state show, for example, different sets of policy orientations and different key conceptual oppositions embedded in different state institutions and sometimes competing within them.38 Class distinction too may be experienced and produced in different ways: Lamont's studies show that bourgeois individuals may choose to define themselves in terms of ethical rather than cultural or material distinction and this choice does not appear to follow the particular bourgeois "fraction" to which the individuals belong.39 More radically again, there are some societies where one finds not just contested variants of one symbolic order but at least two different symbolic orders within deeply contested social institutions. This is most evident when one moves from class to ethno-national or religious divisions. In situations where two groups with their already formed symbolic systems are brought into conflict, the distinct cultural substance associated with each is not lost, even when one group gains decisive socio-political control. The symbolic orders tend to become formally homologous -precisely because the dominated order must answer and oppose each

aspect of the dominant - yet they retain distinct cultural substance.40 It follows that what the child, or adult, internalizes from social practice may be alternate rather than official meanings, recessive aspects of the social order, implicit contradictions between one set of practices and another. What is embodied in the habitus is a transposition of (and in some circumstances an alternative to) the distinctions of the "dominant social order," and it may come more or less into tension with it.

There is also a tendency to narrow the "grammar" of the habitus to a set of binary oppositions based on power relations. Bourdieu's explanation of social reproduction centers on the role of structured social practice as the mechanism by which objective life-chances are transposed into the habitus. More than binary oppositions are thus constituted as dispositions. Complex social practices typically involve at once relations with others (which can build cooperative capacity, skills and virtues) and relations against others (which mark one off from those outside).41 The point is clear in Bourdieu's own empirical studies of crafted interactional patterns or the forming of aesthetic taste.42 Yet the real values, and sometimes also virtues, associated with these practices are subsumed within class distinction, and not prioritized in his analysis.43 Thus one of the internal tensions within the individual habitus - and another potential source of change - is elided. For example, ethno-national distinction is built both from participation-with-others in institutions and social practices, and by exclusion-of-others from these institutions and practices: at once from the multitude of everyday practices and norms that make up "banal nationalism" and from an institutionally credited ranking of these practices that defines the value of cultural capital and constitutes a set of power resources for some and against others.44 National identity at once involves positive values, which grow out of participation and mutual dependence, and implicit oppositions with lesser nations.45 Similarly, in deeply divided societies, ethnic categories are mutually opposed, yet they also resonate with values derived from specific traditions and practices.46 Ethnic and national conflict is all the more intense because the actors know that what is at stake is not empty categories or mere power, but also real values rooted in specific traditions. When we move from consideration of the national category of identity to the interrelations of nationality, class, gender, familial, career, and other categories, the possibility of internal tension within the habitus increases. While, in principle, one is not gendered separately from being national or from being a member

of a given class, the contents and values associated with each category take different prominence in different fields and may co-exist in considerable tension.

A third general point arises from the fact that Bourdieu's work centered on the habitus, that totality of dispositions produced without conscious intent. Identity, however, is not the habitus: there is too much in the habitus, much of it below the level of consciousness, much irrelevant to how one conceives of oneself.47 The habitus rather gives a substratum from which identity is formed in a process of arrangement and rearrangement of some of these elements of embodied meaning and value, with some held close to the center of self, others left on the margins.48 Identity-narratives are the reports of (sometimes the performance of) the arrangement and rearrangement of these blocks of meaning.49 Identity formation and change is thus a continuous process that involves a considerable degree of intentionality. It takes place by the incorporation of new elements of embodied meaning and value, or the rearrangement of old. New elements may be created, not ex ni-hilo but by the choice to foreground particular practices and relations rather than others, so that over time the meanings embedded in these practices become an integral part of the self, while others fade. Such choices are not costless: after years of work and sacrifice, one may find that a sense of belonging in one's adopted society still eludes one; similarly, old meanings may be marginalized but never totally fade. By the same sort of process, this time combining social practices in new ways, new combinations of meanings can come into being. Again, choice is typically constrained, on the one hand by socially entrenched symbolic codes and multiple social pressures not to break them, and on the other hand by personal history that may have laid down some dispositions so early in infancy that they are changeable, if at all, only by depth analysis. Complex patterns of identity change may result. When, for example, peripheral nationalist parties in Catalonia or Scotland or Northern Ireland at once oppose the central state and seek involvement in European networks and global linkages, they begin to constitute a hard-edge anti-statist but non-exclusivist, hybrid national identity. If, however, access to European and global networks is restricted until anti-statism is softened, this tends to create a different sort of national-identity configuration and one that may not meet voter expectations, thus opening the way to future shifts in voter support (from CiU to ERC in Catalonia, from SDLP to Sinn Fein in Northern Ireland).

These points allow us to identify three mechanisms of change in collective categories of identity: dissonances between the social order and the individual habitus; dissonances within the individual habitus; and the moment of intentionality in identity formation. They also allow a more subtle and differentiated analysis of collective identity categories. First, the discussion shows that meaning of collective identity categories is complex, composed of a plurality of elements coexisting in tension. These elements include binary oppositions, substantive values and virtues and normative principles, cognitive assumptions, and understood relations with other categories.50 Each is derived from meanings embedded in the social structure. A mapping of these elements, and of their typical combinations, can be achieved by focused

research.51

Second, there are diverse ways of constructing identity categories. The elements may be intertwined in different permutations and combinations with varying degrees of tension. The meanings of these categories cannot therefore be read off from dominant political discourses: there may be numerous ways to "be Irish" or to "be Basque," quite different from "official" views, pursued spontaneously by different groups and sub-sets of groups and individuals, combining differently regional, local, gender, age, class, religious, and cultural stances. These may not adequately be represented by politicized contest in the public sphere, yet they are in no sense "private languages" of purely personal significance; each variant may be recognizable within the culture and appeal to particular historical traditions within it.52 Arguably this divergence of popular and "official" categorical identification, and variation and flux at the popular level, is a phenomenon of the contemporary age, experienced by many individuals reassessing their options in societies experiencing economic restructuring, Europeanization, globalization, and new phases of immigration.53

Third, there is a "moment" of intentionality in the making and remaking of identity categories, a choice of which permutations and combinations of elements to accept. This choice is, however, highly qualified. Its exercise may require a high level of cultural and social resources. For example, to distinguish oppositional aspects of national categories from the cultural substance of these categories when both are conflated in one's identity requires considerable cultural resources and intellectual stamina, while to separate the meanings in public interactions may require unusual courage. To engage in new social practices that allow a recombination of meanings requires institutional and social

opportunities. Change may be achieved only at high personal cost: as elements of meaning are centered or de-centered, the interpretation of key episodes in the past in which these meanings came into play may also be changed, and the sense of self may be affected.54 Cultural and social resources, not least in the form of public acknowledgment of the costs of widespread change and public recognition of multiple perspectives on the past, can aid such processes of change.55 The form and distribution of key social and cultural resources, therefore, are among the determinants of the likely directions of change in identity categories.

Fourth, change in collective identity categories is on-going. Individuals and sets of individuals often find their intuitive categorizations out of phase with those of their class or group, and still more so with official, state-centered categorizations, or they find that the elements of their own categorizations come into internal tension. Unease, personal discontent, and revision of category structures proceeds, sometimes independent of, sometimes cohering with, sometimes appealing to new modes of identity-construction put forward by political or communal entrepreneurs.56 More radical, society-wide identity change is provoked when socio-political changes bring the elements of collective identity categories into evident contradiction for whole populations. Then individuals are forced to re-sort the elements of their identity. Which directions are chosen depend in part on the prior processes of gradual identity reconstruction within the population, in part on available new resources. When whole populations face such choices, the directions chosen have major political effects.

Social-political change and identity change

Modes of identity transformation

A key cause of change in categories of collective identity is social change. In a society structured throughout by a key set of power relations, radical change in these relations will also cast in doubt the oppositional elements of the collective identity category and their interrelations with other elements. Where, for example, economic position has been correlated with racial or ethnic or gender categories, effective fair employment practices change the "entry tickets" to economic positions and with them the socially sanctioned interrelation

Table 1. Direction of change in collective identity categories.

No change Partial change Total change

Transparence and Reaffirmation Assimilation Conversion

coherence between practice and category Ambiguity and tension Adaptation Ritual appropriation Privatization between practice and category

of these categories with meritocratic norms and pride in achievement. With the official recognition of the value of minority or dominated cultures, the socially sanctioned association of particular cultural categories with norms of progress, civility, or rationality is "decertified."57 With new actors in positions of power, the networks of informal influence change, and with them socially sanctioned expectations of having, or not having, an authoritative voice in society. Where identity has been entwined with power, these changes are experienced not simply as a change of regime, but - for the dominant group - as an overturning of the moral order, an insult to their own integrity and identity, a placing of the undeserving above the deserving. It is a particularly sharp form of dissonance, where the world is not ordered as they had come to expect, and where these expectations were constitutive of their sense of themselves. Those who have long opposed the dominant order may also find that change disrupts the categories in terms of which they had defined themselves. Others, already uneasy with the dominant categorical oppositions, will welcome the opportunity to move beyond them.

