GLOBAL MIGRATION AND THE SELECTIVE REIMAGINING OF RELIGIONS

: It is a commonplace in discussions of immigrant religion to speak of how religion aids in the adjustment of migrants to a new culture and society; how it serves as a dimension of continuity in the process of integration. This article examines theo-retical foundations for reconsidering this perspective in the context of globalization in general and global migration in particular. In a global society, it is far less useful to think of migrants as leaving one society to join another, especially insofar as this optic tends to assume a) that the new “host” culture remains comparatively unaffected while the immigrants culture faces the dilemma of assimilation versus ethnic preservation; and b) that the culture of origin simply loses a few members without much effect by the migrants back onto their cultures of origin. By contrast, the article argues that the consequences of migration are to help (re)deﬁ ne religions in all areas where they are represented; and thus to make distinctions between “core” and “diaspora” far less salient. Instead, different areas where religious traditions are represented are better seen as centres for creating different options for the authentic construction of the same religion; options that are very often in communication with each other.

Palavras-chave: cultura, globalização, migração, religião. I There are two senses in which this article is about the formation and use of concepts. One is sociological in that I wish to reexamine how certain notions such as religion and ethnicity operate within sociological observation and analysis. The second sense has to do with how the people whom we describe with these concepts use them and understand them. One can look at this as two levels of observation: when people describe themselves as, say, Christians or Buddhists, or as Arab or Bengali, that is a fi rst order observation. Sociologists, when they observe how these same people observe themselves, are engaged in second order observations: they observe how others observe (cf. Luhmann, 1995a); and are, of course, themselves subject to further repetition of this refl exive operation. Aside from recalling Weber's notion of "verstehende Soziologie" (Weber, 1978), such an approach places emphasis on the critical role that observation has in creating the social realities in which we live. Although observing may seem to be a rather passive thing to do, it is also a critical aspect of all active social production and reproduction.
The specifi c subject of my observations in this article is global migration, in particular its consequences for the idea and social reality of religion and religions. I understand migration as the relocation of people from one part of the world to another on something other than a very temporary basis. One might therefore contrast it with the more fl eeting notion of "travel", as long as it is understood that migration can still include a good deal of moving back and forth. The reasons for this deliberate ambiguity will become clear as I proceed. The modifi er, global, further restricts my theme to migrants who move large distances and beyond their countries of origin, thus excluding very important phenomena such as urban migration and refugee movements to contiguous territories. The consequences of these related phenomena are in certain respects similar to those of global migration, but also different enough to warrant separate treatment. Within those restrictions, my main thesis is as follows: migration in today's global society is both more complex and simpler than it may have been in the past. It is more complex because the separations between the "old" country and the "new" are neither neat nor complete. Signifi cant communication continues to occur between the two places and that has consequences for social life in both of them. As a result, the adjustments and changes that migrants and host populations make are less predictable and more fl uid because of this more equivocal difference between "home" and "away". For the same reasons, migration is also simpler because the continuity of these two places makes the change at least seem less radical. Globalization processes have made home more like away and away more like home: the difference between here and there makes less of a difference. In this context, religion can and does often play an important role for expressing the continuity and the change. As part of this process, religion also undergoes change: migration helps to form and reform religion, not only in the diaspora, but also at home. It is the two-way process that is of primary concern.
What follows focuses only on the elaboration and clarifi cation of my thesis. This has three parts. A fi rst section discusses the context of global migration, that is, globalization and the formation of a single global society. Migration and its consequences are different because they occur in a different social context than has prevailed in the past. A second section then looks at the situation of migrants as such, and in particular at the role religion, among other concepts, plays in their self-defi nitions. On this basis, a third section then examines the effects that global migration has had, is having, and will likely continue to have in the place and form of religions in contemporary global society more generally. Migration contributes to the globalization of communication, and of religious communication specifi cally.

