Ethnic and Racial Studies Vol. 28 No. 1 January 2005 pp. 1-19 The 'diaspora' diaspora Rogers Brubaker Abstract As the use of 'diaspora' has proliferated in the last decade, its meaning has been stretched in various directions. This article traces the dispersion of the term in semantic, conceptual and disciplinary space; analyses three core elements that continue to be understood as constitutive of diaspora; assesses claims made by theorists of diaspora about a radical shift in perspective and a fundamental change in the social world; and proposes to treat diaspora not as a bounded entity but as an idiom, stance and claim. Keywords: Diaspora; migration; ethnicity; nation-state. Proliferation Just fourteen years ago, writing in the inaugural issue of the journal Diaspora, William Safran observed that most scholarly discussions of ethnicity and immigration paid 'little if any attention ... to diasporas' (1991, p. 83). This claim was beginning to be out of date - as Safran recognized - even by the time it appeared in print. And obviously no one would think of making such a claim today. There has been a veritable explosion of interest in diasporas since the late 1980s. 'Diaspora' and its cognates appear as keywords only once or twice a year in dissertations from the 1970s, about thirteen times a year in the late 1980s, and nearly 130 times in 2001 alone.1 And the diaspora explosion is not confined to academic writing. 'Diaspora' yields a million Google hits; a sampling suggests that the large majority are not academic. As the term has proliferated, its meaning has been stretched to accommodate the various intellectual, cultural and political agendas in the service of which it has been enlisted. This has resulted in what one might call a '"diaspora" diaspora' - a dispersion of the meanings of the term in semantic, conceptual and disciplinary space. © 2005 Taylor & Francis Ltd ISSN 0141-9870 print/1466-4356 online DOI: 10.1080/0141987042000289997 II Routledge jjj^^ Taylors Francis Group 2 Rogers Brubaker Most early discussions of diaspora were firmly rooted in a conceptual 'homeland'; they were concerned with a paradigmatic case, or a small number of core cases. The paradigmatic case was, of course, the Jewish diaspora; some dictionary definitions of diaspora, until recently, did not simply illustrate but defined the word with reference to that case (Sheffer 2003, p. 9). As discussions of diasporas began to branch out to include other cases, they remained oriented, at least initially, to this conceptual homeland - to the Jewish case and the other 'classical' diasporas, Armenian and Greek. When historian George Shepperson introduced the notion of the African diaspora, for example, he did so by expressly engaging the Jewish experience (Shepperson 1966; Alpers 2001; Edwards 2001). The Palestinian diaspora, too, has been construed as a 'catastrophic' diaspora - or in Cohen's (1997) term, a 'victim diaspora' - on the model of the Jewish case. The concept of the trading diaspora - or in John Armstrong's (1976) terms, the 'mobilized diaspora' - was constructed on the model of another aspect of the Jewish, as well as the Greek and Armenian, experience. Chinese, Indians, Lebanese, Baltic Germans and the Hausa of Nigeria are among those often mentioned as trading diasporas.2 An orientation to these paradigmatic cases informs some influential recent discussions as well, including those of Safran (1991), Clifford (1994), and Cohen (1997). As Clifford put it, 'we should be able to recognize the strong entailment of Jewish history on the language of diaspora without making that history a definitive model. Jewish (and Greek and Armenian) diasporas can be taken as non-normative starting points for a discourse that is travelling or hybridizing in new global conditions' (1994, p. 306). In several recent extensions of the term, however, the reference to the paradigmatic cases has become more attenuated. Some emigrant groups - characterized as 'long-distance nationalists' by Anderson (1998) - have been construed as diasporas because of their continued involvement in homeland politics, sometimes involving the support of terrorist or ultra-nationalist movements (Sheffer 1986, 2003; Angoustures and Pascal 1996; Bhatt and Mukta 2000). Albanians, Hindu Indians, Irish, Kashmiri, Kurds, Palestinians, Tamils and others have been construed as diasporas in this way. In a further extension, the term has come to embrace labour migrants who maintain (to some degree) emotional and social ties with a homeland. Algerian, Bangladeshi, Filipino, Greek, Haitian, Indian, Italian, Korean, Mexican, Pakistani, Puerto Rican, Polish, Salvadoran, Turkish, Vietnamese and many other migrant populations have been conceptualized as diasporas in this sense (Sheffer 2003). In other cases, the reference to the conceptual homeland - to the 'classical' diasporas - has become more attenuated still, to the point of The 'diaspora' diaspora 3 being lost altogether. Transethnic and transborder linguistic categories, for example - such as Francophone, Anglophone and Lusophone 'communities' (a word that should be used only in inverted commas [Baumann 1996]) - have been conceptualized as diasporas. So, too, have global religious 'communities', including Hindu, Sikh, Buddhist, Confucian, Huguenot, Muslim and Catholic diasporas.3 It appears to be little more than sheer dispersion that underwrites the formulation of such populations as 'diasporas'. From the point of view of the homeland, emigrant groups have been conceptualized as diasporas, even when they have been largely assimilated. The Italian diaspora is a case in point.4 In yet a further extension, diasporas have been seen to result from the migration of borders over people, and not simply from that of people over borders. Hungarians, Russians and other ethnonational communities separated by a political frontier from their putative national homelands have been conceptualized as diasporas in this manner.5 This still leaves a very large residual set of putative ethnocultural or country-defined diasporas. The academic literature includes references to Belarusian, Brazilian, Cambodian, Colombian, Egyptian, English, Estonian, Ethiopian, Gypsy, Hawaiian, Igbo, Iranian, Iraqi, Japanese, Javanese, Kazakh, Latvian, Lithuanian, Mayan, Polish, Romanian, Scottish, Senegalese, Somali, Soviet, Sudanese, Syrian, Tutsi and Ukrainian diasporas.6 And then there are putative diasporas of other sorts: the dixie diaspora, the yankee diaspora, the white diaspora, the liberal diaspora, the conservative diaspora, the gay diaspora, the deaf diaspora, the queer diaspora, the redneck diaspora, the digital diaspora, the fundamentalist diaspora and the terrorist diaspora.7 One dimension of dispersion, then, involves the application of the term diaspora to an ever-broadening set of cases: essentially to any and every nameable population category that is to some extent dispersed in space.8 