If identity shift is predictable in these circumstances, the form, direction and phasing of that shift is likely to vary depending on the resources available for identity change. Table 1, above, distinguishes responses in terms of the extent of change and the way in which it is expressed in practice. Even those responses that begin by refusing identity change tend to introduce new elements and tensions into identity-content.

Reaffirmation

This option reaffirms the existing core binary oppositions and welcomes or resists change in their name. For those who have benefited,

the elements within the identity-category may be reaffirmed because they sense that they are winning, they are "on a roll," ready to seize the chance to give public prominence to their values, ideas, and concepts at the expense of their former superiors. Those who were once dominant, in contrast, reaffirm their categories in the very process of resisting change. In some cases, where those who resist set up their own networks and social mechanisms of mutual support (as did the French resistance in WWII), the resistance option may be self-confirming. In other cases, resistance introduces new elements into the identity structure: the experience of marginalization, the practice of conservation, and the threat of extinction produce quite different cognitive assumptions, meanings, and values than did the previous experience of successful, expansive social interaction.58 Sometimes, too, key elements of the old identity-categories may be eroded - respect for authority, the belief in progress, the sense of order, and the expectation of influence are unlikely to survive social marginalization and sustained opposition. In these cases, while the oppositional aspect of the collective identity category is retained, its evaluative content, core assumptions, and relations with other categories change, sometimes to the extent that the identity-structure collapses and reaffirmation turns to conversion.

Conversion

For those who were once-dominated and who constructed their identity against the dominant order, as for those who were once-dominant, structural change may render irrelevant older categorical oppositions. As the linkages between the elements of the identity-category collapse, and as many of the old elements are sensed as absurd in the new situation, the symbolic grammar embedded in the new order may simply be accepted as a package, in a form of conversion. Of course, converts must also find within themselves the capacity to work within the new categories, or else they will be liable to disillusion and further change. The stability of conversion depends on how far the substratum of individual dispositions may be reordered within the new structure. In some dramatic and seemingly sudden cases of category change, for example with whites during the South African transition, the conditions for change had long been prepared in the increasing lack of fit of categories to practice, in inchoate unease as self-perceptions did not meet the perceptions of others, in the strain of holding to an ideology that no longer met either interests or experience.59

Privatization

An alternative to conversion, when older oppositional categories and meanings are thrust into irrelevance or official disapprobation but the new categories implicit in practice cannot be accepted, is privatization. This option rearranges the elements of identity, marginalizing all macro-social elements, all national, political, class, and status categorizations, shrinking the core of identity into the private, the familial, perhaps also the religious sphere. This option requires strong family linkages, private cultural and social resources, or religious faith (or alternatively, enough spending power to move to a consumerist individualism). Collective categories may not be rejected fully, but narrowed to the personal sphere: religion becomes pietism and prayer; national identity shrinks to a personal or familial enjoyment of national culture.60 This option has been described in numerous autobiographies of people living in fascist or authoritarian regimes. The pushing aside of collective categories of identity may not be a total rejection of them, simply a matter of making them recessive, thus allowing persons to live relatively happy personal lives under regimes that are alien to them, while re-appropriating the older elements when it becomes possible to do so. In this latter case, privatization begins to converge with adaptation. In other cases, privatization may become impossible if ascribed collective categories affect even the pursuit of private happiness.

Adaptation

In this case, actors adapt to the practices required in the new social order without changing the core elements of their identity. They are "sensible" and agree that the new is here to stay, but they "act" in it while keeping their own values and self-categorizations distinct from this practice. Equally, they keep at a distance the meanings and values embedded in their new social practice, ensuring that they do not touch their core identity. Examples may be found in many social groups in Franco's Spain: Catalans who acted prudently but retained their older values and identity and seized the opportunity for change when it came.61 In this case, however, the meanings embodied in their newly adaptive social practice remain a second language whose logic and rationale always remain alien; adaptation is always grudging, incomplete, with signs of the older values and oppositions always likely to appear.62

Assimilation

In this case, actors reshuffle the elements of identity, marginalizing some binary oppositions that were associated with the old social order, placing other categories closer to the center of identity; differentiating the substantive cultural meanings of, for example, ethnicity from associated oppositions based on status and power hierarchies and retaining the former while casting out the latter. They thus find in themselves the dispositions and expectations necessary to succeed in the new order, while retaining a sense of continuity with their older selves. This is autonomous development of identity, not conversion or the unwelcome intrusion of new elements. This assimilationist option is only open to those who have already available to them, as second nature, elements of value and meaning consistent with the new practices that they can foreground without a sense of total personal upheaval. It was the option taken, for example, by those in old East Germany who responded to unification not by adopting neo-liberal values (conversion), nor by resistance or denial (adaptation), but by combining their own entrepreneurial aims and skills with the strong sense of and valuation of linkages that had been built up in the old regime, thereby retaining a level of continuity with the old (East) German-ness while adapting successfully to the new.63

Ritual appropriation

In this option, new practices are accepted and assimilated within old narrative forms and ritual structures that are used to legitimate, appropriate, and redefine the practices, thus assuring continuity of meaning despite change in practice. Unlike adaptation, however, the new practices and the older categories are not held apart but mutually affect one another, so that the tensions and contradictions between the new elements of meaning and the older categorical oppositions play themselves out in practice. This is explicitly a group-centered rather than an individual response to change; it is how "nations" and the individuals within them modernize while retaining a sense of national tradition and community.64 So, for example, the expansive neo-imperial tradition that once constituted the symbolic content of Britishness is continued ritually by the British royal family's involvement in the Commonwealth, although the changed practical significance of the British world role has given both ritual and identity new meanings and ones that threaten

- for example, in the transition-crises of the British royal family - to break the older ritual form.65

In each of the cases outlined above, change may not affect the collective categories of identification: self-reported identity as German or French may remain. It does, however, affect the meanings, interrelations, and salience of the collective categories and their role in social practice, sometimes in very radical ways.

Transforming identities in Northern Ireland

Contemporary Northern Ireland is a case where major socio-political change associated with the recent peace process and 1998 settlement is triggering shifts and strains in the content of national identity categories.66 This, in turn, is reacting back on the political process, producing crises in the process of implementation of the 1998 Agreement and in the institutions set up by it. In this dynamic situation, identity change is an important part of the causal chain. But recognizing this requires a notion of identity as inertial, out of phase with interaction yet open to quantum leaps of change. This is precisely the concept sketched above.

Identity shift was to be expected in Northern Ireland; changes in the British and Irish states and in Northern Ireland itself were increasingly disconfirming to the oppositional character of national categories there. These changes were spread over some fifteen years, impinging on individuals at different phases.67 It is useful, however, to take the Good Friday Agreement of 1998 as a turning point, because reform policies underway since 1985 intensified with the Agreement.68 In effect, ongoing changes that were gradually ameliorating the inequality experienced by Catholics and the political and cultural marginalization of Irish nationalists in Northern Ireland climaxed in 1998 in an Agreement that promised a strongly egalitarian settlement, economically, politically, and nationally.

For long, unionists had understood being British in Northern Ireland in terms of a set of binary oppositions that were derived from their role as representatives in Northern Ireland of British power, law, order, and justice. This role was institutionalized - unionists had long monopolized positions in the state, civil service, judiciary, and police force - and it was expressed in a whole range of discourses, writings, and symbolic practices - they were loyal to the Crown against

traitors and disloyal Irish nationalists, they were modernizers and realists, aligning with British global power rather than pursuing backward nationalist dreams.69 Unionism varied in its practical and ideological expressions: the habitus of the senior, British-oriented civil servant and his judgments on socially necessary policies differed from those of the part-time rural policeman and of the populist politician.70 Yet, at least until the 1990s, even the most liberal variants of unionism implicitly repeated the old binaries.71 This was similarly the case with nationalism and republicanism. A multiplicity, plurality, and increasing reflexivity in what it meant to be Irish, evidenced in interviews and in writings, coexisted with a continuing tendency to see being Irish in opposition to being British and to associate it with resistance to unjust power.72 Constitutional nationalism radically changed its ideological structure, but remained open to interpretation in terms of the classic binaries; republicanism underwent major strategic and social changes but retained the binary oppositions of British vs Irish; assimilation vs. resistance, compromise vs. idealism, and corruption vs.

purity.73

The Good Friday Agreement of 1998 signaled change in the basic power-configuration in Northern Ireland, and in the British state's role in this.74 The strongly consociational devolved administration equalized the influence of unionists and nationalists, not just for the present but for the longer term, so that republicans and nationalists, as well as unionists, would be responsible for law-making and eventually also law-enforcement in Northern Ireland. Provisions for reform of security, rights, criminal justice, and employment equality meant that the state could no longer be dependent on the Protestant and unionist community for policing and administration. The commitment to parity of esteem in Northern Ireland meant that the British state presence would no longer "certify" the value and meanings embodied in Protestant and unionist rather than Catholic and nationalist practices.