II
The social-scientifi c discussion about globalization is by now quite welldeveloped, with different versions and different emphases. My own position owes a fair amount to the contributions of Roland Robertson (1992), John W. Meyer (Thomas et al., 1987), Niklas Luhmann (1975Luhmann ( , 1995b, and the scholars associated with them and their work; although other infl uences are not absent. I see globalization as the historical process that has lead to the formation of a single global society. The most determinative structural features of this society are a series of differentiated, instrumentally oriented social systems, the most powerful of which are a capitalist economic system, a political system of sovereign states, and an empirically-based science system. Other such systems include ones for health, education, art, law, mass media, sport, and religion. These function systems, while dominant, are by no means the only important structures or typical features of global society. Among the consequences of their development and global spread is what Anthony Giddens (1990) has called the "disembedding" of people and cultures from previous social contexts; which is a way of saying that the traditional structures in which people lived have been undermined, if not completely displaced. One might say, metaphorically, that in global society, everyone is a "migrant" to the extent that the old "homes" have disappeared or had their foundations removed. This raises the obvious question of how people in global society "reembed" themselves in the context of the new societal structures.
The simple answer to this question is that the possibilities for "re-embedding" are quite varied because the dominant function systems do not incorporate people in any complete way. One can say that the systems are highly impersonal in that what matters in each of them is not who you are, but what you do. This, in turn, leads to a high level of individuation as we come to expect people to be responsible for the construction of their own lives and selves, and that they should be able to do so in an undetermined variety of ways. The resulting pluralism has not just individual, but also collective manifestations in that individuals are "free" to identify themselves with groups along various lines of defi nition; and those groups can and do claim a similar legitimacy.
Two persistent problems arise out of this situation, however. The fi rst is that, because the dominant function systems are effectively impersonal and are dependent to a large degree on a concentration of communicative resources (that is, power, infl uence, wealth, knowledge, and so forth), they do not actually treat people equally. Quite the opposite is and must be the case. Yet given that the capacity to construct one's life or identity depends on having access to the benefi ts of these systems, many and, some would argue, even most people in global society do not have the resources to do so. They have no practical choice. We therefore have the expectation that people should be able to live as they choose, but with the constant realization that so many of them cannot. The question of identity then becomes a question of power: the quest for identity (or, morally speaking, dignity) becomes a quest for power; and the struggle for power can be couched in terms of individual and collective identity. Another aspect of this problem is that, even for those who do have power and resources, the choices are still circumscribed by the power of the systems, resulting again in the impression that we have little choice, which is a refl ection of the expectation that we should have. Thus, while marginalized people are oppressed and excluded, the rest of us still may seem subject to, as Habermas puts it, the "colonization of our life-worlds".
The second problem that arises out of the combination of function system dominance and individuation/pluralization of identities is that all these choices or possibilities appear exactly as that: choices that could have been effected otherwise. To the degree that we seek to overcome the manifest contingency, insecurity, or technical rationality of the function systems through recourse to personal or cultural visions that are purportedly "natural", "self-evident", "absolute", or substantively rational; to that degree we expose ourselves rather constantly to comparison with others and other visions that obviously do it differently. What we believe to be constant appears as changeable in a situation where only the means are not optional. The question then becomes one of how we "do" pluralism, how we manage to let the incommensurate live side by side without either deciding for one of them or dissolving all of them? Past societies solved this problem by embedding people more completely in systems differentiated according to hierarchy or location. Modern global society has this option to a far lesser degree, leading almost paradoxically to an increase in the overt claims (that is, self-observations) of "natural" identity.
The two problems, that of exclusion and that of relativization are, of course, in various ways related, for instance, insofar as important personal and group characteristics are still very close to being ascribed as opposed to chosen or achieved. One thinks, for instance, of sex or the colour of one's skin. We can and do, therefore, get very ambiguous situations in which people use what appear to be the unavoidable reasons for their exclusion as the basis of their "natural" identities. The bases of exclusion, say sex or race, become the bases of exclusivities that demand inclusion as such; or in their more extreme forms, seek separation. Overall, the question of "re-embedding" elicits a wide variety of responses, all of which are subject to the relativization of comparison with others; but at the same time, the context encourages the construction of identities that claim to transcend contingency. Moreover, the primary way to assert the latter is not through separation or isolation, the possibilities for which in global society are exceedingly limited. Rather, it is through resort to the dominant systems and their typical by-products. A good example presents itself in the various technological communication media, in particular the mass media: on the one hand, these media so increase communication around the globe that the relativization of comparison becomes unavoidable. On the other hand, these same media allow the repeated and highly selective communication of visions to highly selective audiences, making the formation of particular identities that much easier.
