Language, religion and the politics of difference* ROGERS BRUBAKER University of California, Los Angeles, USA ABSTRACT. Language and religion are arguably the two most socially and politically consequential domains of cultural difference in the modern world. Yet there have been very few efforts to compare the two in any sustained way. I begin by aligning language and religion, provisionally, with ethnicity and nationhood, and by sketching �ve ways in which language and religion are both similar to and similarly intertwined with ethnicity and nationhood. I then identify a series of key differences between language and religion and draw out their implications for the political accommodation of cultural heterogeneity. I show that religious pluralism tends to be more intergenerationally robust and more deeply institutionalised than linguistic pluralism in western liberal democracies, and I argue that religious pluralism entails deeper and more divisive forms of diversity. The upshot is that religion has tended to displace language as the cutting edge of contestation over the political accommodation of cultural difference – a striking reversal of the longer-term historical process through which language had previously displaced religion as the primary focus of contention. KEYWORDS: religion, language, pluralism, multiculturalism, ethnicity, immigration I am deeply honoured by the invitation to deliver the Ernest Gellner Nationalism Lecture. I did not know Gellner well, but I have been reading and teaching his works with pleasure and pro�t for a quarter of a century. And I owe my long and fruitful association with the Nationalism Studies Program at the Central European University, in a sense, to Gellner, for after his death I helped set up the program as a successor to the Center for the Study of Nationalism that he had founded at the newly established university in 1992. Gellner was a deep thinker, a profound analyst not only of nationalism, but also of modernity and Muslim society, conspicuously treating all three – contrary to current fashion – in the singular, not the plural. He was also a great stylist: witty, sharp, irreverent and contrarian, fond of irony, paradox and antithesis, with a remarkable gift for memorable formulations. * Editor’s note: this is a revised version of the ASEN / Nations and Nationalism Ernest Gellner Nationalism Lecture, delivered at the London School of Economics and Political Science, 26 March 2012. Note: Corrections added on 3 December 2012 after initial online publication on 22 November 2012: some minor errors in the text have been corrected in this version of the article. bs_bs_banner EN ASJ OURNAL OF THE ASSOCIATION FOR THE STUDY OF ETHNICITY AND NATIONALISM NATIONS AND NATIONALISM Nations and Nationalism 19 (1), 2013, 1–20. DOI: 10.1111/j.1469-8129.2012.00562.x © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 My lecture brings together two subjects about which Gellner had profoundly interesting things to say: language and religion. Language is obviously at the heart of his theory of nationalism, which highlights the need for context-free communication, identi�cation with literate high cultures, and exo-socialisation through state-organised school systems. And the famous Ruritanian parable turns centrally on language. Religion is less obviously central to Gellner’s theory of nationalism, and Gellner’s unapologetic Enlightenment rationalism no doubt made him even more �religiously unmusical’ – to use Max Weber’s famous self-description – than Weber himself. Yet religion plays a considerably larger role in Gellner’s theory of nationalism than is evident on the surface. The theory turns not on language per se, but on culture and communication; and Gellner emphasises that some forms of religious high culture – notably the scripturalist, egalitarian high culture of Islam – are deeply compatible with nationalism, and indeed often fused with nationalism. What Gellner calls �deeply engrained religiouscultural habits’, moreover, are central to the extended discussion of �entropyresistant traits’ and enduring inequalities in Nations and Nationalism. Beyond the theory of nationalism, religion was also of course central to Gellner’s �eldwork in Morocco and to his synoptic book on Muslim society. My own interest in language and religion arises from a longstanding concern with the social organisation and political expression of cultural difference. Language and religion are arguably the two most socially and politically consequential domains of cultural difference in the modern world. The study of the political accommodation of cultural difference – or what might be called the political sociology of multiculturalism – would therefore seem to require sustained attention to both. Yet there have been few efforts to compare language and religion, the outstanding exception – and an important inspiration for this paper – being a paper by Zolberg and Long (1999). Language and religion are of course often discussed together in the literatures on ethnicity, nationalism, minority rights and multiculturalism. But most such discussions involve passing juxtaposition rather than sustained comparison, and the more sustained discussions (see notably Bauböck 2002) tend to be normative rather than empirical. It might be suggested that the lack of sustained comparison is not surprising, since language and religion are simply not comparable. Now I do not want to get sidetracked here by a discussion of the meaning of comparison or the conditions of comparability. My interests are substantive, not methodological. One can certainly construe religion and language in such a way that they are not comparable. If one were to de�ne religion in terms of beliefs and rituals, for example, there would be little leverage for comparison. And religion (at least �organised religion’) obviously has an organisational dimension and a structure of authority that language lacks. But I want to argue that one can nonetheless construe language and religion in a way that makes comparison both possible and fruitful. 2 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 My strategy for doing so is to begin by aligning language and religion, provisionally, with ethnicity and nationhood, and by sketching �ve ways in which language and religion are both similar to, and similarly intertwined with, ethnicity and nationhood. I will then identify a few key differences between language and religion and draw out their implications for the politics of difference. I obviously cannot undertake a full comparative analysis here, but I hope that the friction of comparison produced by reading one domain of cultural practice against another can suggest some potentially fruitful lines of analysis. In particular, I will argue that religious pluralism tends to be more intergenerationally robust and more deeply institutionalised than linguistic pluralism in contemporary liberal societies, and that religious pluralism entails deeper and more divisive forms of diversity. The upshot, I will suggest, is that religion has tended to displace language as the cutting edge of contestation over the political accommodation of cultural difference – a striking reversal of the longer-term historical process through which language had previously displaced religion as the primary focus of contention. Language and religion in relation to ethnicity and nationhood So let me begin by aligning language and religion with ethnicity and nation- alism.1 To begin with, both language and religion are domains of categorically differentiated cultural practice that simultaneously unite and divide. By �categorically differentiated’, I mean that language and religion are understood by participants and observers alike as partitioned into discrete categories, rather than as a continuous spectrum of variation. (That they are so understood is of course a product of history and politics, not least a history and politics of objecti�cation, individuation and boundary-drawing that has carved out distinct �languages’ from dialect continua, and constructed and institutionalised distinct �religions’ from fluid and varying sets of practices.2 ) In popular understandings, both language and religion sort people into distinct, bounded and largely self-reproducing �communities’, and in this respect they are both analogous to ethnic groups and nations and variously intertwined with them. Second, language and religion are basic sources and forms of social, cultural and political identi�cation. They are ways of identifying oneself and others, construing sameness and difference, and naming fundamental social groups. Language and religion are again both analogous to ethnicity and nationalism in this respect and pervasively intertwined with them. Language, religion or both are generally understood as central to or even constitutive of most ethnic and national identi�cations, and they frequently serve as the key diacritical markers, emblems or symbols of such identi�cations. Third, the family is a primary site of linguistic and religious socialisation, as it is of ethnic and national socialisation. Indeed language and religion are ordinarily more central to primary socialisation in the family than are ethnicity and nationality. Language and religion are therefore often deeply taken for Language, religion and the politics of difference 3 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 granted and embodied identi�cations, and both are routinely represented as primordial. Yet fourth, neither religion nor language is in fact primordial or �xed. As we know from the burgeoning constructivist literature, the same holds for ethnicity and nationhood. Like ethnicity and nationhood, religion and language are powerfully shaped by political, economic and cultural processes, and they change as circumstances change. From an individual point of view, as Anderson (1991: 145) said of nations, both religions and languages are �joinable in time’, and in the contemporary world, both are increasingly chosen rather than given. This shift is particularly marked for religion. Although initial religious identi�cations continue to be inherited, modalities and degrees of religious engagement can no longer be taken for granted, but – in the West at least – are increasingly reflexively negotiated and embraced (or rejected) (see e.g. Taylor 2007). Finally, many of the claims made in the name of religious or linguistic groups are similar to – and again, also intertwined with – claims made in the name of ethnic groups or nations. These include claims for economic resources, symbolic recognition, equal representation, cultural reproduction or political autonomy. Such claims-making is part of a more encompassing process of the politicisation of culture and the culturalisation of politics. In all these respects, language and religion are both similar to ethnicity and nationalism and similarly intertwined with them. This has led many scholars of ethnicity to treat language and religion – implicitly or explicitly – as functionally equivalent. Indeed ethnicity was constituted as an object of study precisely by abstracting from the speci�cities of language, religion and other ascriptive markers such as phenotype, region of origin and customary mode of livelihood. In the words of Joseph Rothschild – whose 1981 study Ethnopolitics was one of the �rst, and remains one of the best, to survey the �eld – it would be pointless to �separate out the notion of ethnic consciousness, solidarity, and assertiveness from religious, linguistic, racial, and other so-called primordial foci of consciousness, solidarity, and assertiveness’. If this were to be done, �it is dif�cult to see what precisely would be left to, or meant by, the residual notion of ethnicity and ethnic groups’ (Rothschild 1981:9; cf. Geertz 1963:109ff; Horowitz 1985:41). The call to abstract from cultural content was given its strongest formulation by Barth (1969), who argued that the study of ethnicity should focus on the nature and dynamics of ethnic boundaries, not on what he somewhat dismissively called the �cultural stuff’ these boundaries enclose. This perspective on ethnicity has been immensely fruitful, and it has been important for my own work. But it is also inevitably flattening, for it neglects, by design, the speci�c cultural practices, understandings and institutions that are implicated in the construction and working of ethnic identities and boundaries. It is not fruitful, I believe, to talk about multiculturalism or the politics of difference in terms of highly generalised notions of ethnicity, culture, identity or difference. It is necessary instead to attend to the speci�c logic and 4 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 properties – the speci�c �affordances’ – of differing modes of cultural difference. So in this paper, following the lead of scholars such as Cornell (1996) and Jenkins (1997: chapter 8), I want to return the �cultural stuff’ – speci�cally language and religion – to the centre of analytical attention. I want to register two caveats, however, about doing so. First, the phrase �cultural stuff’ is potentially misleading if it is taken to imply an opposition between culture and social organisation. I treat language and religion not only as cultural forms but also as modes of social organisation and media of interaction. Second, the speci�c con�guration of the contemporary politics of difference has been shaped not only by the intrinsic properties of language and religion – not only by the cultural and social-organisational �stuff’, considered as an ahistorical constant – but also, and indeed more profoundly, by the speci�c historical trajectories through which states understood as �liberal’ and �national’ emerged, and were transformed, in and through their confrontation �rst with religious and later with linguistic heterogeneity.3 Before proceeding further, it is worth underscoring that religion is a much more elusive analytical object than language. For all their complexity, linguistic phenomena have a de�niteness and regularity that religious phenomena lack. We know what we are talking about when we talk about language, but the same cannot be said for religion. It is not accidental that linguistics is a relatively well-de�ned discipline, while religious study is a loose congeries of undertakings. Some have argued that �religion’ is meaningless or useless as an analytical category (see e.g. Bloch 1996); I am not aware that anyone has made this claim about language. While fully acknowledging that �religion’ is a problematic and deeply contested category – contested both as a category of analysis and as a category of practice – I do not want to enter here into the debate about the category. Since the scope of my argument is limited to contemporary liberal polities, I am content to work here with a relatively unreflexive, common-sense category of �religion’ (cf. Casanova 2009: 5), limiting my attention primarily to what we call �organised religion’, and within that �eld primarily to the Abrahamic religions. Trajectories of politicisation and depoliticisation I want to begin with an observation which suggests that language should be more deeply and chronically politicised than religion under modern conditions. Language, after all, is a universal and pervasive medium of social life, while religion is not. If one de�nes religion broadly enough, to be sure, then religion too can be seen as a universal social phenomenon. But it is not universal in the same way.4 Language is a pervasive, inescapable medium of social interaction; religion is not.5 Moreover, language is a necessary medium of public as well as private life. It is an inescapable medium of public discourse, government, administration, law, courts, education, media and public signage. Language, religion and the politics of difference 5 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 However one de�nes religion, it cannot be said to be an inescapable medium or necessary ground of action in any of these domains. Public life can in principle be a-religious, but it cannot be a-linguistic. The modern state is characterised by direct rule, intensive interaction with citizens, universal public education and a public sector that provides large numbers of jobs. As a result, the rules and practices that govern the language of public life directly affect the material and ideal interests of people with differing language repertoires (Zolberg and Long 1999: 21). This holds a fortiori in an economic context in which