'The term that once described Jewish, Greek and Armenian dispersion now shares meanings with a larger semantic domain that includes words like immigrant, expatriate, refugee, guest-worker, exile community, overseas community, ethnic community' (Tololyan 1991, p. 4). The problem with this latitudinarian, 'let-a-thousand-diasporas-bloom' approach is that the category becomes stretched to the point of uselessness (Sartori 1970). If everyone is diasporic, then no one is distinctively so. The term loses its discriminating power - its ability to pick out phenomena, to make distinctions. The universalization of diaspora, paradoxically, means the disappearance of diaspora. The literature of the 1990s argued effectively that there is no reason to privilege the Jewish experience, not least because that experience is internally complex, ambivalent and by no means straightforwardly 4 Rogers Brubaker 'diasporic' in the strict sense of the term.9 But there is no reason to speak of the diasporization of every more or less dispersed population. Even the editor of the journal Diaspora, a key vehicle for the proliferation of academic diaspora talk, noted in the journal's sixth year that diaspora 'is in danger of becoming a promiscuously capacious category', and argued for at least some 'stringency of definition' (Tololyan 1996, pp. 8, 30). Besides the nomination of new candidates for diaspora status, the 'diaspora' diaspora also involves a dispersion in disciplinary and social space. Within the academy, the term is now used throughout the humanities and social sciences. A sampling of forty recent dissertations on diaspora showed that they were distributed among forty-five different fields and subfields, ranging from various subfields of history, literature, anthropology and sociology through Black studies, women's studies, religion, philosophy, communications, folklore and education, to art history, cinema, dance, music and theatre. As Tololyan (1996, p. 27) has observed, the 'theory-driven revolution in the humanities' has been central to this disciplinary (and trans-disciplinary) dispersion.10 Dispersion has been even more striking outside the academy: in the media,11 on the web, in the self-representations of a wide range of groups and initiatives. In this respect the trajectory of 'diaspora' resembles that of 'identity', which moved from being a technical term of philosophy and psychoanalysis to a key term throughout the humanities and social sciences, and which came to be widely used in the media and popular culture (Gleason 1983, Brubaker and Cooper 2000). The 'diaspora' diaspora involves not only a proliferation of putative diasporas, and a diffusion of diaspora talk throughout the academy and into the wider culture and polity, but also a proliferation of terms. In addition to the concrete noun, 'diaspora', designating a collectivity, there are abstract nouns designating a condition (diaspori-city or diasporism), a process (diasporization, de-diasporization and re-diasporization), even a field of inquiry (diasporology or diaspor-istics). There is the adjective 'diasporist', designating a stance or position in a field of debate or struggle.12 And there are the adjectives 'diasporic' and 'diasporan', which designate an attribute or modality -as in diasporic citizenship, diasporic consciousness, diasporic identity, diasporic imagination, diasporic nationalism, diasporic networks, diasporic culture, diasporic religion, or even the diasporic self (to enumerate only some of the most common conceptual pairings found in recent academic articles).13 The 'diaspora' diaspora 5 Criteria Notwithstanding the dispersion in semantic and conceptual space, one can identify three core elements that remain widely understood to be constitutive of diaspora. Some subset or combination of these, variously weighted, underlies most definitions and discussions of the phenomenon.14 The first is dispersion in space; the second, orientation to a 'homeland'; and the third, boundary-maintenance. Consideration of the changing significance accorded these elements - and of the various ways in which they have been interpreted - provides leverage for a more analytical appraisal of the 'diaspora' diaspora. (1) Dispersion. This is today the most widely accepted criterion, and also the most straightforward. It can be interpreted strictly as forced or otherwise traumatic dispersion;15 more broadly as any kind of dispersion in space, provided that the dispersion crosses state borders; or (in the increasingly common metaphorical extensions of the term), more broadly still, so that dispersion within state borders may suffice. Although dispersion is widely accepted as a criterion of diaspora, it is not universally accepted. Some substitute division for dispersion, defining diasporas as 'ethnic communities divided by state frontiers' (King and Melvin 1999; see also King and Melvin 1998) or as 'that segment of a people living outside the homeland' (Connor 1986, p. 16). This allows even compactly settled populations to count as diasporas when part of the population lives as a minority outside its ethnonational 'homeland'. (2) Homeland Orientation. The second constitutive criterion is the orientation to a real or imagined 'homeland' as an authoritative source of value, identity and loyalty. Here a significant shift can be discerned in recent discussions. Earlier discussions strongly emphasized this criterion. Four of the six criteria specified by Safran (1991), for example, concern the orientation to a homeland.16 These include, first, maintaining a collective memory or myth about the homeland; second, 'regarding the ancestral homeland as the true, ideal home and as the place to which one would (or should) eventually return'; third, being collectively 'committed to the maintenance or restoration of the homeland and to its safety and prosperity'; and fourth, 'continuing] to relate, personally or vicariously', to the homeland, in a way that significantly shapes one's identity and solidarity (Safran 1991, pp. 83-84). Several more recent discussions de-emphasize homeland orientation (Clifford 1994; Anthias 1998; Falzon 2003). Clifford, for example, has criticized what he called the 'centered' model of Safran and others, in which diasporas are by definition 'oriented by continuous cultural connections to a [single] source and by a teleology of "return"'. On this strict definition, as Clifford notes, many aspects of the Jewish 6 Rogers Brubaker experience itself do not qualify. Nor would the experience of dispersed African, Caribbean, or South Asian populations; the South Asian diaspora, for example, is 'not so much oriented to roots in a specific place and a desire for return as around an ability to recreate a culture in diverse locations'. For Clifford, 'decentered, lateral connections may be as important as those formed around a teleology of origin/return' (1994, pp. 305-306). (3) Boundary-Maintenance. The