The upshot of all this complexity is that the construction of selves and group identities in global society does not happen just in contrast to or in opposition to the dominant systems. They are not just anti-systemic protests or reactions against the consequences of functional system dominance because the assertion of an identity requires access to the communicative power of the systems; not all of them, but most often more than one of them. The radical retreat from or rejection of global society and its dominant structures is a real, but also very limited possibility: ecological communes in the wilderness and hyper-isolationist political regimes are possible, but they have to use a disproportionate share of their resources just to keep the outside out. In moving on, now, to a consideration of religion and global migration, the mutuality of identities and systems has to be kept in mind.

III
Global migration as such is, of course, not new to the contemporary era. It has been a feature of most epochs and might even be said to be one of the constants of world history. What is new and quite recent is the extent to which this migration engenders regular and reciprocal communicative ties between different parts of the world. One can go so far as to say that, in the context of global society, global migration is, at one level, merely another manifestation of global communication links. The South Asian migrant to Europe or the African migrant to North America, whatever else they do, establish links between the place of origin and the new home. Postal service, telephone, air travel, and more recently computer links make regular contact both possible and probable. Combined with mass media, worldwide trade (including remittances of migrants), global sport, and other less personal forms, these technologies allow a much higher degree of contact between places than was possible in the past; not only in that any particular migrant can communicate back and forth more easily, but also because those who are not migrants use these same means. Not all migrants, of course, have such links, but I would suggest most do and a very large number use them. The globally extended communication of global migrants is signifi cant and not confi ned to certain privileged elites among them.
The context of richer communicative connections between the sources and destinations of migration leads directly into the question discussed in the previous section: that of identity in a relativizing social context. This question now has to be asked for three groups of people, namely the migrants themselves, their new hosts, and their old compatriots. All three stand to be affected by the fact that the migrants have left "home". I put this last word in quotations because the new situation implies what one can call a disjunction of place and identity, or at least the necessity of reconstructing a presumed unity of place and identity. Who we are and where we are now overlap far less or far less self-evidently. The result is more complex possibilities for who one becomes as a migrant and for the categories that one uses to express these identities. This statement requires some elucidation.
Until relatively recently, arguably the most common model for understanding the consequences of migration operated, often implicitly, on three assumptions that we must, not so much question, as unpack and reformulate. 1

Global migration and the selective reimagining of religions
The fi rst is that the migrant arrives at their new destination with their identity already formed and the only question is the degree to which they will have to adopt a new identity and therefore lose the old one. Migrants are who they are and so we only have to ask how they will change. The second is that the only problem for the new hosts is whether and how they can "integrate" the newcomers, presumably to become a variation of what the hosts already are or to remain separate and marginalized. There is relatively little question of the hosts having to reformulate their identities unless the migration is so massive as to constitute a colonization of the place in question. In which case, as with aboriginal peoples, the question is reversed: the hosts become the migrants even though they did not move. The third assumption is that the consequences for people the migrants left behind in the "old country" are limited to those connected with their simple absence. We may worry about "brain drain" or the economic consequences of migrants' remittances; but the "home" remains the centre of cultural authenticity, unaffected in this legitimacy by what happens in the migrant "diaspora".