work is increasingly �semantic and communicative rather than physical’ (Gellner 1997: 85), involving the manipulation of meanings, not of things. Language is therefore chronically and pervasively politicised in linguistically heterogeneous modern societies (May 2001; Patten and Kymlicka 2003). Religion is also politicised, but it is politicised in different ways and for different reasons. The state must privilege a particular language or set of languages, but it need not privilege a particular religion, at least not in the same way and not to the same degree. Complete neutrality, to be sure, is now widely recognised as a myth (Bader 2007: 82ff), not least because the state cannot help taking a position on the question of what counts as �religion’.6 Moreover, one can easily identify pervasive traces of Christianity in the public life of western liberal democracies, even in those with the strongest traditions of separation of church and state or of laicité (Alba 2005). One need think only of such taken for granted frameworks as the reckoning of dates according to the Christian calendar, the organisation of holidays or the privileging of Sunday as a day of rest – the domain of what Torpey (2010) calls �latent religiosity’. Yet contemporary liberal polities – even those that still have some kind of established church – have made substantial, although contested, moves in the direction of a more neutral stance towards differing religions. Such moves have no counterpart in the domain of language. The state can approach neutrality with respect to religion, even if such neutrality can never be fully realised in practice; but it cannot even approach neutrality with respect to language (Zolberg and Long 1999: 21; Bauböck 2002: 175–6). There is a second reason for thinking that language should be more deeply and chronically politicised than language in the contemporary world. According to secularisation theory, modernity has entailed the progressive privatisation and hence the depoliticisation of religion. Events of the last three decades have made simplistic versions of secularisation theory ripe targets for criticism. But as several leading analysts of religion have argued (Casanova 1994; Gorski and Altinordu 2008; Koenig 2011; Martin 2005; Taylor 2007; Turner 2011), secularisation theory is more complex, interesting and robust than many critics suggest, and it cannot be dismissed out of hand. For many in the modern world, religion has indeed become a more individual, subjective and private experience. To the extent that this is the case, religion indeed becomes depoliticised, and religious pluralism can flourish in the private realm without generating conflicts in the public sphere. Over the course of the last several 6 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 centuries, religion has indeed become much less central to public life and less politically contentious in the West, while language has become much more central and more contentious. Yet while secularisation theory captures an important long-term trend, there is a powerful medium-term counter-trend towards the deprivatisation and therefore the repoliticisation of religion (Casanova 1994). On a time scale of decades rather than centuries, conflicts over religion have intensi�ed, while conflicts over language, as I will argue later, have eased.7 As a result, while religion is not necessarily, chronically and pervasively politicised the way language is, the challenges posed by religious pluralism today – or at least by some forms of religious pluralism – tend to be more complex and dif�cult than those posed by linguistic pluralism. I want to develop this argument in two stages. I will begin by arguing that religious pluralism tends to be more robust than linguistic pluralism in contemporary liberal societies and polities. I will then argue that religious pluralism entails deeper and more divisive forms of diversity than linguistic pluralism. The robustness of religious pluralism So let me begin with the greater robustness of religious pluralism. This results from the differing ways in which religious and linguistic pluralism are generated, reproduced and institutionalised in contemporary liberal societies. I will consider each in turn, starting with the generation of pluralism, and then moving on to address the reproduction and institutionalisation of pluralism. 1. Conquest, colonization, and especially (in the contemporary world) migration generate religious and linguistic pluralism in similar ways, by importing it from without. But religious pluralism is also generated from within. I am not concerned here with relatively rare cases of religious splits and foundings, although historically these have been important internal sources of religious pluralism. I am concerned rather with routine individual-level changes in religious af�liation and identity. Individuals routinely change their linguistic repertoires as well as their religious af�liations. But they do so in differing ways, and with differing consequences. For adults, at least, language change is mainly additive, although there may of course be some attrition of pro�ciency in languages that are seldom used. Religious change, on the other hand, is often substitutive and transformative. When adults add a new language to an existing repertory of languages, this may inflect their identity, but it is unlikely to transform it. Yet when they convert from one religion to another or from one form of religious engagement to another, this can involve a basic transformation of identity. People do not ordinarily simply add a new religion to a repertory of religions, notwithstanding the flourishing of various forms of hybridity and syncretism, nor do they ordinarily �convert’ from one language to another. Language, religion and the politics of difference 7 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 For children of immigrants, to be sure, language change is often substitutive rather than additive; but this reduces heterogeneity within the receiving country, while religious conversion often increases it. Conversion can also reduce heterogeneity, and some immigrant groups to the USA, notably Taiwanese, exhibit high rates of conversion to Christianity. But pressures and incentives for conversion to the prevailing religion are on the whole relatively weak in contemporary liberal societies, while incentives to learn the prevailing language are strong. A whole series of factors, in addition to immigration, promote religious pluralisation in contemporary liberal societies: new religious movements, organised proselytism, transnational religious networks, an open religious marketplace and a general climate of spiritual experimentation. There are no analogous forces generating linguistic pluralisation from within. So religious conversion, broadly understood, is an important source of politically signi�cant cultural heterogeneity, while individual-level language change is not. In contemporary liberal societies, new forms and degrees of linguistic pluralism are almost exclusively imported (through immigration), while new forms and degrees of religious pluralism are both imported and endogenously generated through conversion. 