The Irish state itself, in part through increasing affluence and European integration, in part through effective diplomatic activity, had emerged as a proactive and modernizing player in the Northern Ireland conflict, and a check against the unintended bias of the British state. Meanwhile, the British state with Tony Blair in control was initiating a process of devolution in Scotland and Wales and a new Europeanist and pluralizing political discourse.75 This formed a context where the older oppositions within Northern Ireland were increasingly at odds at once with

the reality of the two states and with the dominant senses of nationality and emerging national habitus within each. The effect was to intensify unease and the sense within Northern Ireland that identity structures there were out of phase with social practice or, perhaps, with modern life.

The Good Friday Agreement, both in its egalitarian institutions and its intergovernmental frame, might have been predicted to cast unionist and republican binaries in doubt. Many unionists saw a world turned upside down. They divided almost equally between those willing to (try to) accept this new world, and those who resisted it.76 Republicans too had to re-conceive their place in Northern Ireland. Yet if identity shift was fully predictable, much of the scholarly literature has failed to recognize it, and some have argued that there has been no change in national identity.77 This seeming paradox is easily explained by the fact that many scholars work solely with macro-categories of national identity. Table 2 shows that there has been little significant change in these categories of self-reported national identity among either Protestants or Catholics since 1978; nor is there significant change when the results are differentiated by gender or by class.78 Only among youth has there been significant change in the macro-categories since the Agreement,79 with Catholic youth (77 percent of 16-year-olds) more likely to declare themselves Irish than older generations, and a significant section of Protestant youth (33 percent of 16-year-olds) now self-identifying as Northern Irish.80

Table 2. Self-reported identity: Northern Ireland.

1968 1978 1986 1989 1994 1998 2000 2001 2002 2003

British Protestant Catholic Irish

Protestant Catholic Northern Irish Protestant Catholic Ulster Protestant Catholic

Not Not 11 asked asked 20

Only when we look beyond the self-reported national categories, can the extent of identity change be recognized. It has, for the most part, taken place in the content of each category rather than involving change between categories. The typology set out in Table One allows us to differentiate and to clarify the directions of change and the possibilities of further change and to give a tentative explanation for this differentiation in terms of the cultural and social resources of the groups in question.

One cluster of Protestants has assimilated to the new order.81 They welcome the stability that it promises and the new institutions allow them to maintain the British linkages that they value, while also building potentially useful Irish linkages.82 They no longer see the point of striving for absolute sovereignty for the British state in Northern Ireland and they have moved away from a notion of British identity as essentially opposed to Irish identity while keeping what they see as the valuable cultural substance of that identity. For some, this change long preceded the Agreement and was newly certified by it.83 For others, the Agreement triggered change. Its provisions (accepting republicans in government, freeing paramilitary prisoners) went against dispositions embodied in the habitus and were intuitively sensed as morally abhorrent. Yet this group still found within themselves the resources to affirm the Agreement. Arthur Aughey, after Lampedusa, speaks of unionists having to "swallow the toad" of the Agreement in order to attain the goal of a "civil" society.84 For Aughey, one deciding factor is moral-political principle: the good of agreement outweighs the moral distaste at the means to it. For other unionists, the deciding factor was practical: the need to secure the union and "get down to business."85 Others found a religious reason for going beyond their immediate reactions and opting for settlement.86 For still others, their contacts with nationalists and republicans in civil society organizations allowed them to de-demonize these political parties and to accept coalition with them, while continuing to disagree with them politically.87 In all of these cases, values or principles deep-set in habitus, which may in the past have sat in some tension with political self-understandings, were called upon in the reinvention of a British identity and unionist politics consistent with the Agreement.

Only a small section of these supporters of the Agreement have changed national categories, and even fewer have changed both category and content.88 They have assimilated, not converted, finding dispositions and values in themselves that they fore-ground, which justify the forms

of interaction and opportunities opened to them in the new situation. It is significant that this section of the population includes the richer, more mobile, more educated section of Northern Irish Protestants, who have the economic opportunities, social linkages, and cultural resources to reconstruct identity more effectively than the marginalized working class and to benefit by so doing. Analysis of the survey data shows that support for the Agreement rises with social class and is strongest in the professional and managerial sectors. Indeed the business community, represented by the Confederation of British Industry in Northern Ireland, welcomed the Agreement and campaigned for a "yes" vote in the 1998 referendum. Individuals in the working class have also assimilated, but in much lesser proportions.89

Other Protestants have privatized. The changes brought by the Agreement have impinged strongly on their identity and they have reacted by withdrawing from politics, turning to God, family or consumption. Increasing numbers - particularly among lower social classes, and most prevalent in the skilled manual working class - say they would not vote at another referendum on the Agreement.90 The elections of 2003 saw a low turn-out and a much commented-upon public indifference. Claire Mitchell, on the basis of qualitative interviews, has noted an increasing privatization among the Protestant population.91 For some of Gladys Ganiel's interviewees, the response to the new order is a turn inwards, to a pietistic concern with God's values, rather than an assertion of these values in the political arena.92 Privatization, however, requires personal resources - familial, religious, career or material resources -and is a costly option for those whose social investments have been in communal activity and solidarity. It may be an interim phase, a period when identity change too difficult or dangerous to accomplish publicly may be tested and controlled in a safe environment.93

Another set of Protestants resisted the Agreement. About half of the skilled manual working class - traditionally the most assertive of unionists - voted against, but it is among the unskilled that support for the Agreement has fallen off most rapidly.94 This section of the population has been increasingly marginalized politically and economically and cannot avail of the new opportunities presented by the Agreement. Instead they see their jobs, status, and even their territory diminished with Catholic equality and demographic growth. In the past, they found status and meaning in identification with their immediate community in loyalist marches and rituals that are assertions at once of Protestant solidarity and of Protestant superiority.95 Some have responded to the

Agreement by reasserting the old binary distinctions between Protestant and Catholic, loyal and disloyal, righteous and unrighteous. They have protested against the Agreement, they have joined paramilitary groups, they have marched at Drumcree against government bans on their "Orange" marches, they have tried to prevent small Catholic schoolgirls walking to school in North Belfast. They have reasserted their identity in the traditional rituals that once expressed Protestant hegemony. But the new order changes the meaning of these rituals. In the past, loyalist rituals served to bind together large segments of the Protestant community; now they can no longer count on wider Protestant support. In the past, loyalist practices celebrated victory, now they are the last stand of those who have come to define themselves as victims instead of victors. The resistants are no longer reconstituting their community but are desperately asserting an older set of opposi-tional meanings that is (quite properly) in the process of being wiped

out.96

By far the largest section of Protestants, however, adapted. They accepted the Agreement pragmatically, as did the majority of the Ulster Unionist Party (UUP), or, if they rejected it, they agreed to work within its institutions, as did the Democratic Unionist Party (DUP).97 They saw this as the best way to defend their interests and did not see it as impinging on their identity and beliefs.98 The daily trials and internal fractures of the UUP show the difficulties of this option.99 These unionists, who have worked within the new institutions while holding the core of their own identity stable, have found that their values and expectations have constantly been challenged by the practical demands made upon them. They have tried to be accommodating, but they are working with conceptual categories in which the old binaries are clearly visible. Because their concept of Britishness is state centered, the "parity of esteem" of national symbols demanded by nationalists seems to them to be radically inappropriate.100 Because their concept of order and authority assumes a continuity with the past, the peace process itself (which has brought ex-paramilitaries into government and required radical reform of the security forces) cannot easily be accepted. Moreover, even where one reform package is - with great difficulty -accepted, its logic and rationale are not. The next phase of implementation of the Agreement is therefore opposed just as strongly as the last, and the UUP appears always to give at best grudging acceptance of the new institutions. It is precisely as if they are applying an inappropriate logic to the institutions of the settlement. The contradiction between

the meanings implicit in the institutions and the meanings and assumptions implicit in the identities of those who are key actors within them is, thus, permanent and has created ongoing political crises.