The vast and complex more recent literature on migration does not, of course, make these assumptions without qualifi cation. Questions of "transnational" networks have of late received some considerable attention, although the formulation of what precisely this means varies a great deal (see Basch et al., 1994;Levitt, 1998). Moreover, all three are valid up to a point, and insightful research can still be done on their basis. Yet to the degree that we use a model which sees global migration essentially as the injection of members of one society into another different society, these assumptions are inadequate because they do not take into account the extent to which global migration is a phenomenon internal to a single global society. The density of global communication, I would argue, makes it fruitful at least to include this latter and separation or resistance. See from among a great many examples, Baub'ck et al. (1996); Driedger (1989); Kim and Gudykunst (1988). A good example of the nuanced persistence of this model is to be found in the current American New Ethnic Immigrant Congregation Project directed by Stephen Warner. See Warner (1997). Here, as I make clear in the text, I do not so much negate this sort of work as seek to set it in a larger global context. One area where the older assumptions are being seriously challenged, and this in the context of rethinking an entire discipline, is in some of the anthropological literature with its talk of "de-essentializing" ethnicity specifi cally, and identity more generally. See, for instance in this regard, Baumann and Sunier (1995); Benmayor and Skotnes (1994). Some anthropologists also address the globalization issue, for instance with talk of "transnationalism". See, e.g. Freston (1997); Levitt (1998);Marshall Fratani (1997); van Dijk (1997) for some very recent examples. See also, in this regard, Basch et al. (1994). starting point, if not abandon the former. One could examine what this means more precisely from various perspectives. Here I concentrate on aspects of the religious dynamics that are frequently involved, and this under three headings: the use of ethnic and religious categories for formulating migrant identities; the phenomenon of, in a broad sense, missions; and the existence of and relations with "indigenous" converts.
Personal and group identities do not simply exist as objective characteristics. They are constructed self-descriptions which serve to control action. The question of "who am I?" or "who are we?" is closely connected to the question of "what shall I or we do?" Moreover, identity and action condition each other mutually, not unidirectionally: at any given moment, it may be just as accurate to say that "I am what I do" as it is to say that "I do what I am". The awareness of such refl exivity is a symptom of the relativization that I discussed in the previous section; and, according to some, a hallmark of contemporary global society (see, e.g. Beck et al., 1994). In addition, just as our actions do not fl ow simply out of our own decisions, the construction of self has at least as much to do with what those around us say we are as it does with our own determinations. The question of attribution can be quite ambiguous: are we seeing ourselves to conform to how others see us? Are we seeing ourselves as we decide we are? How can we tell the difference?
What all this means for global migrants is that their self-observations, much like their life-worlds in the new contexts, can and will vary a great deal. They can vary from person to person, from group to group, and over time. This will reveal itself, for instance, in the categories of identity used, what those categories mean, and in what institutional form they come to be expressed. The examples of the categories of ethnicity and religion can serve to illustrate.
Ethnicity, as a category, can have a number of defi ning features: it can mean a common cultural origin, a common country from which the last migration took place, common language, common region of origin, common race, common religion; it can mean common familial, clan, tribal, caste, or other such origin. Thus, for instance, migrants to Canada who are, according to the Canadian government census bureau (outside observers), of "South Asian" origin may see themselves as just that, South Asians; but they could also divide themselves according to language, place of migration (e.g. Trinidad, East Africa, Britain, India), caste or jati, Indian state origin (e.g. Gujarat, Maharashtra, Tamil Nadu), country of ancestral origin (e.g. Sri Lanka, India, Pakistan, Bangladesh), religion (Muslim, Hindu, Buddhist, Christian, etc.), and so forth for other lines of potential cleavage. With all these possibilities, the question is not just how such migrants might identify themselves, but more importantly according to what categories they will build their institutions: for instance, their associations, organizations, media, social networks, schools, and religious establishments. These latter are in a real sense the ways people "do" their identities, their practical expression. In all these cases, the actual, practical identifi ers will favour certain possibilities as germane, and leave the rest as secondary categories. That includes possible disjunction between what things are named and how others see them: a school, for instance, may present and structure itself as Hindu, but be seen by others as north or south Indian in practice; a temple may be Vaishnavite in name, but regarded by other Hindus and Vaishnavites as Guyanese in practice.
Looked at in this way, religion for global migrants is just one possible locus of identifi cation and organization among others. If, as second order observers, we conceive the term "ethnicity" broadly enough, then religion is a dimension of ethnicity that may or may not be used to express ethnic identity as a fi rst order self-observation. As I indicated in the previous section, however, religion in today's global society is also the locus of a globalized and function-oriented social system; it is not just a possible dimension of group identity, an aspect of collective cultures. This implies a tension between these two senses of religion, between a universal and a particular meaning of religion. Religion can be and is in one sense simply part of a group culture; but it itself represents a cultural complex that can be and is independent of any group culture that identifi es itself on some other basis than religion. Put more concretely, Vietnamese can be Buddhists, for example, but not all Buddhists are Vietnamese and not all Vietnamese are Buddhists. That implies at least the possibility of something called Buddhism which is independent of ethnic or linguistic designations; in which case, we have to ask what that could be. In somewhat different words still, whether a particular religion, such as Hinduism or Buddhism or Daoism, comes to be recognized as a separate something is not a question of whether each simply exists, but whether or not people will treat or construct or imagine each of them as something separable and separate. And beyond that, what each of these religions will become is itself the product of social negotiation: "within" each religion, there will be an ongoing "dialogue" regarding what is authentic, what is orthodox, what belongs within and what is beyond the pale.