2. The second reason for the greater robustness of religious than linguistic pluralism is that religious pluralism is more easily reproduced. Here I shift my perspective from intragenerational to intergenerational change and reproduction. In premodern societies, linguistic pluralism was more or less selfreproducing. Linguistic socialisation occurred in families and local communities, and it did not require any specialised apparatus. Political authorities made no effort to impose linguistic homogeneity (although they often did impose religious homogeneity). In contemporary liberal societies, the situation is reversed. It is now religious pluralism that is more or less self-reproducing. Religious socialisation occurs in families and local religious communities, and political authorities make no effort to impose religious homogeneity. But linguistic reproduction now requires what Gellner (1983: chapter 3) called exo-socialisation. It requires prolonged and expensive schooling on a scale that only the state is ordinarily in a position to provide. So the state is much more central to linguistic than to religious reproduction. Children often acquire basic competence in a minority language from their parents and extended families, and this can be reinforced by minoritylanguage media. But without comprehensive schooling in that language – and I mean schooling with that language as the medium of instruction, not simply as the object of instruction – it is dif�cult for the minority language to be fully reproduced. Some countries with long-established, territorially concentrated linguistic minorities do provide comprehensive minority-language schooling. But even this is not suf�cient to ensure full reproduction. Minority-language populations are shrinking even where such schooling is available – as it is for 8 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 the Swedish minority in Finland or the Hungarian minority in Romania. This happens as some children opt out of minority-language school systems and as intermarriage often leads to intergenerational assimilation (Brubaker et al. 2006: 297–8, 370–1). Beyond comprehensive minority-language schooling, a linguistic regime that constrains people’s choices may be necessary to ensure the reproduction of minority languages. This is what van Parijs (2009: 163ff) has called a �linguistic territoriality regime’. An example is the Quebec policy that restricts who can attend English-language schools (and notably requires almost all new immigrants to attend Francophone schools). This underscores the crucial role of the state in linguistic reproduction. This argument might seem to be blatantly contradicted by sharp increases in linguistic pluralism in the USA and other countries of immigration, which do not provide comprehensive minority-language schooling, or other strong state support for immigrant languages. Immigration does of course generally increase linguistic heterogeneity, and the effect is intensi�ed when immigrants cluster in metropolitan areas that sustain dense networks of mothertongue institutions. But this speaks to the generation of pluralism, not to its reproduction. Continuing large-scale immigration masks substantial intergenerational linguistic assimilation. The Fishman model of language shift among secondand third-generation immigrants, set forth a half century ago (Fishman 1966; Veltman 1983), remains valid in its broad outlines. Thickening transnational ties, weakening assimilationist pressures and the growth of substantial foreign-language media markets may have slowed down the process, at least for some groups. As Alba et al. (2002) and others have shown, this is notably the case for the descendants of Spanish-speaking immigrants in the USA. But even in this group, a majority of the third generation speak only English at home.8 Huntington’s (2004) alarmist scenario of ethnonational conflict in the American southwest, based on a deepening language divide, has no basis in fact. So the reproduction of minority languages in contemporary liberal states requires a massive and ordinarily state-provided educational apparatus, and it may also require a territorial regime that limits language choice. Such arrangements are in place in some historically multilingual states, as a legacy of earlier nationalist struggles over the language of public life. Examples include Canada, Belgium, Spain, Switzerland and India. But no such arrangements protect minority languages generated by recent immigration. The various limited forms of de facto bilingualism or multilingualism that have emerged in the USA and other countries of immigration are signi�cant as pragmatic ways of accommodating linguistic pluralism, but they neither aim at, nor are capable of, reproducing that pluralism intergenerationally.9 The religious pluralism generated by immigration is more easily reproduced. Of course it is not automatically reproduced. The religious landscape of contemporary liberal societies is fluid, especially in the USA, and Language, religion and the politics of difference 9 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 I noted above the importance of conversion. But the intergenerational transmission of minority religions requires no state apparatus like a minority-language school system. And it requires no particular legal regime beyond the commitment to religious freedom that is a constitutive element of liberal polities. The transmission of religion, moreover, is not particularly costly. The transmission of a language – beyond what is simply picked up in the home – requires a major effort and carries a substantial opportunity cost. But the transmission of a religious af�liation or identi�cation does not. What is transmitted, to be sure, may be little more than a nominal religious af�liation or identi�cation (Gans 1994). But this nominal identity can later be revived or reconstructed. Some second- and third-generation Muslim immigrants in western countries are more pious than their parents or grandparents or have constructed new forms of Muslim religiosity (Roy 2004; Duderija 2007); the same has been true of many American immigrant groups (Hirschman 2004).10 The intergenerational staying power of religion results in signi�cant part from the ways in which religious practices can be flexibly adapted to changing circumstances. This has no real analogue in the domain of language. So the religious pluralism generated by immigration is more likely to be intergenerationally persistent than the linguistic pluralism so generated. Admittedly, one should distinguish between nominal and substantive religious pluralism. In the USA, immigration has sharply increased the nominal pluralism of an already pluralistic religious landscape; but at the same time, immigrant religions have become Americanised, notably by adopting prevailing congregational forms of religious organisation and worship (Yang and Ebaugh 2001). Still, among descendants of immigrants, religion offers a more enduring locus for cultural pluralism than language does.11 This is especially true in the American context, characterised by high levels of religiosity. But elsewhere too there is nothing in the domain of religion analogous to the characteristic pattern of language shift for second- and third-generation immigrants. While linguistic competences and identi�cations erode substantially across generations, religious practices and identi�cations are more likely to persist and in some cases may even grow stronger. 3. The �nal reason for the greater robustness of religious than linguistic pluralism is that religious pluralism is institutionalised and legitimated as an enduring presence in liberal societies in ways that linguistic pluralism is not. Both ideologically and institutionally, as Zolberg and Long (1999: 31) have observed, contemporary liberal states tend to be pluralist with respect to religion, and monist or assimilationist with respect to language. Their stance towards religion is an attenuated pluralism, to be sure. A more far-reaching pluralism is found in some empires and postcolonial polities, where differing systems of personal law govern members of different religious communities. But this kind of legal pluralism is �incompatible with the structural character 10 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 of modern nation-states’ (Zolberg and Long 1999: 14; see also Hirschl and Shachar 2009). Still, even this attenuated pluralism towards religion represents a sharp reversal of the historical pattern in the Christian world, in which states were strongly monist with respect to religion and pluralist with respect to language (or, more precisely, simply indifferent to linguistic diversity). Ideologically and normatively, the clearest expression of this difference in the stance of contemporary liberal states towards religion and language is that immigrants are not expected to adopt the prevailing religion, but are expected to learn the prevailing language (or one of the prevailing languages). The liberal state is expected to be neutral with respect to religion, even if it can never be fully neutral in practice; but there is no such expectation of neutrality with respect to language. Language tests for citizenship are routine, but a religious test would be unthinkable in a liberal polity. Enduring religious