third constitutive criterion is what, following Armstrong (1976, pp. 394-7), I will call boundary-maintenance, involving the preservation of a distinctive identity vis-a-vis a host society (or societies). Armstrong invokes Barth's seminal 1969 contribution to emphasize the importance of boundaries for collectivities that do not have 'their own' territorial polity: Clearly, a diaspora is something more than, say, a collection of persons distinguished by some secondary characteristic such as, for example, all persons with Scottish names in Wisconsin ... [T]he mobilized diaspora ... has often constituted for centuries a separate society or quasi-society in a larger polity (1976, pp. 393-4). Boundaries can be maintained by deliberate resistance to assimilation through self-enforced endogamy or other forms of self-segregation (Armstrong 1976, pp. 394-5; Smith 1986) or as an unintended consequence of social exclusion (Laitin 1995). On most accounts, boundary-maintenance is an indispensable criterion of diaspora (e.g. Armstrong 1976; Safran 1991, p. 83; Tololyan 1996, p. 14; Cohen 1997, p. 24). It is this that enables one to speak of a diaspora as a distinctive 'community', held together by a distinctive, active solidarity, as well as by relatively dense social relationships, that cut across state boundaries and link members of the diaspora in different states into a single 'transnational community'. Yet here there is an interesting ambivalence in the literature. Although boundary-maintenance and the preservation of identity are ordinarily emphasized, a strong counter-current emphasizes hybridity, fluidity, creolization and syncretism. In an oft-quoted remark by Stuart Hall, the 'diaspora experience ... is defined, not by essence or purity, but by the recognition of a necessary heterogeneity and diversity; by a conception of 'identity' which lives with and through, not despite, difference; by hybridity' (Hall 1990, p. 235, italics original). This counter-current is especially characteristic of the literature on transnationalism, which has tended to fuse in recent years with the literature on diaspora. There is thus a tension in the literature between boundary-maintenance and boundary-erosion. The tension is only occasionally acknowledged and then sometimes only implicitly: in his discussion of Gilroy, for example, Clifford resorts to The 'diaspora' diaspora 7 oxymoron, referring to the problem of the 'changing same', to 'something endlessly hybridized and in process but persistently there' (1994, p. 320). A final point about the boundary-maintenance criterion is that it must occur over an extended time. This is seldom made explicit, but it is crucial. The erosion of boundaries through assimilation is always a temporally extended, inter-generational process (Alba and Nee 1997, 2003; Brubaker 2001). As a result, boundary maintenance only becomes sociologically interesting, as it were, when it persists over generations. That migrants themselves maintain boundaries is only to be expected; the interesting question, and the question relevant to the existence of a diaspora, is to what extent and in what forms boundaries are maintained by second, third and subsequent generations. Classical diasporas - as Armstrong (1976, 1982, pp. 206-13) and others have emphasized - were a phenomenon of the longue duree. Whether the various instant diasporas being nominated into existence today will have this kind of multigenerational staying power is by no means clear.17 A radical break? Having reviewed the changing meanings of 'diaspora', and pointed out differences and tensions in the way core constitutive elements are understood, I want now to raise some more general questions, and to make some analytical suggestions. What are we to make of the proliferation of diasporas and of diaspora-talk, inside and outside the academy? And how should one interpret it? Are we seeing a proliferation of diasporas in the world, or perhaps even the dawning of an age of diaspora? Or, on the other hand (to put the question in deliberately exaggerated form), are we seeing simply a proliferation of diaspora talk, a change in idiom rather than in the world? I want to consider two sorts of claims about novelty and discontinuity. One concerns a putatively sharp break in ways of looking at the world, the other a putatively sharp transformation in the world itself. The two claims are, of course, closely related and are usually advanced together: the radical shift in perspective is presented as a way of coming to terms, analytically and politically, with fundamental changes in the world. On the one hand, the literature on diaspora claims to mark a sharp shift in perspective. The old perspective - it is suggested - was immigrationist, assimilationist, (methodologically) nationalist,18 and teleological. It took nation-states as units of analysis and assumed that immigrants made a sharp and definitive break with their homelands, that migration trajectories were unidirectional, and that migration 8 Rogers Brubaker inexorably led to assimilation. The new perspective does not make these assumptions. It is said to 'transcend' the old assimilationist, immigrationist paradigm. In one representative statement: it is no longer assumed that immigrants make a sharp break from their homelands. Rather, premigration networks, cultures, and capital remain salient. The sojourn itself is neither unidirectional nor final.... [M]ovements ... follow multifarious trajectories and sustain diverse networks. Rather than the singular immigrant, scholars now detail the diversity of immigration circumstances, class backgrounds, gendered transitions, and the sheer multitude of migration experiences (Lie 1995, p. 304).19 This greatly exaggerates the shift in perspective, at least in the American context. Long before 'diaspora' became fashionable, historians and sociologists of immigration had abandoned - if indeed they ever held - simplistic assumptions about unidirectional trajectories, sharp and definitive breaks with home countries, and a singular path of assimilation. If Glazer and Moynihan's (1963, p. v) observation that 'the point about the melting pot... is that it did not happen' was iconoclastic when first made, it had become widely accepted by the end of the 1960s. So much emphasis was placed on ethnic persistence in the historical and sociological literature between about 1965 and 1985 - again, before the 'diaspora' explosion - that there has even been, in reaction, a certain 'return of assimilation' (Brubaker 2001) in the last two decades (albeit of a more subtle, multidimensional, and normatively ambivalent concept of assimilation). More important than the alleged novelty and originality of the literature is the alleged novelty and import of the phenomenon itself. Does 'diaspora' - along with kindred terms such as transnationalism, post-nationalism, globalization, deterritorialization, postcolonialism, creolization, transculturalism, and