Catholics and nationalists had different and easier sets of choices because, for most, the Agreement represented movement to a situation where it would be easier to sustain their traditional national identification, even in its classic oppositional form. Here, too, however, there was a dynamic of change whose trajectory remains open.

Only a very few rejected and resisted the Agreement, and they (the dissident republicans) quickly yielded to superior force and moved to adaptation. Few, if any, "converted"; the Agreement newly certified Irish identity rather than giving an incentive to move away from it. Some, however, moved toward privatization. Those Catholics whose nationalism was primarily a resistance to unjust rule have responded to change by absorption in ordinary everyday activity, enjoying the peace and the new opportunities opened now that the major problems associated with British rule are being removed.

Moderate nationalists willingly assimilated. They have given up the binary opposition between British and Irish, while retaining much of the cultural substance of Irish identity. These changes had already been prefigured in nationalist ideology, and many nationalists within the SDLP found the new institutional changes confirming and certifying.101 Others, previously unconvinced, find the political changes sufficient to allow them to put aside older enmities and even to identify with the category of Britishness in some of its aspects, now that the category of Irishness is also institutionally recognized.102 This is aided by contemporaneous changes in the Irish state, where there is now radical openness to influences and inputs from Britain, Europe and the United States, where the official notion of Ireland and Irish identity is modernizing and pluralist, and where an opposition of Irish and British identities is no longer prominent, pressing or felt necessary.103 Indeed, it is a plausible hypothesis that the factors that most favor assimilation among nationalists are linkages with Great Britain, the European Union, and, perhaps paradoxically, the Irish state, whose input into Northern Ireland is now more likely to favor assimilation among nationalists than reassertion of exclusivist identities.

Yet there is another side to the Agreement that is also evident in the data. Some nationalists who accept the Agreement find that it confirms their traditional identity. Their Irishness is rising in status in the new institutional context corresponding to an equivalent drop in the status of Britishness.104 In this respect, and against the drafters' intentions, the new institutions may favor and reproduce oppositional aspects of Irish national identity. Survey results suggest that the opposition of Irish and British identity continues for about 60 percent of the Catholic population, particularly the young (most likely to declare themselves Irish, least likely to see any British aspect to their identity).105 Yet this is not simply a reassertion of older aggressive forms of nationalism. In circumstances where Irish identity no longer has to be asserted against state and official opposition, where cultural activities are now funded and "chill factors" (British flags, military installations) removed, an evident cultural "relaxation" has occurred in some areas.106 Even if the identity of this section of the population is "not at all British," some can now accept cohabitation with a British dimension.

For republicans, however, cohabitation threatens to destabilize identity. The Agreement implies acceptance of the legitimacy of the present status of Northern Ireland as part of the United Kingdom and willingness to help administer the expenditure of British money there.107 Forrepub-licans in the past, an Irish identity carried with it a strong opposition to the British state; the present situation therefore threatens to create cognitive problems parallel to those of unionists. The republican leadership could have opted for a strategy of adaptation. Instead they chose ritual appropriation. They have worked within the new institutional structure, but interpreted the meaning of their new practices in terms of older legitimating symbols and rituals. Thus, they have asserted an essential continuity between past and present, and brought with them an increasingly diverse, but predominantly young, working class constituency.108 Yet the conceptual oppositions expressed in the older symbols are coming into tension with practices that assume that the path to Irish unity lies in compromise, gradualism, and work within British institutions. Dissident republicans and liberal unionists predict that the older republican identity system will collapse (if it has not already done so), and that pro-Agreement republicans will become assimilationist nationalists.109 Pro-Agreement republicans portray their practice as that of a new modernizing republicanism that retains its radical thrust. The outcome of this mix of developing symbol system and developing practical action is not predictable; it is in principle

open-ended, allowing a movement towards reflective equilibrium between identity and practice. Where that equilibrium will lie depends, in large part, on other actors. Increasing resistance and division among unionists (not reassured by this republican strategy) allows republicans to confirm much of their older identity, while also showing their willingness to reshuffle its elements. Assimilation among unionists would remove the ambiguity from the situation and demand either adaptation or assimilation by republicans.

The changes outlined above are ongoing, worked out in practical life and interactions, conditioned by the uneven distribution of cultural and social resources, sometimes reflected upon and opened to conscious choice. In and between these changes and choices are the multiple ethnic entrepreneurs, political parties and strategic maneuverings of the political process. All of this is important: had the leadership of the proAgreement unionists been less grudging in its support for the Agreement, perhaps more unionists would have been won to assimilation from adaptation. Had pro-Agreement republicans adopted a different strategy, there would almost certainly have been a major republican split. But the one lesson that should be learned from over thirty years of conflict in Northern Ireland is that the public does not follow the admonitions of its leaders unless these grasp its worries and uncertainties and hopes. Before analyzing the impact of political discourse and competition on the public, it is necessary to map the inchoate but discernible trends of change in that public. The map proposed here is in need of development and revision as more research results emerge; it is intended, however, to give a useful framework for that research.

Conclusion

This article puts forward a Bourdieu-ian perspective that allows us to recognize the embeddedness of identity-categories, their internal complexity, their effects in framing interaction, and the possibility of both gradual and sudden change. It explains identity change in terms of three variables: existing identity structure (the content and constitutive values of the identity category and its relations with other categories), power-relations (as expressed in institutions and institutionally patterned interactions), and resources. The fact of identity change is explained when changing power relations conflict with embedded categorical understandings, expectations, and values. The direction of identity change is crucially affected by the resources available to different sectors of the

population. The article sketches a typology of six possible directions of change.

The value of this approach is illustrated in the case of Northern Ireland. The analysis above shows the post-Good-Friday situation in Northern Ireland in flux, not simply politically but also in terms of the meanings of national identity. It explains why pro-Agreement unionists have delayed the process of implementation (because of its threat to the content of their identity categories), and why pro-Agreement republicans have been unable to assuage unionist fears (because of their mode of identity shift). It shows an uneven process of reshuffling of the contents of national identities. In short, it identifies and explains subtle shifts and strains in identity-structure that were provoked by the Good Friday Agreement, and suggests how they in turn explain the difficulties of implementation of the Agreement. It predicts farther reaching change, as the costs of retaining older oppositional concepts of identity increase, as the costs of new choices become clear and as resources for further change emerge. This sketch of identity change cannot claim to be definitive. Change is on-going, and the results of relevant research are still coming in; identity structures may stabilize in tension with existing institutions, while creating endemic crises in them. Yet this perspective lets us raise the question of identity change, look at emerging evidence, and evaluate trends. In this, it is superior to approaches that fail to recognize the difficulty of identity change or fail to recognize its possibility.

The case study shows the importance of including identity and identity change in the study of socio-political transformations. Without it, we lose a key factor that differentiates groups, motivates action, and explains the emerging fission of old and fusion of new clusters of the population. It shows the value of Bourdieu's approach: the linking of habitus and power relations explains not only the entrenchment of oppositions but also the shifts in content of identity categories as habitus is disconfirmed by socio-structural and political change. It suggests that these processes of identity change follow discernible patterns, both in terms of the "logic" or "rationale" of possible responses as given in the typology, and in terms of patterns of resource distribution that predispose towards one choice or another. Indeed, the discussion suggests that trajectories of change exist, with initial choices tending to develop into others, so that, for example, one pattern of change leads from reaffirmation through ritual appropriation to assimilation, with the

timing of each phase dependent on social positioning and resources. But to pursue this line of thought is to go beyond the scope of this article.

It might be questioned whether the notion of "identity" is indeed central to the analysis. Could the whole analysis be rephrased in terms of identifications, values, assumptions, and self-categorizations? The potential price is a loss of the sense of coherence with which individuals tie together these factors and a lack of recognition of the cost to them in pulling them apart and reshuffling these elements. The benefit of this analytic perspective is to link individuals' reports of their own experience of social and self change to wider socio-theoretical explanations of change, while retaining the sense and complexity of the individual's experience.

Acknowledgments

Earlier versions of this article were presented at the Culture/Power International Conference, in Lisbon, Portugal, November 3, 2003, and at the Identity, Diversity and Citizenship workshop, Institute for the Study of Social Change, University College Dublin (UCD), March 2004. I thank participants for their comments, and Lorenzo Canas-Bottos, John Coakley, Iseult Honohan, Gladys Ganiel, Claire Mitchell, Nathalie Rougier, Joseph Ruane, and the Theory and Society Editors and reviewers for comments on a draft of the article. It has been written as part of the "Intergenerational transmission and ethno-national identity" project, funded by the EU Programme for Peace and Reconciliation, within the Identity, Diversity Citizenship programme at the Geary Institute, UCD. Herbert Preiss provided research assistance in analysis of the quantitative data.