pluralism is not simply normatively accepted in liberal states, but institutionally supported. To be sure, as I noted above, some historically multilingual states provide strong institutional support for linguistic pluralism.12 But this strongly pluralist stance nowhere applies to immigrants. I do not mean to suggest that liberal states generally adopt harshly or even actively assimilationist stances towards immigrant languages, although there has been a shift in the last decade back towards a moderately assimilationist stance (Brubaker 2001). The point I want to underscore here is the sharp distinction, both normative and institutional, between endogenous and imported linguistic pluralism. International minority rights regimes mandate expansive protections for long-established minority languages, but only minimal protection for immigrant languages. And states that provide elaborate institutional support for historically established minority languages provide nothing comparable for immigrant languages.13 Liberal countries of immigration do of course accommodate the linguistic pluralism generated by immigration in various ways. They may provide signage, information, voting materials or bureaucratic forms in minority languages; translators in medical, legal or administrative settings; or various forms of bilingual education. But these pragmatic accommodations are categorically distinct from the comprehensive parallel school systems or regimes of territorial autonomy that seek to facilitate the long-term reproduction and preservation of multiple languages within a single state. So there is a sharp distinction between endogenous and imported linguistic pluralism. But there is no sharp distinction between endogenous and imported religious pluralism. I want to underscore this point, so let me restate it in differing form. Rights and protections for long-established minority languages are nowhere extended to immigrant languages. Linguistic settlements, in other words, are not expandable to include immigrant-borne languages. But religious settlements are expandable: not easily or automatically expandable, but expandable nonetheless. Many of the rights and recognitions enjoyed by long-established religions have been extended to immigrant religions. Liberal states have differing historically conditioned modes of accommodating Language, religion and the politics of difference 11 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 religious pluralism. But whatever their established mode of accommodation, they face nontrivial pressures to accommodate immigrant religions on similar terms. And these pressures have no counterpart in the domain of language. The most salient contemporary instance of course concerns the accommodation of Islam in northern and western Europe. It is impossible to do justice to this vexed and complex issue here. But consider just one example, from the domain of education. Accommodation on similar terms would mean providing or permitting Islamic education in public schools in countries in which religious instruction in public schools is the norm, or subsidising private Islamic schools in countries where such subsidies are provided to other private religious schools. Moves to accommodate Islam in this and other domains have been halting, uneven and controversial; many Muslims claim with considerable justice that the measures taken have not even come close to realising equal treatment. And of course one can point to spectacular counter-examples in other domains, such as the ban on the face-covering niqab that has been enacted in France and is under discussion elsewhere, and the Swiss referendum banning the construction of minarets. Yet if one looks beyond cases of highly mediatised contestation, one can see a steady if slow, contested and often grudging movement towards accommodation in the educational sphere and in other domains. This has been driven by the courts on the one hand, which have been receptive to parity claims (Koenig 2010; Joppke and Torpey forthcoming), and by a statist and securitarian concern to manage and supervise Muslim populations on the other (Laurence 2012). So to sum up this part of the argument: normative expectations, institutional frameworks and individual incentives converge in fostering a deeper and more robust religious than linguistic pluralism in liberal societies. Not simply immigration but other factors too make for increasing, persistent and institutionalised religious pluralism. Immigration generates at least as much linguistic as religious pluralism, but this migration-generated linguistic pluralism is neither intergenerationally persistent nor institutionally supported. Continuing immigration and clustered settlement patterns sustain the appearance of increasing and persistent linguistic pluralism, but an ongoing intergenerational language shift tends to prevent the consolidation of self-reproducing linguistic minorities. Religious pluralism and deep diversity Having argued that religious pluralism tends to be more robustly generated, reproduced and institutionalised than linguistic pluralism, I want to move on to argue that the accommodation of religious pluralism in liberal polities is also more likely to generate dif�cult and sometimes intractable problems of what political theorists in a line of work deriving most immediately from Rawls’ (1993) Political Liberalism (e.g. Galston 1995) have called �deep diversity’. This is obviously not true for all forms of religious pluralism. Insofar as religious pluralism involves individualised, �subjectivised’ or otherwise 12 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 privatised forms of religious experience, it is easily accommodated in liberal polities. Much of the recent pluralisation of the religious landscape in liberal societies has involved the proliferation of new forms of individualised religiosity and spirituality that do conform to the expectations of secularisation theory about the long-term privatisation and depoliticisation of religion. But as I mentioned earlier, recent decades have also witnessed a signi�cant counter-trend towards the de-privatisation and re-politicisation of religion. And I am concerned here with public, organised and collective forms of religious life, not with private, individualised forms.14 Much of the discussion of public or political religion has focused on Islam, and for good reason. Privatised and individualised forms of religiosity are more common among Muslims, especially those living in the West, than essentialist accounts of Islam as an intrinsically public and political religion would suggest (Cesari 2002). But these have been overshadowed by the centrality of various forms of public or political Islam to political contestation in both Muslim-majority and Muslim-minority settings. Public religion is of course not unique to Islam. Strong forms of public religion can be found in Christian, Jewish, Hindu and Buddhist traditions, among others. Yet the claims of public Islam pose a particularly dif�cult challenge to liberal states. But I want to keep the focus on religion and language more generally. In the era of modern nationalism, language has been widely understood as the chief criterion and main cultural substrate of nationhood. Territorially concentrated linguistic minorities have therefore been understood – by ethnopolitical entrepreneurs on the one hand, and central state elites on the other – as potential nations, and linguistic pluralism has therefore been construed as a threat to the national identity and territorial integrity of the state. Even where secession or territorial autonomy has been implausible, language conflicts have been endemic. The expansion of state employment, the introduction of universal schooling and universal male military service, and the growing importance of what Gellner (1983: chapter 3) called �context-free communication’ in an urban, mobile, literate society have made language a crucial form of cultural capital, a central focus of personal and collective identity and a key terrain of political struggle. Yet I want to argue that language conflict has lost some of its intensity and transformative potential in recent decades, as the high noon of language-based nationalist conflicts appears