postmodernity - name something fundamentally new in the world? Do these terms mark, or at least augur, an epochal shift, as some theorists (Kearney 1991; Appadurai 1996) have suggested? Have we passed from the age of the nation-state to the age of diaspora? More specifically, does the 'unprecedented porosity' of borders (Sheffer 2003, p. 22) - the unprecedented circulation of people, goods, messages, images, ideas and cultural products - signify a basic realignment of the relationship between politics and culture, territorial state and de-territorialized identities? Does this entail the transcendence of the nation-state, based on territorial closure, exclusive claims on citizens' loyalty, and a homogenizing, nationalizing, assimilationist logic? Does the age of diaspora open up new possibilities for what Clifford has called 'non-exclusive practices of community, politics and The 'diaspora' diaspora 9 difference' (1994, p. 302)? Does it offer 'an alternative to life in territorially and nationally marked groups'?20 Obviously, the world has changed, and so have our ways of talking about it. But one should be sceptical of grand claims about radical breaks and epochal shifts (Favell 2001). Can one, in fact, speak of an unprecedented porosity of borders? Not with regard to the movement of people. Over the course of the nineteenth and, especially, the twentieth centuries, states have gained rather than lost the capacity to monitor and control the movement of people by deploying increasingly sophisticated technologies of identification and control including citizenship, passports, visas, surveillance, integrated databases and biometric devices.21 The shock of 9/11 has only pushed states further and faster along a path on which they were already moving. No liberal state, to be sure, can absolutely seal its borders. On balance, however, the world's poor who seek work or refuge in prosperous and peaceful countries encounter a tighter mesh of state regulation and have fewer opportunities for migration to prosperous and peaceful countries than they had a century ago (Hirst and Thompson 1999, pp. 30-31, 267). Is migration today unprecedented in volume and velocity? How one answers this question depends, of course, on one's units of analysis. Migrant flows of recent decades to the United States are, in fact, much smaller, in relation both to the population of the United States and to the population of the rest of the world, than those of a century ago. And while contemporary migrations worldwide are 'more geographically extensive than the great global migrations of the modern era', they are 'on balance slightly less intensive' (Held et al. 1999, p. 326, italics added). Even if there are as many as 120 million international migrants (including refugees and asylum seekers) today, as one expert has suggested, this amounts to less than 2 per cent of the global population (Castles and Miller 1998, pp. 4-5); the mobility of the great majority remains severely limited by the morally arbitrary facts of birthplace and inherited citizenship and by the exclusionary policies of states. Is migration today neither unidirectional nor permanent? Of course not, in many cases; but it was neither unidirectional nor permanent, in many cases, a century ago. Historians have long highlighted the very high rates of return migration from North America to various European countries of origin in the late nineteenth and early twentieth centuries. Do migrants make a sharp and definitive break with their homelands? Of course not. Nor did they do so a century ago, as an abundant historical literature has made clear. Do migrants sustain ties with their country of origin? They do indeed; but they managed to do so by non-electronic means a century ago (Hollinger 1995, pp. 151 ff; Morawska 2001; Waldinger and Fitzgerald 2004). This is not to say that nothing has changed, or that distance-eclipsing technologies of 10 Rogers Brubaker communication and transportation do not matter; it is, however, to caution against exaggerated claims of an epochal break. Have the exclusive claims of the nation-state been eroded? They have indeed, but the nation-state - as opposed to the multifarious particular nation-states - is a figment of the sociological imagination. 'The' nation-state is the primary conceptual 'other' against which diaspora is defined - and often celebrated (Tololyan 1991; Clifford 1994, p. 307). But there is a risk of essentializing 'the' nation-state, a risk of attributing to it a timeless, self-actualizing, homogenizing 'logic'. Sophisticated discussions are sensitive to the heterogeneity of diasporas; but they are not always as sensitive to the heterogeneity of nation-states. Discussions of diaspora are often informed by a strikingly idealist, teleological understanding of the nation-state, which is seen as the unfolding of an idea, the idea of nationalizing and homogenizing the population.22 The conceptual antithesis between nation-state and diaspora obscures more than it reveals, occluding the persisting significance (and great empirical variety) of nation-states (Mann 1997). Entity or stance? Like nation, ethnic group or minority - terms with which it shares an overlapping semantic field (Tololyan 1991, 1996) - diaspora is often characterized in substantialist terms as an 'entity'. As one example among many, consider the beginning of a new book by Gabriel Sheffer: The highly motivated Koreans and Vietnamese toiling hard to become prosperous in bustling Los Angeles, the haggard Palestinians living in dreary refugee camps near Beirut and Amman, the beleaguered Turks dwelling in cramped apartments in Berlin, and the frustrated Russians in Estonia, all have much in common. All of them, along with Indians, Chinese, Japanese, Africans, African-Americans, Jews, Palestinians, Greeks, Gypsies, Romanians, Poles, Kurds, Armenians and numerous other groups permanently residing outside of their countries of origin, but maintaining contacts with people back in their old homelands, are members of ethno-national diasporas (2003, p. 1). The problem here is with the definite article. Diasporas are treated as 'bona fide actual entities' (ibid, p. 245) and cast as unitary actors. They are seen as possessing countable, quantifiable memberships. And indeed they are counted. We learn from the 'first attempt to estimate the real numbers of the main historical, modern and incipient diasporas' that among 'historical diasporas', the Chinese diaspora The 'diaspora'diaspora 11 numbers 35 million, the Indian diaspora 9 million, the Jewish and Gypsy diasporas 8 million each, the Armenian diaspora 5.5 million, the Greek diaspora 4 million, the German diaspora 2.5 million and the Druze 1 million. Among 'modern' diasporas, the African-American diaspora numbers 25 million, the Kurdish diaspora 14 million, the Irish diaspora 10 million, the Italian diaspora 8 million, the Hungarian and Polish