An important aspect of this latter process that can help us understand further the complexity of these migrant situations is the phenomenon of missions, that is, the deliberate spreading of religions; and within the context of such missions, the existence of "indigenous converts", which is to say converts that are deemed not to belong to the migrant group. From one perspective, global migration is an aspect of mission to the extent that migrants carry their religious expressions with them; and inasmuch as certain migrants move abroad for the purpose of spreading religion. The migrant guru or evangelist is just as much a part of global migration as are others. An important question in this regard, however, is the object of such missions. One can divide them into two groups: There are those ethnically identifi ed with the religion in question, usually other migrants of fi rst or subsequent generations; and there are those who have no previous connection with the religion concerned, most often people of the "host" culture(s). The consequences for both the migrant groups and the religion in question can be quite varied and also potentially important. Four illustrations will concretize these implications. I select three North American examples, black Muslims, ISKCON, and Buddhist groups; and one more global movement, Christian Pentecostalism. Many more examples could be used, both from the North American context and elsewhere, but these can serve adequately the purposes of illustration.
The American phenomenon of those African-Americans who have identifi ed themselves more or less with orthodox versions of Islam as dominate in the Muslim heartlands of the Middle East is of peculiar importance because these Muslims, although "indigenous" converts, probably constitute a larger group than Muslims who have arrived in the United States as recent migrants from other places. Moreover, it is not at all rare to see indigenous and migrant Muslims belong to the same congregations (see Smith, 1993Smith, , 1994. Unlike with the other two examples I cite here, there is no consistent organizational segregation between the two varieties. Evidently, this has much to do with the fact that Islam has always been a missionizing religion; but probably more salient is the sense that many African-American Muslims have, that their Islamic identity makes them part of a more universal movement where precisely race, the prime characteristic of their exclusion in American society, is at least offi cially and frequently practically irrelevant. Here we have one of those many examples where identity and power are intimately related. In addition, we must note that this movement of indigenous conversion to Islam would in all likelihood have been impossible without the signifi cant and simultaneous migration of Muslims from other parts of the world to the United States; and without the communication links discussed above which allow continuous contact and communication with non-migrant Muslims in those other parts of the world. Islam in the United States is therefore neither just an immigrant religion nor an indigenous one; nor is it simply both, side by side. Instead, what it is and what it will become has to do with events and developments both within the United States and elsewhere: it is not just a case of Islam "inculturating" or "adapting" to American culture, but an aspect of wider, indeed global, transformations as well. In due course, we should expect American Islam to serve as both a dependent and independent model of what Islam in today's global society can look like. Migration and global communicative links are critical to this development. Turning to the Hindu example of ISKCON, this movement is a product of global migration in that its founder deliberately moved to the United States to propagate his version of Bengali Vaishnavism. Since its founding, the movement has spread to other parts of the West, for instance Britain and Israel. It has also established itself in India, thus representing a religious product of global migration that has re-imported itself into the "homeland" of this religion. The vast majority of its adherents are still Western converts and not South Asians. Since the death of its founder in 1977, the movement has suffered a certain amount of schism, but what one might call the main movement has attempted to remained solidly within "orthodox Hinduism", implying, of course, that there is such a thing. The existence of ISKCON in fact raises the question of the identity of Hinduism in a quite practical fashion: whether or not the movement becomes accepted as a part of Hinduism has everything to do with how it is observed, by insiders and outsiders (cf. Carey, 1987). And the outcome of those observations will help to determine what Hinduism is.