to have passed. The vast reorganisation of political space along national (and for the most part broadly linguistic) lines throughout Europe and Eurasia has reduced, although of course not eliminated, the scope for new language-based nationalist claims. This has involved not only the disintegration of multinational empires into linguistically more homogeneous successor states, but also the internal reorganisation of multilevel states to create linguistically more homogeneous constituent states, as in India. It also includes forms of federalism and devolution that have allowed autonomous but nonsovereign polities like Quebec, Catalonia, the Basque Country and Wales to Language, religion and the politics of difference 13 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 pursue their own language agendas. Older language-based nationalist and ethnopolitical conflicts of course remain alive, but–with some exceptions—they have become less urgent and less destabilising. In the geopolitically relaxed zones of northern and western Europe, the USA and Australia-New Zealand, states no longer seek to impose the tight coupling of culture, territory and population that was central to the nationalising projects of a century ago; linguistic diversity is not only tolerated, but in some cases even celebrated. Even in central and eastern Europe, historically the locus classicus of nationalist language conflicts, the eastward expansion of the European Union and the institutionalisation of minority-language rights have taken some of the edge off formerly intractable ethnolinguistic conflicts. In the USA, conflicts over the status of Spanish flare up periodically, focused in many cases on bilingual education or the symbolic question of an of�cial language. But more striking is the continuing piecemeal, pragmatic and largely uncontested accommodation of Spanish and other languages in a variety of less visible settings. Language continues to be a terrain of chronic struggle in multilingual polities worldwide, especially where linguistic minorities are territorially concentrated. But in liberal polities, those struggles – again with some exceptions, most obviously in Belgium – have become less intense and intractable. Yet while language conflicts have eased somewhat in recent decades, conflicts over religion have intensi�ed, driven by the resurgence of public religion. As a universal and inescapable medium of public life, language can never be fully privatised or depoliticised. Religion could in principle be fully privatised and depoliticised, but the mid-twentieth century Western vision of a fully privatised religion has proved entirely chimerical. And to the extent that religion is not privatised or depoliticised, the conflicts arising from religious pluralism tend to be deeper and more intractable than those arising from linguistic pluralism. The reasons for this are found in the most elementary differences between language and religion. Language is a medium of communication; it is not a structure of authority, and it has no intrinsic normative content. Defenders of the Sapir-Whorf hypothesis – the idea that language constitutes thought, culture and world views – would argue with this characterisation. But strong versions of Whor�anism have long been discredited, so this would be an argument only at the margins. Whatever normative content languages might carry is relatively thin.15 But religion (and especially public religion) often involves an authoritative, binding and comprehensive set of norms. These norms do not simply regulate private behaviour. They reach into the public realm, addressing such matters as gender, sexuality, family life, education, social policy, the economy and even international affairs and war.16 Gender, sexuality and family life are particularly important (and of course contested) domains of religious regulation (Friedland 2002; Casanova 2009: 17–8). Some religious norms constitute systems of law that directly and 14 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 comprehensively regulate family matters, as Jewish and Islamic norms do for marriage, divorce and inheritance. But nearly all forms of organised religion seek to regulate gender, sexuality and family life. The claims of public religion to provide binding and authoritative norms for the regulation of public and private life challenge the state’s claim to monopolise the regulation of public life (and to authoritatively regulate certain areas of private life as well). They also create conflicts with competing forms of public religion and with those segments of the public (including those who profess the same religion) who reject the claims of public religion. These are often deep conflicts of principle, involving fundamental differences of world view. It is these that warrant speaking of �deep diversity’. Language conflicts do not involve such conflicts of principle or world view. As Gellner (1983: 117–8) put it in another context, they are conflicts between people who �speak the same language’, as it were, even when they do not speak the same language. Liberal states are committed to a far-reaching accommodation of religious pluralism, but this commitment can generate quandaries. Liberal states may be obliged to accommodate forms of religion that promote illiberal ideas or practices; or they may be obliged to act illiberally in restricting religious or other freedoms in the name of other values (see e.g. Joppke 2009: 4–5; Triada�lopoulos 2011). Consider a few examples from the domain of education. Should the state exempt Christian children from exposure to �secular humanist’ views in school, as some fundamentalist Christian parents in Tennessee requested (Stolzenberg 1993)? Should it exempt Muslim children from coeducational physical education classes, as some Muslim parents in European countries have requested (see e.g. German Islam Conference 2009: 20–2)? Should it allow teachers or students to wear religious clothing, including the face-covering niqab (Joppke 2009; Joppke and Torpey forthcoming: chapter 2)? How much leeway should it grant, and what kind of �nancial or other support should it provide, to conservative religious schools (or to forms of home schooling) that cultivate ways of life at odds with the state’s interest in fostering the development of autonomous individuals and responsible citizens (Reich 2002: chapter 6)? Or consider the question that was brought into focus by the Rushdie affair in the late 1990s and revived by the Danish Cartoon affair some years later: should the state restrict potentially hurtful or offensive speech or expression so as to protect the sensibilities of members of religious communities (Parekh 2000: chapter 10)? No comparable quandaries arise in the domain of language. Conclusion Language and religion have seldom been studied together in a sustained way. To specialists in either subject, language and religion have seemed too different, while to students of ethnicity, they have seemed too similar. I have Language, religion and the politics of difference 15 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 argued that language and religion are similar enough, if construed in a certain way, to make comparison possible, yet different enough to make it interesting. As fundamental domains of cultural difference, language and religion have much in common. Both are ways of identifying oneself and others, of construing sameness and difference. In Bourdieusian language, both are basic principles of vision and division of the social world. Both divide the world, in popular understandings, into distinct, bounded and self-reproducing communities. And claims are made in the name of both kinds of communities for recognition, resources and reproduction. These and other similarities have led students of ethnicity to treat language and religion as functionally equivalent, and as theoretically uninteresting forms of �cultural stuff’, signi�cant primarily as grist for the mill of ethnic classi�cation and boundary formation. But this perspective is flattening. It neglects important differences in the social organisation and political expression of language and religion in liberal societies and polities. Language is an inescapable medium of public as well as private