diasporas 4.5 million each, the Turkish and Iranian diasporas 3.5 million each, the Japanese diaspora 3 million, the Lebanese (Christian) diaspora 2.5 million and the 'Black Atlantic' diaspora 1.5 million. A similar list with numbers is given for thirty 'incipient diasporas' (Sheffer 2003, pp. 104-6, my italics). Sheffer, to be sure, distinguishes 'core', 'marginal', and 'dormant' members of diasporas (2003, p. 100), but then goes on to quantify diaspora memberships, without any grounds for indicating that those so counted identify as members of the respective diasporas at all. And the large numbers he provides - 25 million, for example, for the incipient Russian diaspora - suggest that the numbers take no account of these distinctions but are maximally inclusive. The very notion of 'dormant members' of a diaspora is problematic; if they are really dormant - if they have 'assimilated or fully integrated' into a host society (Sheffer 2003, p. 100) and merely 'know or feel that their roots are in the diaspora group' - then why should they count, and be counted, as 'members' of the diaspora at all? What is it that Sheffer and others are counting when they count 'members' of diasporas? It appears that what is usually counted, or rather estimated, is ancestry. But if one takes seriously boundary maintenance, lateral ties to fellow diaspora members in other states and vertical ties to the homeland, then ancestry is surely a poor proxy for membership in a diaspora. Enumerations such as this suggest that discussions of diaspora opportunistically combine elements of strong and weak definitions.23 Strong definitions are used to emphasize the distinctiveness of diaspora as a social form; weak definitions, to emphasize numbers (and thereby the import of the phenomenon). Not all discussions of diaspora, to be sure, emphasize boundary-maintenance. Some discussions, as indicated above, emphasize hy-bridity, fluidity, creolization and syncretism and offer an alternative to the groupist portrayal of diasporas as tangible, quantifiable, and bounded entities (though these discussions, too, tend to speak of diasporas as distinctive communities with distinctive identities, without explaining how such distinctive communities and identities can emerge if all is hybrid, fluid, creolized and syncretic). Where boundary maintenance and distinctive identity are emphasized, as they are in most discussions, familiar problems of 'groupism' arise (Brubaker 2002). The metaphysics of the nation-state as a bounded territorial community may have been overcome; but the 12 Rogers Brubaker metaphysics of 'community' and 'identity' remain. Diaspora can be seen as an alternative to the essentialization of belonging;24 but it can also represent a non-territorial form of essentialized belonging. Talk of the de-territorialization of identity is all well and good; but it still presupposes that there is 'an identity' that is reconfigured, stretched in space to cross state boundaries, but on some level fundamentally the same. Yet if, as Homi Bhaba put it in a discussion of Rushdie's Satanic Verses, 'there is no such whole as the nation, the culture, or even the self',25 then why should there be any such whole as the Indian or Chinese or Jewish or Armenian or Kurdish diaspora? To overcome these problems of groupism, I want to argue that we should think of diaspora not in substantialist terms as a bounded entity, but rather as an idiom, a stance, a claim. We should think of diaspora in the first instance as a category of practice, and only then ask whether, and how, it can fruitfully be used as a category of analysis.26 As a category of practice, 'diaspora' is used to make claims, to articulate projects, to formulate expectations, to mobilize energies, to appeal to loyalties.27 It is often a category with a strong normative change. It does not so much describe the world as seek to remake it. As idiom, stance, and claim, diaspora is a way of formulating the identities and loyalties of a population. Those who do the formulating may themselves be part of the population in question; or they may be speaking in the name of the putative homeland state.28 In either case, though, not all those who are claimed as members of putative diasporas themselves adopt a diasporic stance. Indeed, those who consistently adopt a diasporic stance, as Tololyan (1996, p. 19) has noted, are often only a small minority of the population that political or cultural entrepreneurs formulate as a diaspora. According to one comprehensive sociological analysis (Bakalian 1992, cf Tololyan 1996, p. 15), for example, what is casually called 'the Armenian diaspora' is, in the US, not very diasporic at all and is becoming less rather than more so over time, as the large majority of those who identify as Armenians distance themselves from diasporic stances, from links to the homeland, and from links to Armenians in other countries. Their 'Armenianness' is closer to what sociologist Herbert Gans (1979) long ago called 'symbolic ethnicity'. There is, of course, a committed diasporan or diasporic fraction, as Tololyan (1996, p. 18) calls it, among Armenians and many other dispersed populations. And they have good reason to refer to all dispersed Armenians as a 'diaspora'. For them, diaspora is a category of practice and central to their project. But why should we, as analysts, use diaspora as a category of analysis to refer to all persons of Armenian descent living outside Armenia? The disadvantage of doing so is that it occludes the difference between the actively diasporan The 'diaspora' diaspora 13 fraction and the majority who do not adopt a diasporic stance and are not committed to the diasporic project. In sum, rather than speak of 'a diaspora' or 'the diaspora' as an entity, a bounded group, an ethnodemographic or ethnocultural fact, it may be more fruitful, and certainly more precise, to speak of diasporic stances, projects, claims, idioms, practices, and so on. We can then study empirically the degree and form of support for a diasporic project among members of its putative constituency, just as we can do when studying a nationalist project. And we can explore to what extent, and in what circumstances, those claimed as members of putative diasporas actively adopt or at least passively sympathize with the diasporic stance,29 just as we can do with respect to those who are claimed as members of putative nations, or of any other putative collectivity. Conclusion One of the virtues of 'diaspora', scholars have suggested, is that it provides an alternative to teleological, nation-statist understandings of immigration and assimilation. But theories of 'diaspora' have their own teleologies. Diaspora is often seen as destiny - a destiny to which previously dormant members (or previously dormant diasporas in their entirety) are now 'awakening' (Sheffer 2003, p. 21). Embedded in the teleological language of 'awakening' - the language, not