Although it is an almost insignifi cantly small movement, it poses with particular clarity the question of religion as separate system versus religion as dimension of group or ethnic culture. For most global migrants, directly or indirectly from South Asia, Hindu religion is not a missionizing religion, but a dimension of their culture: to become a Hindu one has already to be a Hindu. Yet, just as the presence, through global migration, of ethnically Turkish people in Germany or North Africans in France challenges these national designations to redefi ne themselves away from their current ethnic or "blood line" emphasis, so ISKCON at least raises the question of whether Hinduism -wherever its adherents live -can long continue as only ethnically defi ned religiosity. Without global migration and its consequences, such questions could hardly have been raised. For the global migrants especially, ISKCON eliminates the easy possibility of Hinduism being a self-evident identifi er of South Asian migrant groups, for themselves and their "hosts". One of the surer possibilities for self-observation in the new "home" becomes equivocal. Still, it must be said that at the moment such a challenge is only a possibility, because, for the most part, unlike with Islam in the United States, the two types of Hinduism (one ethnic, one open and missionizing) receive expression in different organizations and institutions.
The example of Buddhist groups presents another variation on this theme. Here we have a religious tradition that historically has been missionizing, a trend which continues into the contemporary era. Yet, because Buddhism has also been a or the main religious expression of certain regions of the globe, notably Southeast and East Asia, it is also a potential identifi er for migrant populations. Refl ecting this double aspect, North American Buddhists are for the most part "ethnic", for instance Vietnamese, Chinese, Sinhalese, or Japanese Buddhists. These ethnic varieties maintain their separate institutions, especially as much of the migration to this continent has been recent and thus of the fi rst generation. Beside them are various "indigenous convert" Buddhists groups, mostly founded by migrating Buddhist monastic leaders. Here the most popular sub varieties are Tibetan Buddhism and Zen Buddhism, although other types are also represented (cf. Layman, 1976;Prebish, 1979). Thus far, there has not been a really signifi cant mixing of the ethnic and indigenous varieties, at least at the level of institutions. One partial exception is the North American branch of the Japanese Nichiren group, the Soka Gakkai; but Nichiren Shoshu America, like its parent, represents a quite exclusivist version of Buddhism, and is in any case largely controlled by its ethnically Japanese members.
The fi nal example I wish to examine is that of Pentecostal Christianity. Here missionizing and global migration inform each other in complex, but particularly clear ways. Most analyses of Pentecostalism locate its origins in early 20th century America; from here, missionaries supposedly spread it to various other parts of the world (cf. Cox, 1995). And now we face a situation in which Pentecostal migrants who are the products of these missionary efforts are returning to the original "heartlands" of Pentecostalism, informing the original movement in complex ways. This vision, I suggest, is only partly correct because it misses the degree to which the origins, spread, and "return" of Pentecostalism are much more a matter of self-observation, of fi rst order observation than of some sort of essence forming, spreading, and migrating.
Three features of contemporary "global" Pentecostalism are noteworthy in this regard. First, it is entirely possible that modem "Pentecostal" forms of Christianity arose fi rst in other parts of the world, not Azusa Street in Los Angeles. As such, the Azusa Street origins are an aspect of a self-observing reconstruction of what Pentecostalism is, not a matter of absolute historical origins. Second, what Pentecostalism actually has become in different parts of the world shows such an enormous variation -in part, one suspects, because local "proto-Pentecostal" forms already existed in places as varied as India, Africa, and Latin America -that it becomes another act of self-observation to see all of this as variations of a single "global Pentecostal culture" (cf. Poewe, 1994). Third, the global migration of Pentecostals is only one aspect of a larger, more general communication among all these varied Pentecostal forms. These latter take the form of traveling missionaries, world conferences, a great deal of printed material, television contacts, correspondence, and so forth. In other words, even though the different local versions of Pentecostal Christianity are quite different from one another, they are also in far from insignifi cant contact with one another.