1. Doug McAdam, Charles Tilly, and Sidney Tarrow, Dynamics of Contention: Cambridge Studies in Contentious Politics (Cambridge: Cambridge University Press, 2000), for example, 56.

2. Manuel Castells, The Power of Identity, vol. II of The Information Age (Oxford: Blackwell, 1997); M. Keating, Plurinational Democracy: Stateless Nations in a Post-Sovereignty Era (Oxford: Oxford University Press, 2001).

3. S. Kakar, The Colors of Violence, Cultural Identities, Religion and Conflict (Chicago: Chicago University Press, 1996); R. Brubaker, Nationalism Reframed: Nationhood and the National Question in the New Europe (Cambridge: Cambridge University Press, 1996).

4. "The coincidence of the changing of circumstances and of human activity or self-changing can be conceived and rationally understood only as revolutionary practice'' K. Marx and F. Engels, "Theses on Feuerbach," in C.J. Arthur, editor, The German Ideology (London: Lawrence and Wishart, 1970), 121. Short of coincidence, however, there are many types of interrelations that require analysis.

5. They expand the role ofinterpretative framing from that assumed in classic social movement theory: see, for example, the discussion of "opportunities and threats," McAdam et al., Dynamics of Contention, 44-47.

6. McAdam, Tarrow, and Tilly discuss these aspects of relationality in Dynamics of Contention, 57-58. Although they focus on the interactive process, and in particular criticize the assumption that identities exist "within people's heads" (57) they also acknowledge the other aspects, see 274ff, 310, 344. Charles Tilly tends to be more robust in dispensing altogether with the subjective and intentional aspect of individual identity: see C. Tilly, Stories, Identities and Political Change (Oxford: Rowman and Littlefield, 2002). In the present article, individuals and intentionality are brought back in, but explicitly as moments of a broader analysis that inter-relates intentionality, institutions, and interactions.

7. Rawi Abdelal, Y.M. Herrera, A.I. Hohnston, R. McDermott, "Identity as a Variable" (May 10, 2003 version), Weatherhead Initiative in International Affairs, Harvard University, www.wcfia.harvard.edu/misc/initiative/identity/papers/index.htm, 1.

8. See, for example, Donald L. Horowitz, Ethnic Groups in Conflict, 2nd edition. (Los Angeles: University of California Press, 2000), 66-67.

9. Tobias Theiler, Political Symbols and European Integration (Manchester: Manchester University Press, 2004).

10. A number of examples are discussed in J. Ruane and J. Todd, The Dynamics ofConflict in Northern Ireland: Power, Conflict and Emancipation (Cambridge: Cambridge University Press, 1996). This process of transposing new elements within older meaning-systems is exemplified in the ritual of "la sape," analyzed by J. Friedman as a case of "weak globalization"; see Friedman, "Global system, globalization and the parameters of modernity," in Mike Featherstone, Scott Lash, and Ronald Robertson, editors, Global Modernities (London: Sage, 1995), 69-90.

11. For example, they explain the peaceful character of the Spanish post-Franco transition in part by the movement towards a European identity (McAdam et al., Dynamics of Contention, 181-183), but the diversity of, for example, Europeanness, Spanishness, Catalanness, and Basqueness, and of their combinations within Spain from the 1970s to the present, which at once underlies the possibility of consensus on the 1978 constitution and the ongoing contention over it, is elided. On this diversity, see La Question de L'Espagne, Hérodote: revue de géographie et de geopolitque, no. 91, 1998. On the expression of these diverse understandings within constitutional debate, see J-M Comas, "Spain: the 1978 constitution and centre-periphery tensions," in J. Ruane, J. Todd, and A. Mandeville, editors, Europe's Old States in the New World Order: The Politics of Transition in Britain, France and Spain (Dublin: University College Dublin Press, 2003), 38-61.

12. On the multiple dimensions of any collective identity category, see Richard D. Ashmore, Kay Deaux, and Tracy McLaughlin-Volpe, "An organizing framework for collective identity: articulation and significance of multidimensionality," Psychological Bulletin 130/1 (2004): 80-114.

13. On theories ofchanging forms ofnational identity, see Keating, Plurinational Democracy; Castells, Power of Identity; J. Nederveen Pieterse, "Globalization as hybridization," in M. Featherstone, S. Lash, and R. Robertson, editors, Global Modernities (London: Sage, 1995), 45-68.

14. On the role of political entrepreneurs, see McAdam et al., Dynamics of Contention, 142ff. See also Rogers Brubaker, "Ethnicity without groups," Archives Européennes de Sociologie 42/2(2002): 163-189, 166-167.

15. The unexpected nature of entrepreneurial success is striking even to the entrepreneurs themselves: a veteran republican activist once described to me (in an interview in 1988) his

involvement in the Civil Rights Movement (CRM) in Northern Ireland in the late 1960s. He was involved in the organization of the CRM from the start, as he had been involved in every anti-state or protest movement in Northern Ireland for as long as he could remember. At the time, he saw the civil rights movement as no different, another phase and strategy of anti-state activism: he had no idea that this one was going to have such major consequences.

16. Behavior and boundaries are themselves richly meaningful, although their meaning may be multiple, ambiguous, and disguised. On the need to go beyond analysis of cultural boundaries to analysis of cultural content, see Nick Hopkins and Steve Reicher, "The construction of social categories and processes of social change: arguing about national identities," in G. M. Breakwell and E. Lyons, editors, Changing European Identities: Social Psychological analyzes of Social Change (Oxford: Butterworth-Heinemann, 1996), 69-93.

17. Cerulo, "Identity construction"; Rogers Brubaker and Frederick Cooper, "Beyond 'identity,"' Theory and Society 29/1 (2002): 1-47; Margaret Somers, "The narrative construction of identity: A relational and network approach," Theory and Society 23/5 (1994): 605-649.

18. T. Edensor, National Identity, Popular Culture and Everyday Life (Oxford: Berg, 2002).

19. Brubaker and Cooper, "Beyond 'identity'," 7, 9-11; Tilly, Stories, Identities and Political Change, 46, 49, 75; Tarrow et al., Dynamics of Contention, 56-57.

20. Brubaker and Cooper, "Beyond 'identity'."

21. Taking this notion of identity categories, Tilly rightly sees no point in looking at identity apart from interaction; see Tilly, Stories, Identities and Political Change, 49, 75. Brubaker and Cooper see the very term as dispensable, "Beyond 'identity'," 1-2.

22. I focus particularly on the following works where Bourdieu emphasizes class distinction: P. Bourdieu, Outline of a Theory of Practice, translator R. Nice (Cambridge: Cambridge University Press, 1977); P. Bourdieu, Distinction: A Social Critique of the Judgement of Taste, translator R. Nice (Cambridge: Harvard University Press, 1984); P. Bourdieu, The Logic of Practice, translator R. Nice (Cambridge: Polity, 1990). In some ofhis later articles, he analyzes collective categories more generally. See P. Bourdieu, Language and Symbolic Power, edited and introduction by John B. Thompson, trans. G. Raymond and M. Adamson (Cambridge: Polity, 1991).

23. Ashmore et al., "An organizing framework," 83ff. note at least fourteen dimensions of each collective identity category, including salience, affect, group-belonging, as well as different aspects of content, together with the further dimensions of interrelations of categories and their social contextualization. The present article deals only with the cognitive content and interrelations of collective identity categories, and their relation to the social context.

24. Bourdieu, Outline of a Theory of Practice, 78-79; Bourdieu, Logic of Practice, 58.

25. Bourdieu, Outline of a Theory of Practice, 86.

26. The phenomenological depth of, for example, ethnic identity is emphasized by D.L. Horowitz, Ethnic Groups in Conflict; Walker Connor, Ethnonationalism: The Quest for Understanding (Princeton: Princeton Univ. Press, 1994); A. D. Smith, The Ethnic Origins of Nations (Oxford: Basil Blackwell, 1986). Ethnic and national categorizations are not normally as deeply embedded in the individual as are gender categorizations, but they can come close: see Rosemary Harris, Prejudice and Tolerance in Ulster: A Study of Neighbours and 'Strangers' in a Border Community (Manchester: Manchester University Press, 1972), 148.

27. Bourdieu, Logic ofPractice, 56.

28. See Bourdieu's classic study of genderized distinction within the Kabyle house, Logic of Practice, 271-283. On class, gender, and local variations on national identity, see Raphael Samuel, editor, Patriotism: The Making and Unmaking of British National Identity, 3 vols. (London: Routledge, 1989).