life; religion is not. The state must privilege a particular language or set of languages, but it need not privilege a particular religion. Language is chronically and pervasively politicised in the modern world, while much of religion has become privatised and depoliticised. Yet deprivatisation is an important counter-trend, and the claims of public religion to authoritatively regulate public and private life have no counterpart in the domain of language. Immigration generates new forms and degrees of both linguistic and religious pluralism, but the religious pluralism generated by immigration is more intergenerationally robust and more deeply institutionalised than the linguistic pluralism. The result is that religion has tended to displace language as the cutting edge of contestation over the political accommodation of cultural difference in Western liberal democracies – a striking reversal of the longer-term process through which language had previously displaced religion as the primary focus of contention. Acknowledgments I would like to thank Kristen Kao and Amy Zhou for research assistance, and Regina Bendix, John Bowen, David Gellner, Phil Gorski, Ann Swidler and Andreas Wimmer for helpful comments. I am particularly indebted to Matthias Koenig and Peter Stamatov for their detailed and insightful comments on an earlier draft, and to Susan Ossman, with whom I talked through much of the argument. Notes 1 For an alternative perspective on the commonalities of religion and language, see Safran (2008). 2 On language, see e.g. Haugen (1966); on religion, Beyer (2001). 16 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 3 To analyse these trajectories is beyond the scope of this paper. From the large literature, see illustratively Kaplan (2007) and Madeley (2003) on religion, and Hobsbawm (1990) and Barbour and Carmichael (2000) on language. 4 Whether or not one accepts Chomsky’s notion of a universal grammar, language is universal not only in the sense that it is found in all human societies, but also in the sense that all humans ordinarily develop pro�ciency in at least one language. This has no clear analogue in the domain of religion. While everyone may have a capacity for religious experience (depending on how this is de�ned), it cannot be said that everyone is pro�cient in at least one religion. Quite apart from explicit irreligiosity or anti-religiosity, people differ much more widely in religiosity than in basic �linguosity’. Of course people differ substantially in linguistic pro�ciency or linguistic capital as well, but the differences in religious quali�cation are much greater and more consequential; they are among other things the basis for the opposition between virtuoso and mass religiosity that is central to Weber’s sociology of religion. 5 In traditions of virtuoso religiosity, religion may be normatively understood as pervasively relevant to all aspects of life, but that does not make religion a universal medium of social life in the sense that language is. 6 On the broader debate in political theory about the ideal of state neutrality vis-à-vis competing understandings of the good, see Koppelman (2004). Koppelman acknowledges the incoherence of the idea of complete neutrality, yet he af�rms the continued relevance and value of neutrality as an ideal. 7 This is obviously a gross generalisation, and exceptions spring immediately to mind. Still, it and the equally gross generalisation in the preceding paragraph hold up pretty well for the Western world on a time scale of decades and centuries, respectively. But one would need to shift to another level and mode of analysis altogether, of course, to account for the varying contexts, contours and trajectories of conflicts over language and religion in particular places and times. 8 Merle et al. (2010) present their study as a challenge to Fishman, but their �nding that only a third of migrants’ grandchildren in Basel and Geneva understand or speak the language of their grandparents seems broadly consistent with Fishman. 9 This is not to minimise the thoroughgoing institutionalisation of Spanish in the USA, anchored not only in government policies but also, as was already evident to Zolberg and Long (1999: 26), in a substantial media market. Yet as noted in the text, the process of intergenerational language shift continues among the children and grandchildren of Spanish-speaking immigrants. 10 The literature that points to new or intensi�ed forms of religiosity among some second- or third-generation Muslim immigrants is largely ethnographic (see e.g. Glynn 2002). Intensi�ed religiosity may be highly visible, but there is no evidence that it is broadly representative. Quantitative studies have reported intergenerational stability in levels of religiosity (Diehl and Koenig 2009, examining Germany), or slight intergenerational declines (Ersanilli and Koopmans 2009, considering Germany, France and The Netherlands). Kashyap and Lewis 2012, interestingly, �nd both decreased observance among younger British Muslims and an increased salience of Islam for personal identity. 11 This point was already made by Herberg (1960), and has more recently been emphasised by Warner (1993) and Kurien (1998). 12 In many of these cases, however (e.g. in India, Switzerland, Belgium and Canada, or at least Quebec), linguistic pluralism on the statewide level coincides with linguistic monism – or at least with the strongly institutionalised primacy of a single language – at the level of federal component states or provinces. Linguistic pluralism, in other words, generally exists as a collection of lowerlevel linguistic monisms. This observation supports the argument of Zolberg and Long that modern states (or at least their component substate polities) tend towards monism in the domain of language. 13 Political theorists are divided about the justice of this sharp difference in the treatment of long-established and recently imported linguistic pluralism; see e.g. Kymlicka (1995); Patten (2006). Language, religion and the politics of difference 17 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 14 Deprivatisation and ongoing privatisation are not mutually exclusive; given the complexity of the contemporary religious landscape, it is not surprising that both are happening at the same time (Casanova 2009: 29). 15 To underscore the relative normative and cultural �thinness’ of language vis-à-vis religion is not to deny that language may carry �thicker’ cultural meanings and commitments in some contexts than in others. See Carens (2000: 128–9) and Bauböck (2002: 177–8) on �thin’ and �thick’ theories of language in relation to cultural meanings and commitments. 16 The meanings of and boundaries between �public’ and �private’, to be sure, are richly ambiguous and chronically contested (Casanova 1994: chapter 2). References Alba, R. 2005. �Bright vs. blurred boundaries: second generation assimilation and exclusion in France, Germany, and the United States’, Ethnic and Racial Studies 28, 1: 20–49. Alba, R. et al. 2002. �Only English by the third generation? Loss and preservation of the mother tongue among the grandchildren of contemporary immigrants’, Demography 39, 3: 467–84. Anderson, B. 1991. Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso. Bader, V. 2007. Secularism or Democracy: Associational Governance of Religious Diversity. Amsterdam: Amsterdam University Press. Barbour, S. and Carmichael, C. 2000. Language and Nationalism in Europe. New York: Oxford University Press. Barth, F. 1969. �Introduction’ in F. Barth (ed.), Ethnic Groups and Boundaries: The Social Organization of Cultural Difference. London: George Allen & Unwin. Bauböck, R. 2002. �Cultural minority rights in public education, religious and language instruction for immigrant communities in Western Europe’ in A. Messina (ed.), West European Immigration and Immigrant Policy in the New Century. Westport, CT: Greenwood Press. Beyer, P. 2001. �What counts as religion in global society?’ in P. Beyer (ed.), Religion in the Process of Globalization. Würzburg: Ergon Verlag. Bloch, M. 1996. �Religion and ritual’ in A. Kuper, J. Kuper (eds.), The Social Science Encyclopedia. London: Routledge. Brubaker, R. 2001. �The return of assimilation?’ Ethnic and Racial Studies 24, 4: 531–48. Brubaker, R. et al. 2006. Nationalist Politics and Everyday Ethnicity in a Transylvanian Town. Princeton, NJ: Princeton University Press. Carens, J. 2000. Culture, Citizenship, and Community. Oxford: Oxford University Press. Casanova, J. 1994. Public Religions in the Modern World. Chicago, IL: University of Chicago Press. Casanova, J. 2009. �Religion, politics and gender equality: public religions revisited’, in A Debate on the Public Role of Religion and its Social and Gender Implications. Gender and Development Programme Paper Number 5. Geneva: United Nations Research Institute for Social Development. Cesari, J. 2002. �Islam in France: the shaping of a religious minority’ in Y. Haddad-Yazbek (ed.), Muslims in the West, from Sojourners to Citizens. Oxford: Oxford University Press. Cornell, S. 1996. �The variable ties that bind: content and circumstance in ethnic processes’, Ethnic and Racial Studies 19, 2: 265–89. Diehl, C. and Koenig, M. 2009. �Religiosität türkischer Migranten im Generationenverlauf: Ein Befund und einige Erklärungsversuche’, Zeitschrift für Soziologie 38, 4: 300–19. Duderija, A. 2007. �Literature review: identity construction in the context of being a minority immigrant religion: the case of Western-born Muslims’, Immigrants & Minorities 25, 2: 141– 62. 18 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 Ersanilli, E. and Koopmans, R. 2009. �Ethnic retention and host culture adoption among Turkish immigrants in Germany, France and The Netherlands: a controlled comparison’, Working Paper, Social Science Research Center, Berlin. Fishman, J. 1966. �Language maintenance in a supra-ethnic age: summary and conclusions’ in J. Fishman et al., Language Loyalty in the United States. The Hague: Mouton. Friedland, R. 2002. �Money, sex, and god: the erotic logic of religious nationalism’, Sociological Theory 20, 3: 381–425. Galston, W. A. 1995. �Two concepts of liberalism’, Ethics 105, 3: 516–34. Gans, H. J. 1994. �Symbolic ethnicity and symbolic religiosity: towards a comparison of ethnic and religious acculturation’, Ethnic and Racial Studies 17, 4: 577–92. Geertz, C. 1963. �The integrative revolution’ in C. Geertz (ed.), Old Societies and New States: The Quest for Modernity in Asia and Africa. New York: Free Press. Gellner, E. 1983. Nations and Nationalism. Ithaca, NY: Cornell University Press. Gellner, E. 1997. �Reply to critics’, New Left Review 221: 81–118. German Islam Conference. 2009. �Interim résumé by the working groups and the round table’, Paper for the 4th Plenary Session of the German Islam Conference. Berlin, 25 June 2009. http://www.deutsche-islam-konferenz.de/SharedDocs/Anlagen/EN/DIK/Downloads/Plenum/ DIK-viertes-Plenum-en,templateId=raw,property=publicationFile.pdf/DIK-viertes-Plenumen.pdf. Accessed on 4 May 2012. Glynn, S. 2002. �Bengali Muslims: the new east end radicals?’ Ethnic and Racial Studies 25, 6: 969–88. Gorski, P. S. and Altinordu, A. 2008. �After secularization?’ Annual Review of Sociology 34: 55–85. Haugen, E. 1966. �Dialect, language, nation’, American Anthropologist 68, 4: 922–35. Herberg, W. 1960. Protestant, Catholic, Jew: An Essay in American Religious Sociology. Garden City, NY: Anchor Books. Hirschl, R. and Shachar, A. 2009. �The new wall of separation: permitting diversity, restricting competition’, Cardozo Law Review 30: 2535–60. Hirschman, C. 2004. �The role of religion in the origins and adaptations of immigrant groups in the United States’, International Migration Review 38, 3: 1206–33. Hobsbawm, E. J. 1990. Nations and Nationalism since 1780. Cambridge: Cambridge University Press. Horowitz, D. L. 1985. Ethnic Groups in Conflict. Berkeley, CA: University of California Press. Huntington, S. P. 2004. �The Hispanic challenge’, Foreign Policy 141: 30–45. Jenkins, R. 1997. Rethinking Ethnicity: Arguments and Explorations. London: Sage. Joppke, C. 2009. Veil: Mirror of Identity. Cambridge: Polity. Joppke, C. and Torpey, J. Forthcoming. Legal Integration of Islam: A Transatlantic Comparison. Cambridge, MA: Harvard University Press. Kaplan, B. J. 2007. Divided by Faith: Religious Conflict and the Practice of Toleration in Early Modern Europe. Cambridge, MA: Harvard University Press. Kashyap, R. and Lewis, V. A. 2012. �British Muslim youth and religious fundamentalism: a quantitative investigation’, Ethnic and Racial Studies. iFirst http://dx.doi.org/10.1080/ 01419870.2012.672761 Accessed on 12 September 2012. Koenig, M. 2010. �Gerichte als Arenen religiöser Anerkennungskämpfe’ in A. Reuter, H. G. Kippenberg (eds.), Religionskonflikte Im Verfassungsstaat. Göttingen: Vandenhoeck & Ruprecht. Koenig, M. 2011. �Jenseits des Säkularisierungsparadigmas’, Kölner Zeitschrift für Soziologie une Sozialpsychologie 63, 4: 649–74. Koppelman, A. 2004. �The fluidity of neutrality’, The Review of Politics 66, 4: 633–48. Kurien, P. 1998. �Becoming American by becoming Hindu: Indian Americans take their place at the multicultural table’ in S. R. Warner, J. G. Wittner (eds.), Gatherings in Diaspora. Philadelphia, PA: Temple University Press. Kymlicka, W. 1995. Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Oxford University Press. Language, religion and the politics of difference 19 © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 Laurence, J. 2012. The Emancipation of Europe’s Muslims. Princeton, NJ: Princeton University Press. Madeley, J. 2003. �A framework for the comparative analysis of church-state relations in Europe’, West European Politics 26, 1: 23–50. Martin, D. 2005. On Secularization: Towards a Revised General Theory. Aldershot: Ashgate. May, S. 2001. Language and Minority Rights: Ethnicity, Nationalism, and the Politics of Language. New York: Longman. Merle, M. et al. 2010. �De la langue d’origine á la langue héritée: le cas des familles espagnoles á Bâle et á Genéve’, Travaux neuchâtelois de linguistique 52: 9–28. Parekh, B. 2000. Rethinking Multiculturalism: Cultural Diversity and Political Theory. Cambridge, MA: Harvard University Press. Patten, A. 2006. �Who should have of�cial language rights?’ Supreme Court Law Review 31, 2: 101–15. Patten, A. and Kymlicka, W. 2003. Language Rights and Political Theory. Oxford: Oxford University Press. Rawls, J. 1993. Political Liberalism. New York: Columbia University Press. Reich, R. 2002. Bridging Liberalism and Multiculturalism in American Education. Chicago, IL: University of Chicago Press. Rothschild, J. 1981. Ethnopolitics: A Conceptual Framework. New York: Columbia University Press. Roy, O. 2004. Globalized Islam: The Search for a New Ummah. New York: Columbia University Press. Safran, W. 2008. �Language, ethnicity and religion: a complex and persistent linkage’, Nations and Nationalism 14, 1: 171–90. Stolzenberg, N. M. 1993. � “He drew a circle that shut me out”: assimilation, indoctrination, and the paradox of liberal education’, Harvard Law Review 106: 581–667. Taylor, C. 2007. A Secular Age. Cambridge, MA: Harvard University Press. Torpey, J. 2010. �A (post-) secular age?’ Social Research 77, 1: 269–96. Triada�lopoulos, T. 2011. �Illiberal means to liberal ends? Understanding recent immigrant integration policies in Europe’, Journal of Ethnic and Migration Studies 37, 6: 861–80. Turner, B. S. 2011. Religion and Modern Society. New York: Cambridge University Press. Van Parijs, P. 2009. �Grab a territory! How equal linguistic dignity can be reconciled with English Dominance in the European Union’ in J. Poirier et al. (ed.), Ties That Bind: Accommodating Diversity in Canada and the European Union. Bern: Peter Lang. Veltman, C. 1983. Language Shift in the United States. Berlin: Walter de Gruyter. Warner, S. R. 1993. �Work in progress: toward a new paradigm for the sociological study of religion in the United States’, American Journal of Sociology 98, 5: 1044–93. Yang, F. and Ebaugh, H. R. 2001. �Transformations in new immigrant religions and their global implications’, American Sociological Review 66, 2: 269–88. Zolberg, A. R. and Long, L. W. 1999. �Why Islam is like Spanish: cultural incorporation in Europe and the United States’, Politics and Society 27, 1: 5–38. 20 Rogers Brubaker © The author(s) 2012. Nations and Nationalism © ASEN/Blackwell Publishing Ltd 2012 Copyright of Nations & Nationalism is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.