The three features combine to give us a fairly complex picture. We might say that Pentecostalism was indeed the product of American religious developments, but only insofar as the larger global phenomenon received its overall identity from this source, not as concerns the actual sources of Pentecostal practice around the world. Global migrants thus carry their own versions of Pentecostalism with them, transplanting them to other areas of the world as authentic differences within the felt whole of Pentecostalism as such. Accordingly, there are vast differences between say, African, Korean, Indian, Latin American, and North American Pentecostal forms. Yet, as a result of global migration and "reverse" missionizing, these different versions come to be located in areas outside their respective "heartlands", especially in North America and Europe, the destination of so much current global migration. In the Pentecostal case, the migrant's religion and the "indigenous" religion are already the same one, and therefore the relation between "ethnic" Pentecostal migrant and "indigenous" Pentecostal becomes even more susceptible of diverse possibilities. Certainly in this case, looking at the religion of global migrants strictly in terms of "accommodation" to the host society misses far too much of what is actually happening for that perspective to be adequate.
What these and other possible examples then show is that the consequences of global migration for religion is not the simple and straightforward spread of the religion of the migrants by those migrants and for those migrants. Different currents are at least potentially involved, making migrant communities and their typical religions in their new homes unpredictable but thereby also possibly fruitful areas for research into global religious trends and developments. Above all, in the context of this article, migration raises the practical question of what the various religions will become as a result of the fact that so many of them have become multi-centred, no longer nearly as easily identifi ed with a particular culture or region of the world. Obviously, this multi-centredness has in fact always already been the case with these same religions historically, notably Islam, Christianity, Judaism, Buddhism, and even Hinduism if one restricts one's view to the subcontinent. The difference of the contemporary global situation, however, is that the multiple centres are in communication with each other on a fairly consistent and constant basis: the contacts are not occasional and accidental; and this adds a further unpredictable dynamism to the process of what these religions will become. It is to the question of the construction and reconstruction of religions in the global context that I now turn.

IV
To some degree, what I have described in the previous section is not new to the contemporary era, anymore than migration, even global migration is. Before the 19th century, the histories of particularly Islam and Christianity have been affected by such migration. This is not all that surprising given that both these religions, more clearly than others, separated what we now call "ethnicity" from religious adherence. Thus, for instance, the pre-modern history of Islam in Africa may bear comparison with the situation of black Muslims in the United States today. And the entire colonizing pattern of North and South America by European Christians can yield many parallels for examining how religion plays a role in the identity-building of migrants and their hosts. That said, the differences between these situations and the contemporary one are at least as important. At the heart of the differences, as I indicated at the outset, is the context of global society with its dominant functional structures.
Two aspects of this context need to be underlined. The fi rst concerns the difference between religion and non-religion. The structural logic of a dominance of functional differentiated systems means that these will be relatively autonomous in their internal operations and also resistant to "interference" of one system in the operation of the others. Yet the systems are also mutually interdependent, which means that each assumes the others as a condition of its own autonomy; and they are each totalizing in the sense that most anything can become the subject of their processes: in principle, anything can be commodifi ed, politicized, medicalized, sacralized, and so forth. There is therefore the potential, not just for interference, but for self-conceptions of the society as a whole as primarily characterized by the norms and values of one system as opposed to others. Thus we have observations of global society as essentially capitalistic, technological, or as a network of nation-states. Religion in this context, however, suffers under certain disadvantages having to do with the diffi culty of reconstructing it as one specialized instrumentality among others. Here cannot be the place for defending this idea in detail. Suffi ce it to say that the modern/ global circumstance presents religion with the reality or at least possibility of the secularization of all the most powerful spheres of social communication. That does not mean the disappearance of religion, but certainly its potential privatization, meaning its loss of collective authority (Beyer, 1994(Beyer, , 1997cf. Chaves, 1994). This brings us to the second aspect.
As with the formation of the global political system, the emergence of a corresponding religious system has resulted in internal, segmentary differentiation. In the case of the political into territorial states; in the case of religion into religions. The modem tendency has been toward the formation of a plurality of equally legitimate states, with their corresponding nations as self-descriptions. It has also produced a tendency toward principled religious pluralism as exemplifi ed in the constitutions of most states which grant freedom of religion even in cases where such freedom is in practice denied. As with nations, we attempt to solve the problem of ordering the relations of juxtaposed religions by stating that they are all equal in principle and, moreover, equally accessible in principle. Unlike states, however, religions do not have the practical and neat territorial metaphor for deciding where one ends and the other starts. Religious pluralism is further problematic because religions generally claim some sort of absolute, non-relative, but above all holistic meaning, thereby making the question of the relation among religions more acute. Religious pluralism therefore tends to lead toward the privatization of religions because there is no religious way of establishing precise collective boundaries. Pluralism, in this case, resolves itself at the level of the individual because it cannot, by itself, do so at a collective level. The early modern and still today quite popular response to the pressure toward privatization or to the frequent diffi culty of juxtaposing religious collectivities has been to politicize religion: to bring about the identifi cation of certain states or political subunits, and their nations, with particular religions which can then be given privilege or made binding through political means. Included in this strategy are variations of the "millet" system where different religious communities are defi ned politically and, through their representative authorities, are given control over certain areas of collective authority such as personal and family law, education, health facilities, and so forth. The limits to this solution are the same as identifying certain cultural or ethnic groups with certain states: the adherents to religions do not neatly divide themselves according to politically established boundaries. Here the question of conversion and of people with "no religion" can prove particularly challenging. Religious pluralism, in fact, exists practically everywhere. And not only does the population distribution of religions ignore political boundaries; so too, in principle, do religious organizations, institutions, and authorities, these being other primary ways of attempting to draw more precise boundaries between what we treat and imagine to be separate religions. All these matters bring us directly back to the question of the effect on religions of global migration.
As I stressed in the previous section, migrants constitute an important moment in global communicative ties: they challenge other boundaries to communication, especially territorial ones, but also those based on self-observed and especially purportedly self-evident identifi cation. They raise the question of identity in a practical form. The overall social context is one that globally favours the differentiation of religion from other social forms and a corresponding pluralization of religions as particular systemic subunits of religion. Among the many consequences is an increase of both the privatization and the politicization of religion, each of which has to do above all with religious authority: what shall count as authentic in a particular religion, and what shall not? Global migration, among other forces such as the prevalence of organizations and social movements in contemporary global society, effectively encourages the development of multiple centres for addressing these issues, but without undermining but rather encouraging self-descriptive unity among these multiple centres. It is no longer at all self-evident that the traditional "heartlands" of a religion shall be the favoured locus of authority and authenticity. Or, at the very least, the "heartlands" have to establish and reproduce their authoritative status deliberately. "Diaspora" centres, in dialogue and competition, can generate their own authentic versions or variations for several reasons. In important ways, they are outside the reach of religious authorities in traditional centres, especially if they are not fi nancially dependent on these. They exist in different contexts, in particular insofar as they represent minority and even new religions in a region. That will call for greater clarity in terms of self-identifi cation. The existence of converts from outside traditional ethnic or cultural groups means that easy identifi cation of cultural identity and religious identity will be more diffi cult. Moreover, as in the case of ISKCON, but much more importantly in the case of a variety of Christian expressions ranging from liberation theological movements to Seventh Day Adventism to Pentecostalism, the "diaspora" can, through global migration, introduce itself directly into the "heartland" and thereby, if not challenge the latter's authenticity, then at least bring about reformation processes in the "heartland" versions.
The upshot of all these interwoven considerations is that we are dealing with a high degree of unpredictability when we ask about the effects on religion of global migration; and about the role of religion in the identity constructions of global migrants. Certainly, at least in part, previous patterns in which ethnically identifi ed religion becomes for a time a strategy for maintaining a sense of identity-continuity for the relatively recent migrant will continue. But beyond such temporary responses, the consequences for the formation and reformation of religions in global society does not seem to show such recognizable precedents. Overall, the relative disjunction of religion and place would seem to make a further dissociation of religion and group-cultural identity more likely, yielding perhaps a situation in which all religions become missionizing religions and in which conversions to and disaffi liations from religions become a more regular and more individualized occurrence. Yet we need to be cautious in this regard. The frequency with which we witness the politicization of religion indicates that further differentiation of religion and cultural group identity is not the only possibility; that "voluntary" religion may continue and grow along with variations of collectively enforced religion, although the two can hardly happen in the same territory. In any case, the trends that are establishing themselves are worth observing and therefore should continue to be the subject of ongoing research, if only to forestall easy but untenable conclusions about what directions religions and their adherents are taking.