29. Being male or female is not separate from being Kabyle or being French working class, but rather it forms a way of being Kabyle or French working class.

30. For a discussion of the opposition between social and individual identity, see Abdelal et al., "Identity as a Variable." The distinction between instrumental and affective modes

of identification is pervasive in discussions of ethnic identity, see J. Hutchinson and A. D. Smith, editors, Ethnicity (Oxford: Oxford University Press, 1996). On the interrelations, see Bourdieu, Outline of a Theory ofPractice, 80-82. For a discussion of the transcendence of the related material/ideal and economic/cultural dualisms in Bourdieu's work, see David L. Swartz, "Drawing information from Bourdieu's sociology of symbolic power," Theory and Society 32/5-6 (2003): 519-528, 525; and Frederic Lebaron, "Pierre Bourdieu: Economic models against economism," Theory and Society 32/5-6 (2003): 551-565, 560-563.

31. B. De Wita, French Bourgeous Culture, trans. J. A. Underwood (Cambridge: Cambridge University Press, 1994).

32. For a measured critique, see John R. Hall, "The Capital(s) of cultures: a non-holistic approach to status situations, class, gender and ethnicity," in Michele Lamont and Marcel Fournier, editors, Cultivating Differences: Symbolic Boundaries and the Making of Inequality (Chicago: U. of Chicago Press, 1992), 257-285.

33. For a useful discussion, see Nick Crossley, "The phenomenological habitus and its construction," Theory and Society 30/1 (2001): 81-120.

34. Gradual social evolution that allows historical continuity through the capacity of the habitus to apply old distinctions in new circumstances is, however, at the core of Bourdieu's theory. For discussions of this and of more radical change, see Outline of a Theory of Practice, 95; Logic of Practice, 67.

35. But see Crossley, "Phenomenological habitus," 92-93.

36. On the "indigenization" of new elements through the habitus, see Bourdieu, Outline of a Theory of Practice, 78-81. For one set of historical examples, see Ruane and Todd, Dynamics of Conflict.

37. Hall, "Capital(s) of culture," points out that the notion of "field" does not resolve the difficulty because a dominant order is presupposed within each field.

38. The very concept of "governance," for example, dispenses with the notion of a central ordering point within the state, in favor of a plurality of disparate potentially competing institutions, embodying different modes of practice and norms, and interlinked in a variety of ways. In a quite different tradition of thought, but referring to comparable phenomena, Marxists point to "contradictions" within the state, see C. Offe, Contradictions of the Welfare State, ed. John Keane (London: Hutchinson, 1984); or to multiple sites of struggle within it, see N. Poulantzas, State, Power, Socialism (London: New Left Books, 1978). See also B. Jessop, The Future of the Capitalist State (Cambridge: Polity, 2002).

39. Michele Lamont, Money, Morals and Manners, The Culture of the French and American Upper-Middle Class (Chicago: U. of Chicago Press, 1992).

40. On opposed ethno-religious symbol systems, see J. Ruane and J. Todd, Dynamics of Conflict and Transition in Northern Ireland (Cambridge: Cambridge University Press, forthcoming). For historical examples, see S. Farrell, Rituals and Riots: Sectarian Violence and Political Culture in Ulster 1784-1886 (Lexington, KY: University of Kentucky Press, 2000).

41. On the formation of skills and virtues from complex social practices, see Alasdair MacIntyre, After Virtue: A Study in Moral Theory (London: Duckworth, 1981).

42. Bourdieu, Distinction.

43. Although she does not theorize it in these terms, one might say that it is precisely these values that some of Lamont's interviewees emphasize in Money, Morals and Manners.

44. On these different aspects within national identity categories, see M. Billig, Banal Nationalism (London: Sage, 1995); Edensor, National Identity; Bourdieu, Language and Symbolic Power.

45. For different views, see Iseult Honohan, "Friends, strangers or countrymen? Citizens as colleagues," Political Studies 49/1 (2001): 51-69; D. Miller, On Nationality (Oxford: Clarendon, 1995); B. Parekh, "Ethnocentricity in the nationalist discourse," Nations and Nationalism 1/1 (1995): 25-52.

46. George Schopflin, Nations, Identity, Power: The New Politics of Europe (London: Hurst, 2000).

47. For example, whether and to what extent the meanings and values embedded in work and working relations which color one's expectations, perceptions, and responses are incorporated into identity varies dramatically depending on the satisfaction and fulfillment or tediousness and alienation associated with work. If the norms and skills of creative professional work can willingly be assumed into identity, routine work can be factored out.

48. Here the notion of the active subject or agent is brought in; for a discussion of why this is necessary and how it is (conceptually) possible, see Crossley, "Phenomenological habitus."

49. Somers, "Narrative construction of identity."

50. This list of elements parallels three of the elements of identity-content discussed in Abdelal et al., "Varieties of identity" (they mention constitutive norms, relational comparisons, and cognitive models as well as social purposes), although I give a rather different reading of each.

51. This was precisely Bourdieu's achievement in Distinction. Such mapping requires a triangulation of methods: theoretical hypotheses drawn from existing research and comparative work, refined through individual interviews and tested by a range of methods, from surveys to participant observation. See Abdelal et al., "Varieties of identity," for discussion of some relevant methods.

52. The variants of Irish identity are, for example, satirized in television situation comedies and represented in television drama, associated with different cultural styles, party political loyalties, and with identification with different aspects of the Irish national tradition.

53. See J. Ruane and J. Todd, "Old and new Irish nationalisms: reading the Belfast Agreement," 121-145, in Ruane et al., editors, Europe's Old States.

54. On the ontological role of narrative, see Somers, "Narrative construction of identity," e.g. 618-619,629-631.

55. C. Vicencio and W. Verwoerd, editors, Looking Back, Reaching Forward: Reflections on the South African Truth and Reconciliation Commission (London: Zed Books, 2000).

56. This type of process was evidenced in (unpublished) semi-structured interviews (80 in all) conducted in Northern Ireland in 1988 by the author and a colleague: accessing national identity was one aim of the interviews. There was much individual variation in the contents and values of the same national categories: significant numbers reported periods (sometimes the bloody 1970s, sometimes traumatic social events like the republican hunger-strikes of 1981, sometimes individual experiences of injustice) when they had felt it necessary to rethink their sense of nationality.

57. On certification and decertification, see McAdam et al., Dynamics of Contention, 145-148, 204-207.

58. Marc Auge shows the changing meanings of symbolism when actors face the threat of cultural extinction. M. Auge, The War of Dreams, translator, L. Heron (London: Pluto, 1999), 19. Terence McCaughey has pointed to the changing character of the Irish language now that it is "conserved" rather than a living encounter with the world. Terence McCaughey, "Religion and reconciliation in Irish politics," article presented to the conference "Old Structures, New Beliefs: Religion, Community and Politics in Contemporary Ireland," Institute for British-Irish Studies, University College Dublin, 15 May, 2003.

59. On individual processes of change, see Vicencio and Verwoerd, Looking Back, Reaching Forward. For a classic discussion of such a collapsed identity system (which has itself become subject of historical contention), see M. Sahlins, Historical Metaphors and Mythical Realities: Structure in the Early History of the Sandwich Islands Kingdom (Ann Arbor, University of Michigan Press, 1981).

60. Josep R. Llobera, "The role of commemorations in (ethno)nation-building: the case of Catalonia" in C. Mar-Molinero and A. Smith, editors, Nationalism and the Nation in the Iberian Peninsula: Competing and Conflicting Identities (Oxford: Berg, 1996), 191-206; 193, notes that the family was a prime site of reproduction of Catalan national identity during the Franco years.

61. Llobera, "Commemorations."

62. In one recent autobiography, this distancing of practice and belief has been called "the silent no." H. Hamilton, The Speckled People (London: Fourth Estate, 2003). On the linguistic analogy, see Bourdieu, Logic of Practice, 67.

63. This process of using old linkages to act effectively in a new social order has been much more discussed in the globalization literature than in the literature on identity: see, for example, S. Lash and J. Urry, Economies of Signs and Space (London: Sage, 1994), chapter 4. For an overview of the literature on different responses to reunification in the old East Germany, see Jan Werner Muller, "East Germany: Incorporation, Tainted Truth and the Double Division," in Alexandra Barahona de Brito, Carmen Gonzale-Enriquez, Paloma Aguilar, editors, The Politics of Memory: Transitional Justice in Democratizing Societies (Oxford: Oxford University Press, 2001), 248-274, 270-274.