coin-cidentally, of many nationalist movements - are essentialist assumptions about 'true' identities. Little is gained if we escape from one teleology only to fall into another. The point of this analysis has not been to deflate diaspora, but rather to de-substantialize it, by treating it as a category of practice, project, claim and stance, rather than as a bounded group. The 'groupness' of putative diasporas, like that of putative 'nations', is precisely what is at stake in [political, social and cultural] struggles. We should not, as analysts, prejudge the outcome of such struggles by imposing groupness through definitional fiat. We should seek, rather, to bring the struggles themselves into focus, without presupposing that they will eventuate in bounded groups. Acknowledgements I would like to thank Natan Sznaider for the opportunity to present an initial version of this article at the conference on 'Diaspora Today', Elmau, Germany, 16-17 July 2003. Thanks also to ERS reviewers for comments, and to Rob Jansen and Kristin Surak for helpful research assistance. 14 Rogers Brubaker Notes 1. Calculated from Dissertation Abstracts, http://wwwlib.umi.com/dissertations/search. 2. On trading diasporas generally - or what have also been called 'middleman minorities' (Bonacich 1973) - see Fallers 1962, Curtin 1984, Cohen 1997, pp. 83-104. See also Wang 2000 on the overseas Chinese, Winder 1962 on the Lebanese, Armstrong 1978 on Baltic Germans, and Cohen 1971 on the Hausa. 3. Vertovec (2000) distances himself from the generalized extension of the term 'diaspora' to dispersed adherents of religious faiths, but, in agreement with Parekh (1993), makes the case for a Hindu diaspora on the grounds that Hinduism, like Judaism and Sikhism, is a special case of a non-proselytizing 'ethnic religion'. On the Sikh diaspora, see Axel 2001; on the Muslim diaspora, Kastoryano 1999, Saint-Blancat 2002, Werbner 2002. 4. For a sceptical discussion of the notion of an Italian diaspora, see Gabaccia 2000, pp. 5-12. 5. I have myself contributed to this form of proliferation with a paper on 'accidental diasporas' (Brubaker 2000). See also Kolstoe 1995 and Mandelbaum 2000. 6. The list would be much longer if one included the putative diasporas discussed in the journalistic or activist literatures. 7. Of these, the academic literature includes articles or books on the Dixie, white, liberal, gay, queer, and digital diasporas. 8. A survey of library catalogues confirmed this dispersion. A search of the WorldCat union catalogue for the keyword 'diasp?' revealed that nearly all (17 of 18) books published on diasporas between 1900 and 1910 addressed the Jewish case; as late as the 1960s, this remained true of 15 of 20 books sampled (out of a total of 78 books matching the keyword published between 1960 and 1970). In 2002, by contrast, the top 20 books sampled (of 253 published that year) addressed 8 different cases; only two of the twenty addressed the Jewish case. 9. Clifford (1994, pp. 305, 325-327) has made the point powerfully. 10. Dissertations on diaspora appear to be more interdisciplinary, in terms of the mean number of subject keywords associated with them, than other dissertations. For a sample of 15 'modern history' dissertations with the keyword 'diaspora', there was a pool of 19 other subject terms represented, and an average of 3.20 terms per entry. For non-diaspora modern history dissertations, the subject pool was 14 terms, with an average of 2.73 terms per entry (calculated from Dissertation Abstracts, http://wwwlib.umi.com/dissertations/search). 11. On the importance of the media for the popularization - and conceptual stretching -of the term, see Tololyan 1996, p. 10. 12. 'Diasporist', it should be noted, can have two quite different meanings. With respect to Jews, and sometimes in other cases (see, for example, Clifford 1994, p. 321 on the African diaspora), it designates a positive orientation to the diaspora at the expense of an actual or putative homeland, a valorization of lateral over centripetal (homeland-oriented) connections, in Clifford's terms. Thus in the Jewish case, 'diasporist' is opposed to Zionist (see, for example, Boyarin and Boyarin 1993), an opposition taken to an extreme in Philip Roth's novel Operation Shylock. In many other contexts, however, 'diasporist' designates a positive orientation to the diasporic condition (which may include a constitutive commitment to a homeland) in the face of the exclusive claims of the nation-state of residence on loyalty and identity. 13. The whimsically inclined might find further proliferation tempting. One might, for example, speak of diasporosity (to designate the permeability of the boundaries of a diaspora); diasporific (to characterize the catastrophic origins of many diasporas); diaspornography (as a new term for the global trafficking industry); diasportfolio (a new global investment strategy); diaspersion (an unkind remark about a diaspora); diasporapathy (to characterize putative members of a diaspora who do not respond to the appeals of diasporactivists) ; and diasperanto (a project for a common language of the diaspora). The 'diaspora'diaspora 15 14. For sustained discussions of definitional issues, see Safran 1991; Clifford 1994; Tololyan 1996; Cohen 1997; Sheffer 2003. 15. This is Safran's formulation: 'they, or their ancestors, have been dispersed' (1991: 83); the passive formulation does not allow for voluntary dispersion. Cohen (1997) and others see this as too limiting. 16. This point was noted by Cohen, who preserves three homeland-related criteria in his own enumeration of nine 'common features of a diaspora' (1997, pp. 23, 26). 17. On the temporal dimension, see Marienstras 1989, p. 125; Cohen 1997, pp. 185-6. 18. On methodological nationalism, see Centre for the Study of Global Governance 2002; Wimmer and Glick Schiller 2003. Although she does not use the term, Soysal (2000) makes a partially similar argument. 19. For related articles about a fundamental shift in perspective, see Glick Schiller et al. 1995; Kearney 1995; Beck 2000. 20. Natan Sznaider, opening remarks to the conference on 'Diaspora Today', Schloss Elmau, Germany, 17 July 2003. 21. For a critique of the view that states have lost their ability to control their borders, see Brubaker 1994; Freeman 1994; and, for a more detailed account, Zolberg 1999. On the mid-nineteenth century codification of citizenship as a means of controlling migration, see Brubaker 1992, pp. 64 ff. On the historical development of passports and related techniques of identification, see Torpey 2000 and the studies collected in Caplan and Torpey 2001. 22. For a nuanced argument about 'cosmopolitics' - as a mode of 'thinking and feeling beyond the nation' - that does not treat the nation-state and cosmopolitanism as antithetical, see Robbins 1998. David Hollinger has also argued eloquently that the 'nation' need not be antithetical to cosmopolitan or transnational engagements, but can sometimes mediate effectively 'between the ethnos and the species' (1998, p. 87; see also 1995, ch. 6). 