64. Some examples of the ways (national) continuity is asserted in times of change are to be found in Wright's study of facets of everyday life and symbolism in Thatcher's Britain. P. Wright, On Living in an Old Country: The National Past in Contemporary Britain (London: Verso, 1995). Of course, the "oldness" of the ritual in terms of which the new is integrated may itself be radically exaggerated in the ritual itself. See D. Cannadine, "The context, performance and meaning of ritual: the British monarchy and the 'invention of tradition,' c. 1820-1977," in E. Hobsbawm and T. Ranger, editors, The Invention of Tradition (Cambridge: Cambridge University Press, 1983): 101-164.

65. See E. Frazer, "Citizenship and culture," in P. Dunleavy, A. Gamble, I. Holliday, and G. Peele, editors, Developments in British Politics 6 (London: Macmillan, 2000), 203-218; T. Nairn, The Enchanted Glass: Britain and Its Monarchy (London: Hutchinson Radius, 1988).

66. I rely here on a range of published studies and research reports of contemporary change in Northern Ireland, referred to below, supplemented by interviews conducted in the late 1980s and after by the author and a colleague and results to date of interviews conducted by the author for the "Intergenerational transmission" research project at the Geary Institute, University College Dublin. Much research has begun since 1998, not all ofwhich is yet completed or published so the analysis here necessarily remains provisional. I thank Joseph Ruane, Claire Mitchell, and Gladys Ganiel for discussions about their findings.

67. For example, ex-loyalist paramilitaries in the new loyalist parties gave very public testimony of change in the 1990s, and, to a man, dated their changing conceptions to their experiences in prison in previous decades: prison changes loyalists because it pits their self-conception (as defenders of the state) against the palpable reality of power relations (the evident fact that the state is punishing them for their activity). See J. Todd, "Two traditions in unionist political culture," Irish Political Studies 2 (1987): 1-26, 20-21.

68. On fair employment policy, see C. McCrudden, "Equality and the Good Friday Agreement" in J. Ruane and J. Todd, editors, After the Good Friday Agreement: Analysing Political Change in Northern Ireland (Dublin: University College Dublin Press, 1999), 96-121. Changes instituted after 1998 also include reform of the policing system, far-reaching guarantees of human rights, and "parity of esteem" to be formulated by a new Northern Ireland Human Rights Commission, reform of the criminal justice system, and support for the Irish and Ulster Scots languages.

69. J. Whyte, "How much discrimination was there under the Unionist regime?" in Tom Gallagher and James O'Connell, editors, Contemporary Irish Studies (Manchester: Manchester University Press, 1983), 1-35; Ruane and Todd, Dynamics of Conflict in Northern Ireland, chapters 5-7; Fionnuala ni Aoilain, The Politics of Force: Conflict Management and State Violence in Northern Ireland (Belfast: Blackstaff, 2000); Todd, "Two traditions"; D. Kennedy, The Widening Gulf: Northern Attitudes to the Independent Irish State 1919-1949 (Belfast: Blackstaff, 1988). This is a classic Bourdieu-ian situation, where unionists' position within the power structure supports self-understandings and dispositions, which in turn create an expectation and legitimation of this position.

70. P. Bew, P. Gibbon, and H. Patterson, The State in Northern Ireland: 1921-71: Political Forces and Social Classes (Manchester: Manchester University Press, 1979), chapter

3. For an indication of sub-groups within the Protestant population, see Susan McKay, Northern Protestants: An Unsettled People (Belfast: Blackstaff, 2000); J. Coakley, editor, Changing Shades of Orange and Green: Redefining the Union and the Nation in Contemporary Ireland (Dublin: University College Dublin Press, 2002); H. Patterson, "Missing the point: academic analyzes of the 'new unionism,"' article presented to the Identity, Diversity and Citizenship conference, Institute for the Study of Social Change, University College Dublin, October 6, 2004.

71. On the binary oppositions, see Ruane and Todd, Dynamics of Conflict and Transition, chapter 3. For a clear statement of the liberal unionist position, see Arthur Aughey, Under Siege: Ulster Unionism and the Anglo-Irish Agreement (Belfast: Blackstaff, 1989).

72. On the diversity, see F. O'Connor, In Search of a State: Catholics in Northern Ireland (Belfast: Blackstaff, 1993). On the continuing binaries, see Ruane and Todd, Dynamics of Conflict in Northern Ireland, chapter 4, and Dynamics ofConflict and Transition, chapter 3.

73. J. Ruane, "Contemporary republicanism and the strategy of armed struggle," in M.J. Bric and J. Coakley, editors, From Political Violence to Negotiated Settlement. (Dublin: University College Dublin Press, 2004), 115-132; J. Todd, "Nationalism, republicanism and the Good Friday Agreement," in Ruane and Todd, editors, After the Good Friday Agreement, 49-70; Coakley, editor, Changing Shades of Orange and Green; A. McIntyre, "Modern Irish republicanism and the Belfast Agreement: Chickens coming home to roost or turkeys celebrating Christmas?" in R. Wilford, editor, Aspects of the Belfast Agreement (Oxford: Oxford University Press, 2001), 202-222.

74. See, variously B. O'Leary, "The Nature of the British-Irish Agreement," New Left Review 233 (1999): 66-96; J. Ruane and J. Todd, "The Politics of Transition: Explaining the Crises in the implementation of the Belfast Agreement," Political Studies 49 (2001): 923-940; Ruane and Todd, Dynamics of Conflict and Transition; M. Cox, A. Guelke, and F. Stephen, editors, A Farewell to Arms: From 'long war' to long peace in Northern Ireland (Manchester, Manchester University Press, 2000); Ruane and Todd, editors, After the Good Friday Agreement; J. McGarry and B. O'Leary, The Northern Ireland Conflict: Consociational Engagements (Oxford: Oxford University Press, 2004); and R. Wilford, editor, Aspects of the Belfast Agreement.

75. The point being made here is not that unionists mistrusted Blair or his cabinet or their policies - on some accounts, David Trimble trusted Blair - but that the changes being instituted in Great Britain itself, and in particular the "rebranding" of Britain and Britishness, went against the assumptions about territory, morality, and politics embedded in the unionist habitus: see, for example, Eric Waugh, "What Cheriegate means to Ulster," Belfast Telegraph, 18/12/02.

76. Life and Times surveys, however, show a steady decrease in the percentage of Protestants who supported the Agreement, declining from just over a half in 1998 to less than a third in 2003. Significantly it was the unskilled working class who found the new regime hardest to cope with and among whom support fell off most steeply (from 50 percent who reported in 1999 that they had voted yes in the referendum the previous year, to 15.4 percent who would vote "yes" in 2003). (Life and Times survey results and data are available on the webpage <http://www.ark.ac.uk/nilt/ visited October 2004). This division also existed at the party political level, between the main unionist parties and within the predominantly pro-Agreement Ulster Unionist Party.

77. In a trenchant and incisive defense of the 1998 settlement, John McGarry and Brendan O'Leary counterpose the "consociationist" argument that national identities in Northern Ireland are "durable" to the "integrationist" view that they are transformable, Conso-ciational Engagements, 16-24, 32. For an argument that such identifications may be understood at once as durable and transformable, see J. Ruane and J. Todd, "The roots of intense ethnic conflict may not themselves be ethnic: categories, communities and path dependence," Archives Européennes de Sociologie 45/2 (2004).

78. The biggest single change in self-reported identity occurred among Protestants, over a quarter of whom switched to a British identity between 1968 and 1978. An explanation

of this lies beyond the scope of this article, but it is relevant to note that the criteria used by Protestants may themselves have changed due to the political context - by 1976 they focused on political loyalty rather than on wider cultural issues. For discussion, see E. Moxon-Browne, Nation, Class and Creed in Northern Ireland (Aldershot: Gower, 1983), 6-11. Gender makes no significant difference. In the late 1980s, Catholic identification as Northern Irish was more likely among the higher social classes, but this class differentiation is no longer stark, perhaps reflecting the fact that it is now socially acceptable (rather than politically "extreme") to claim Irish identity. See E. Moxon-Browne, "National identity in Northern Ireland," in P. Stringer and G. Robinson, editors, Social Attitudes in Northern Ireland (Belfast: Blackstaff, 1991), 23-30; Life and Times2003, http://www.ark.ac.uk/nilt/.

79. Young Life and Times survey, 2003. See http://www.ark.ac.uk/ylt/. Age is a significant variable also among adults, particularly evident in the recent period: self-identification as British decreases as one moves down age-categories, self-identification as Irish and as Northern Irish rises as one moves to the younger age-categories, even when one controls for the greater number of Catholics in the younger age-groups.

80. Sources: R. Rose, Governing Without Consensus: An Irish Perspective (London: Faber and Faber, 1971); Moxon-Browne, Nation, Class and Creed in Northern Ireland; D. J. Smith, Equality and Inequality in Northern Ireland (London: Policy Studies Institute, 1987); K. Trew, "National Identity" in R. Breen, P. Devine, and L. Dowds, editors, Social Attitudes in Northern Ireland: The Fifth Report (Belfast, Appletree, 1996), 140-52; Life and Times, (Northern Ireland) 1997-2002. Life and Times Survey, website, (community relations module, identity) <http://http://www.ark.ac.uk/nilt/.

81. This is a relatively small minority. No more than a third of Protestants continued to support the Agreement in 2003, and some of these fall into the category of "adaptation" rather than assimilation. See note 76.

82. Claire Mitchell, "Protestant identification and political change in Northern Ireland," Ethnic and Racial Studies 26/4 (2001): 612-631, 620-621.

83. From the New Ulster Movement of the late 1960s through the Alliance Party to new ecumenical groups and cross-community parties, there has been a (minority) tradition of Protestants and unionists who took this stance.

84. Aughey, "Learning from the Leopard," in Wilford, editor, Aspects of the Belfast Agreement, 184-201, 193. A. Aughey, "A new beginning: the prospects for a politics of civility in Northern Ireland," in Ruane and Todd, editors, After the Good Friday Agreement, 122-144.

85. These are what Henry Patterson calls the "structural unionists," who are willing to accept change if the structure of the union is secured. H. Patterson, "Missing the point," H. Patterson, "The limits of the 'new unionism': David Trimble and the Ulster Unionist Party," Eire-Ireland 39/1&2 (2004): 163-188. One unionist voter exemplified this approach when she said to the author after the first act of republican decommissioning in 2001, "they [republicans] have done what we asked for, now let's get on with it."

86. Gladys Ganiel, "Conserving or changing? The theology and politics of Northern Irish fundamentalist and evangelical protestants after the Good Friday agreement," IBIS working paper no. 20, Institute of British Irish Studies, University College Dublin.

87. Involvement in cross-community civil society organizations is an important factor in moderating conflict in other societies, see A. Varshney, "Ethnic conflict and civil society: India and beyond," World Politics 53 (2001): 362-398. This may be because they give resources for identity change. For example, one Protestant woman interviewed by the author in 2004 recounted that the crisis-provoking event that broke her own oppositional categories was tied to her evening classes in an adult education center: the (personal) crisis arose with the onset of negotiations towards the Agreement, when she learned from the media that her much-liked and respected language teacher was also a high-ranking republican. The resource gained from the civil society organization in this case was the strength of her personal intuitions, which she was able to reconcile with her democratic and pacifist principles only after she had distinguished the latter from her specific political reactions (that republicanism was simply evil).

88. Although occasionally one meets individuals who have switched national category and content, the only significant section of the population to have done so in the recent period is the segment of Protestant youth who have adopted a Northern Irish identity. For analyses, see K. Trew, "The Northern Irish Identity," in A. Kershan, editor, A Question ofIdentity (Aldershot: Ashgate, 1998), 60-76; L. Dowds and P. Devine, "Unleashing the apathy of a lost generation?" in L. Dowds, P. Devine, and R. Breen, editors, Social Attitudes: the Sixth Report (Belfast: Appletree, 1997), 47-69.

89. P. Shirlow, "'Who fears to speak,"' Fear, mobility and ethno-sectarianism in the two 'Ardoynes,"' Global Review of Ethno-Politics 3/1 (2003): 76-91. Edwards and Bloomer report the arguments of assimilating loyalists that the socio-political resources necessary to win other working class loyalists to this position are not available. Aaron Edwards and Stephen Bloomer, "A watching brief? The political strategy of progessive loyalism since 1994," Conflict Transformation Papers no. 8 (Belfast: LINC Resource Centre, 2004).

90. See Life and Times data, http://www.ark.ac.uk/nilt/.

91. Mitchell, "Protestant Identification."

92. G. Ganiel, "Evangelical Political Identity in Transition: mapping the intersections of religion, politics, and change in post Belfast Agreement Northern Ireland." ISSC working paper 04/01. (Dublin: University College Dublin, Institute forthe Study of Social Change, 2004).

93. The trajectory of future change may well depend on the type of civil society institutions in which these individuals are involved: it is well known, for example, that some sports in Northern Ireland are intra-community, some inter-community, some all-Ireland in organization, with values differing accordingly See J. Sugden and A. Bairner, Sport, Sectarianism and Society in a Divided Ireland (Leicester: Leicester University Press, 1993)

94. See Life and Times data, http://www.ark.ac.uk/nilt/.

95. Desmond Bell, Acts of Union: Youth Culture and Sectarianism in Northern Ireland (London: Macmillan, 1990).

96. P. Shirlow and I. Shuttleworth, "'Who is going to toss the burgers?' Social class and the reconstruction of the Northern Irish economy," Capital and Class 69 (1999): 27-46; Shirlow, "Who fears to speak"; McKay, Northern Protestants; Bell, Acts of Union. See Dominic Bryan, Orange Parades: The Politics of Ritual, Tradition and Control (London: Pluto, 2000), for the tensions between rough and respectable within loyalism. Some community activists and political entrepreneurs are highly pessimistic about the consequences if this position is defeated without resources available for endogenous change. See Edwards and Bloomer, "A watching brief?"

97. The DUP quickly moved beyond resistance, but swayed between adaptation and ritual appropriation: on occasion their discourse was against the agreement while they served as ministers within the new executive.

98. This is the category that Patterson calls "conservative realists." See Patterson, "Missing the point," and "The limits of the 'new unionism'."

99. For a narrative account of the continuing crises within the Ulster Unionist Party, the constant challenges to the pro-Agreement leadership, that party's delays in joining the institutions set up by the Agreement, and its eventual defeat in the 2003 elections by the Democratic Unionist Party, see Northern Ireland monitoring reports (Nations and Regions: Dynamics of Devolution. Constitution Unit website http://www.ucl.ac.uk).

100. This is shown in numerous articles by liberal unionist columnists in the pro-Agreement news-paper, The Belfast Telegraph. See for example, Eric Waugh's column, Belfast Telegraph, 04/07/01.

101. They had been arguing for them for twenty-five years. On nationalist ideology, see J. Todd, "The reorientation of constitutional nationalism" in Coakley, editor, Changing Shades of Orange and Green, 71-83.

102. About a third of Catholics have a weak sense of British identity, which they combine with a sense of Irish identity: the rest either have a strong sense of Britishness and no sense of Irishness (about 10 percent), or a strong sense of Irishness and no sense of Britishness at

all. Just less than half of Protestants have some sense of Irish identity. (Life and Times, 1999: http://www.ark.ac.uk/nilt/)

103. See, for examples, articles in W. Crotty and D. E. Schmitt, editors, Ireland and the Politics of Change (London: Longman, 1998). In 2003-2004, Theresa O'Keefe conducted over 60 interviews on Changing Irish Identities for the Identity, Diversity and Citizenship program in ISSC, UCD; the individuals whom she interviewed do not, in general, fore-ground their Irish identity nor do they explicitly oppose it to British identity.

104. By 2001, more Catholics than Protestants felt that their tradition was protected and more Protestants than Catholics felt that they were underdogs, a marked reversal of previous perceptions (Life and Times surveys, http://www.ark.ac.uk/nilt/).

105. Sixty-two percent of Catholics in 1999 felt that they were "not at all" British (compared to 51 percent of Protestants who felt that they were "not at all" Irish), the figure in 1998, the only other time the question was asked was 59 percent (http://www.ark.ac.uk/nilt/). Shirlow's interviews in one marginal working class area of Belfast suggest strong opposi-tionalism there, see Shirlow, "Who fears to speak." However, there are no direct correlations between Irish identity and unskilled class status. There is some interview evidence of a reaffirmation of Irish identity in majority Catholic rural areas that had experienced security force harassment, but I have not found direct survey evidence of regional distinctions.

106. This is how a border Protestant interviewee reported her experience of change in Catholics' interactions with her family since the Agreement (interview with the author, 2004). It is not the case among those in sectarian interface areas. See Shirlow, "Who fears to speak."

107. For a clear statement of the contradictions, see Ruane, "Contemporary republicanism"; and McIntyre, "Modern Irish republicanism."

108. Support for Sinn Fein has grown steadily, so that in the 2003 Northern Ireland elections they became the largest nationalist party in Northern Ireland.

109. Paul Bew, "The unionists have won, they just don't know it," Sunday Times, 17/5/98. McIntyre, "Modern Irish republicanism."