23. For an argument that discussions of identity are bedeviled by a mix of strong and weak definitions, see Brubaker and Cooper 2000. 24. The former possibility has been emphasized by Gilroy 1997, p. 328 and by Natan Sznaider, in his opening remarks to the conference on 'Diaspora Today', Schloss Elmau, Germany, 17 July 2003. The latter possibility has been noted by Anthias 1998, pp. 560, 563, 567. 25. Quoted in Tambiah 2000, p. 178. 26. For a very different argument criticizing the use of diaspora as an analytical category in the study of immigration, see Soysal 2000. For an argument about categories of analysis and categories of practice in the study of ethnicity, race, and nation, see Brubaker 2002. 27. Writing on the African diaspora, Patterson and Kelly (2000, p. 19) observe that 'the presumption that black people worldwide share a common culture was not... the result of poor scholarship. It responded to a political imperative - one that led to the formation of political and cultural movements premised on international solidarity'. They quote Hall's (1990, p. 224) remark that unitary images of diaspora offered 'a way of imposing an imaginary coherence on the experience of dispersal and fragmentation'. 28. On the changing historical stances of sending states towards emigrant populations and their descendants, see, for example, Smith 2003 (on Mexico, Italy, and Poland), Gabaccia 2000 (on Italy), Itzigsohn 2000 (on the Dominican Republic, Haiti, and El Salvador), and Wang 1993 (on China). 29. There is no reason to expect that people will be consistent in this respect. They may well adopt a diasporic stance at some moments or in some contexts, and distance themselves from such a stance in other times and contexts. References ALBA, RICHARD D. and NEE, VICTOR 1997 'Rethinking assimilation theory for a new era of immigration', International Migration Review, vol. 31, no. 4, pp. 826-74 16 Rogers Brubaker -2003 Remaking the American Mainstream: Assimilation and Contemporary Immigration , Cambridge, MA: Harvard University Press ALPERS, EDWARD A. 2001 'Defining the African Diaspora'. Paper presented at Comparative Social Analysis Workshop, UCLA ANDERSON, BENEDICT 1998 The Spectre of Comparisons: Nationalism, Southeast Asia, and the World, London: Verso ANGOUSTURES, ALINE and PASCAL, VALERIE 1996 'Diasporas et financement des conflits', in Francois Jean and Jean-Christophe Rutin (eds), Economie des guerres civiles, Paris: Hachette, pp. 495-542 ANTHIAS, FLOYA 1998 'Evaluating "diaspora": Beyond ethnicity', Sociology, vol. 32, no. 3, pp. 557-80 APPADURAI, ARJUN 1996 Modernity at Large: Cultural Dimensions of Globalization, Minneapolis, MN: University of Minnesota Press ARMSTRONG, JOHN A. 1976 'Mobilized and proletarian diasporas', American Political Science Review, vol. 70, no. 2, pp. 393-408 - 1978 'Mobilized diaspora in Tsarist Russia: The case of the Baltic Germans', in Jeremy Azrael (ed.), Soviet Nationality Policies and Practices, New York: Praeger, pp. 63-104 - 1982 Nations Before Nationalism, Chapel Hill: University of North Carolina Press AXEL, BRIAN KEITH 2001 The Nations Tortured Body, Durham and London: Duke University Press BAKALIAN, ANNY P. 1992 Armenian-Americans: From Being to Feeling Armenian, New Brunswick: Transaction Publishers BARTH, FREDRIK 1969 'Introduction', in Fredrik Barth (ed.), Ethnic Groups and Boundaries: The Social Organization of Cultural Difference, London: George Allen & Unwin, pp. 9-38 BAUMANN, GERD 1996 Contesting Culture: Discourses of Identity in Multi-Ethnic London, Cambridge: Cambridge University Press BECK, ULRICH 2000 'The cosmopolitan perspective: Sociology of the second age of modernity', British Journal of Sociology, vol. 51, no. 1, pp. 79-105 BHATT, CHETAN and MUKTA, PARITA 2000 'Hindutva in the West: Mapping the antinomies of diaspora nationalism', Ethnic and Racial Studies, vol. 23, no. 3, pp. 407-41 BONACICH, EDNA 1973 A theory of middleman minorities', American Sociological Review, vol. 38, no. 583-94 BOYARIN, DANIEL and BOYARIN, JONATHAN 1993 'Diaspora: Generation and the ground of Jewish identity', Critical Inquiry, vol. 19, pp. 693-725 BRUBAKER, ROGERS 1992 Citizenship and Nationhood in France and Germany, Cambridge, MA: Harvard University Press - 1994 Are immigration control efforts really failing?', in Wayne A. Cornelius, Philip L. Martin, and James F. Hollifield (eds), Controlling Immigration: A Global Perspective, Stanford, CA: Stanford University Press, pp. 227-31 -2000 Accidental diasporas and external "homelands" in central and eastern Europe: past and present', Institute for Advanced Studies, Vienna, Political Science Series no. 71 - 2001 'The return of assimilation?', Ethnic and Racial Studies, vol. 24, no. 4, pp. 531-48 -2002 'Ethnicity without groups', Archives europeenes de sociologie, vol. XLIII, no. 2, pp. 163-89 BRUBAKER, ROGERS and COOPER, FREDERICK 2000 'Beyond "identity"', Theory and Society, vol. 29, no. 1, pp. 1-47 CAPLAN, JANE and TORPEY, JOHN 2001 Documenting Individual Identity: The Development of State Practices in the Modern World, Princeton: Princeton University Press CASTLES, STEPHEN and MILLER, MARK J. 1998 The Age of Migration: International Population Movements in the Modern World, New York: The Guilford Press CLIFFORD, JAMES 1994 'Diasporas', Cultural Anthropology, vol. 9, no. 3, pp. 302-38 The 'diaspora' diaspora 17 COHEN, ABNER 1971 'Cultural strategies in the organization of trading diasporas', in Claude Meillassoux (ed.), The Development of Indigenous Trade and Markets, London: Oxford University Press, pp. 266-81 COHEN, ROBIN 1997 Global Diasporas: An Introduction, Seattle, WA: University of Washington Press CONNOR, WALKER 1986 'The impact of homelands upon diasporas', in Gabriel Sheffer (ed.), Modern Diasporas in International Politics, London: Croom Helm, pp. 16-46 CURTIN, PHILIP D. 1984 Cross-Cultural Trade in World History, Cambridge: Cambridge University Press EDWARDS, BRENT H. 2001 'The uses of diaspora', Social Text, vol. 19, no. 1, pp. 45-73 FALLERS, LLOYD A. 1962 'Comments on "The Lebanese in West Africa'", Comparative Studies in Society and History, vol. 4, no. 3, pp. 334-6 FALZON, MARK-ANTHONY 2003 '"Bombay, our cultural heart": Rethinking the relation between homeland and diaspora', Ethnic and Racial Studies, vol. 26, no. 4, pp. 662-83 FAVELL, ADRIAN 2001 'Migration, mobility and globaloney: Metaphors and rhetoric in the sociology of globalization', Global Networks, vol. 1, no. 4, pp. 389-98 FREEMAN, GARY P. 1994 'Can liberal states control unwanted migration?', Annals of the American Academy of Political and Social Science, vol. 534, pp. 17-30 GABACCIA, DONNA R. 2000 Italy's Many Diasporas, London: UCL Press GANS, HERBERT J. 1979 'Symbolic ethnicity: The future of ethnic groups and cultures in America', Ethnic and Racial Studies, vol. 2, no. 1, pp. 1-20 GILROY, PAUL 1997 'Diaspora and the detours of identity', in Kathryn Woodward (ed.), Identity and Difference, London and Thousand Oaks, CA: Sage Publications in association with the Open University, pp. 299-343 GLEASON, PHILIP 1983 'Identifying identity: a semantic history', The Journal of American History, vol. 69, no. 4, pp. 910-31 GLAZER, NATHAN and MOYNIHAN, DANIEL PATRICK 1963 Beyond the Melting Pot: The Negroes, Puerto Ricans, Jews, Italians, and Irish of New York City, Cambridge, MA: MIT Press GLICK SCHILLER, NINA, BASCH, LINDA, and BLANC, CHRISTINA SZANTON 1995 'From immigrant to transmigrant: Theorizing transnational migration', Anthropological Quarterly, vol. 68, pp. 48-63 HALL, STUART 1990 'Cultural identity and diaspora', in Jonathan Rutherford (ed.), Identity: Community, Culture, Difference, London: Lawrence & Wishart, pp. 222-37 HELD, DAVID, MCGREW, ANTHONY, GOLDBLATT, DAVID, and PERRATON, JONATHAN 1999 Global Transformations: Politics, Economics, and Culture, Stanford, CA: Stanford University Press HIRST, PAUL and THOMPSON, GRAHAME 1999 Globalization in Question, Cambridge: Polity Press HOLLINGER, DAVID A. 1995 Postethnic America: Beyond Multiculturalism, New York: Basic Books - 1998 'Nationalism, cosmopolitanism, and the United States', in Noah M. J. Pickus (ed.), Immigration and Citizenship in the Twenty-First Century, Lanham: Rowman and Littlefield, pp. 85-99 ITZIGSOHN, JOSE 2000 'Immigration and the boundaries of citizenship: The institutions of immigrants' political transnationalism', Immigration and the Boundaries of Citizenship, vol. 34, no. 4, pp. 1126-54 KASTORYANO, RIVA 1999 'Muslim diaspora(s) in western Europe', South Atlantic Quarterly, vol. 98, no. 1-2, pp. 191-202 KEARNEY, MICHAEL 1991 'Borders and boundaries of state and self at the end of empire', Journal of Historical Sociology, vol. 4, no. 1, pp. 52-74 - 1995 'The local and the global: The anthropology of globalization and transnationalism', Annual Review of Anthropology, vol. 24, pp. 547-65 18 Rogers Brubaker KING, CHARLES and MELVIN, NEIL J. 1998 Nations Abroad: Diaspora Politics and International Relations in the Former Soviet Union, Boulder: Westview Press - 1999 'Diaspora polities', International Security, vol. 24, no. 3, pp. 108-24 KOLSTOE, PAUL 1995 Russians in the Former Soviet Republics, London: Hurst and Company LAITIN, DAVID D. 1995 'Marginality: A microperspective', Rationality and Society, vol. 7, no. 1, pp. 31-57 LIE, JOHN 1995 'From international migration to transnational diaspora', Contemporary Sociology, vol. 24, no. 4, pp. 303-6 MANDELBAUM, MICHAEL 2000 The New European Diasporas: National Minorities and Conflict in Eastern Europe, New York: Council on Foreign Relations Press MANN, MICHAEL 1997 'Has globalization ended the rise and rise of the nation-state?', Review of International Political Economy, vol. 4, no. 3, pp. 472-96 MARIENSTRAS, RICHARD 1989 'On the notion of diaspora', in Gerard Chaliand (ed.), Minority Peoples in the Age of Nation-States, London: Pluto, pp. 119-25 MORAWSKA, EWA 2001 'Immigrants, transnationalism, and ethnicization: A comparison of this great wave and the last', in Gary Gerstle and John Mollenkopf (eds), E Pluribus Unum? Contemporary and Historical Perspectives on Immigrant Political Incorporation, New York: Rusell Sage, pp. 175-212 PAREKH, BHIKHU 1993 'Some reflections on the Indian diaspora', Journal of Contemporary Thought, vol. 3, pp. 105-51 PATTERSON, TIFFANY RUBY and KELLEY, ROBIN D. G. 2000 'Unfinished migrations: Reflections on the African diaspora and the making of the modern world', African Studies Review, vol. 43, no. 1, pp. 11-45 ROBBINS, BRUCE 1998 'Introduction Part I: Actually existing cosmopolitanism', in Pheng Cheah and Bruce Robbins (eds), Cosmopolitics: Thinking and Feeling Beyond the Nation, Minneapolis, MN: University of Minnesota Press, pp. 1-19 SAFRAN, WILLIAM 1991 'Diasporas in modern societies: Myths of homeland and return', Diaspora, vol. 1, no. 1, pp. 83-99 SAINT-BLANCAT, CHANTAL 2002 'Islam in diaspora: Between reterritorialization and extraterritoriality', International Journal of Urban and Regional Research, vol. 26, no. 1, pp. 138-51 SARTORI, GIOVANNI 1970 'Concept misinformation in comparative polities', American Political Science Review, vol. LXIV, no. 4, pp. 1033-53 SHEFFER, GABRIEL 1986 A new field of study: Modern diasporas in international polities', in Gabriel Sheffer (ed.), Modern Diasporas in International Politics, London: Croom Helm, pp. 1-15 -2003 Diaspora Politics: At Home Abroad, Cambridge: Cambridge University Press SHEPPERSON, GEORGE 1966 'The African diaspora - or the African abroad', African Forum, vol. 1, no. 2, pp. 76-93 SMITH, ANTHONY D. 1986 The Ethnic Origins of Nations, Oxford: Basil Blackwell SMITH, ROBERT C. 2003 'Diasporic memberships in historical perspective: Comparative insights from the Mexican, Italian, and Polish cases', International Migration Review, vol. 37, no. 3, pp. 724-59 SOYSAL, YASEMIN 2000 'Citizenship and identity: Living in diasporas in post-war Europe', Ethnic and Racial Studies, vol. 23, no. 1, pp. 1-15 TAMBIAH, STANLEY J. 2000 'Transitional movements, diaspora, and multiple modernities', Daedalus, vol. 129, no. 1, pp. 163-94 TOLOLYAN, KHACHIG 1991 'The nation-state and its others: In lieu of a preface', Diaspora, vol. 1, no. 1, pp. 3-7 - 1996 'Rethinking diaspora(s): stateless power in the transnational moment', Diaspora, vol. 5, no. 1, pp. 3-36 TORPEY, JOHN 2000 The Invention of the Passport: Surveillance, Citizenship and the State, Cambridge: Cambridge University Press The 'diaspora' diaspora 19 VERTOVEC, STEVEN 2000 The Hindu Diaspora: Comparative Patterns, London: Routledge WALDINGER, ROGER and FITZGERALD, DAVID 2004 'Transnationalism in question', American Journal of Sociology, vol. 109, no. 5, pp. 1177-95 WANG, GUNGWU 1993 'Greater China and the Chinese overseas', The China Quarterly, no. 136, pp. 926-48 - 2000 The Chinese Overseas: From Earthbound China to the Quest for Autonomy, Cambridge, Massachusetts: Harvard University Press WERBNER, PNINA 2002 Imagined Diasporas Among Manchester Muslims, Oxford: James Currey WIMMER, ANDREAS and GLICK SCHILLER, NINA 2003 'Methodological nationalism, the social sciences, and the study of migration: An essay in historical epistemology', International Migration Review, vol. 37, no. 3, pp. 576-610 WINDER, R. BAYLY 1962 'The Lebanese in West Africa', Comparative Studies in Society and History, vol. 4, no. 3, pp. 296-333 ZOLBERG, ARISTIDE 1999 'Matters of state: Theorizing immigration policy', in Charles Hirschman, Philip Kasnitz, and Josh DeWind (eds), The Handbook of International Migration: The American Experience, New York: Russell Sage Foundation, pp. 71-93 ROGERS BRUBAKER is Professor of Sociology at the University of California, Los Angeles. ADDRESS: Department of Sociology, UCLA, 264 Haines Hall, 375 Portola Plaza, Los Angeles, CA 90095-1551, USA. Email: