Handbook of the Sociology of Religion

Edited by

Michele Dillon University of New Hampshire

CAMBRIDGE

UNIVERSITY PRESS

Nancy T. Ammerman

..

CHAPTER SIXTEEN

Religious Identities and Religious Institutions

207

For modern social theory, as well as for many ordinary people, religious identities have been a problem.' Just what does it really mean to claim a Jewish or Christian identity? To think of oneself as Presbyterian or Baptist? What do we know of that new church down the road that simply calls itself "Fellowship Church"? And do any of those things have anything to do with how we might expect someone to perform their duties as a citizen or a worker? As modern people have loosened their ties to the families and places that (perhaps) formerly enveloped them in a cocoon of faith (or at least surrounded them with a predictable round of religious activity), they can choose how and whether to be religious, including choosing how central religion will be in their lives. Religious practices and affiliations change over a complicated lifetime, and the array of religious groups in a voluntary society shifts in equally complex ways. If religious identity ever was a given, it certainly is no longer.

In his influential work on religion and personal autonomy, Philip Hammond posits that, given the mobility and complexity of the modern situation, individual religious identities are of various sorts - either ascribed (collectivity-based) or achieved (individual) and either primary (a core or "master" role) or secondary (Hammond 1988). In the premodern situation, religion was presumably collective and core," In the modern situation, taking up a collective, core religious identity is a matter of (exceptional) choice, not determinism." We neither all share one religious identity nor know quite what to make of the many identities with which we are surrounded.

While social theory has taught us that maintaining a religious identity is a problem in the "mainstream" of culture, at the margins, religious identities seem still to playa role. Indeed, much of recent research on religious identity has focused on the margins and the interstices, on the times and places where religious identities clash and/or must be remade. Lively work is now underway, for instance, on the struggle to

1 Classic theories predicting religion's demise include Marx (1878/1964) and Weber (1904-5/ 1958), with Berger (1967) providing the most elegant theoretical formulation and Lechner (1991) among the most cogent current defenders.

2 Mary Douglas (1983) debunks the notion that premodern people were thoroughly religious.

3 John Hewitt (1989) uses the example of the totally dedicated fundamentalist or orthodox person to illustrate the uncommon modern identity strategy of "exclusivity."

r

I I I

208

Nancy T. Ammerman

..

maintain or recreate immigrant religious Identities." Circumstances and demands in a new culture inevitably reshape the beliefs and practices that were taken for granted in a home country. Thrown together both with "anglo" hosts and with more proximal, yet often strange, ethnic compatriots, immigrants use religious gatherings as places to sustain old cultural ways, but also as places where new ways are hammered out (Warner and Wittner 1998). The clash of cultures is across generations, as well, as second and third generations arrive at their own relationships to ethnic and religious traditions.

Two earlier sets of immigrants now fuel another stream of writing about religious identity. Both American Catholics and American Jews have, in the last generation, passed into the mainstream of culture, have begun to experience high rates of intermarriage, and have consequently generated a good deal of identity anxiety among their leaders. Can religious institutions support distinct ways of life that are both ethnic and religious in American middle class society? Researchers have attempted to disentangle the beliefs, practices, relationships, institutions, and conscious self-identity that mayor may not be essential to perpetuating community and tradition. Whether the object of study is independent-minded post-Vatican-Il Catholics or intermarried nonreligious Jews, questions of religious identity have emerged in both practical and theoretical discussions.f

Another set of questions about religious identity is raised by seemingly incongruous religiosocial pairings (Warner 1997). Where significant collective identities stand in opposition to one another, individuals who find themselves in both warring camps at the same time must engage in active identity work. Thumma (1991) examines, for instanc~, th.e case of gays who are also evangelical. He demonstrates that special purpose orgamzatrons can engender both the rationale and the practices by which a "gay evangelical" identity can be built and sustained, but such practices take intentional work. By replicating much of evangelical culture, but within a gay environment, people create and tryout new religious solidarities.

Equally interesting has been the attempt to understand conversion. Especially at the height of sociology's attention to new religious movements, we had opportunities to see actions and affiliations transformed in ways that brought identity construction visibly to the fore (e.g., Bromley and Hammond 1987; Robbins 1988). Here were people who chose, in a thoroughly modern way, a seemingly pre-modern absorption in a religious community, trading a multilayered and complicated modern identity for one organized around a single set of core religious beliefs, practices, and assoctations."

Among the most helpful of the work on conversion that emerged from that era was Mary jo Neitz's portrayal of the process by which charismatic Catholics gained that new identity (Neitz 1987). She describes conversion as the gradual building up of a new "root reality" (Heirich 1977) at the same time that the old one is being discarded. The change is made as people engage in a kind of practical/rational process

: See, for example, Chong 1998; Kim 2000; Lawson 1999; Peria and Frehill1998; Yang 1999.

Hoge (2?00) has recently ~ade this argument. Among the key recent studies of]ewish identity are Davidman (1990), Hellman (1996), and Goldstein and Goldstein (1996). For Catholics, see Dillon (1999a) and McNamara (1992).

6 Even t~a~ construal is, of c~urse, l_llore "ideal typical" than real. Even the most tightly bounded new relIgIOUS movement still retamed complex layers of involvement and dissent and therefore complex versions of identity. See, for example, Barker (1984).

Religious Identities and Religious Institutions

209

of testing faith claims against their everyday experience to see what makes practical sense. She notes that conversion can take many forms, given that we all live with varying degrees of complexity in our worlds and begin from different degrees of religious salience. To move from a high-salience Catholic to a low-salience Catholic is a process to be explained no less than the move from a low-salience Catholic to a highsalience charismatic. And her insistence that we take practical reason into account moves us helpfully into questions of the social conditions under which religious actors, ideas, and relationships become salient within the complicated lives of modern persons.

Two things are striking to me about this literature. First, much of it proceeds with little attention to a definition or theory of identity. The assumption seems often to be that "we know it when we see it." Even careful ethnographers charting the process by which identities are under siege or being remade, write a text between the lines that asserts identity (especially an authentically religious one) to be a singular guiding "core" that shapes how others respond to us and how we guide our own behavior. We either have it or we don't. Other identities may be partial, but "real" religious ones surely must be total. The task in transitional and contradictory situations, this subtext reads, is to get the core back together again. In what follows I want to question and nuance that basic assumption.

The second thing that strikes me is that so little of our thinking about religious identity has taken the everyday world of ordinary people into account. In looking - understandably - at the places where identity work was obvious, we have perhaps avoided the basic questions about social life that ought to inform any attempt to understand the place of religion in it. How and why do people act as they do? What guides and constrains that action? Under what conditions do people orient themselves toward religious institutions and realities? By beginning with a look at recent thinking about social identity - both personal and collective - I hope to move our discussion of religious identity to include such questions.

CONSTRUCTING AND DECONSTRUCTING SOCIAL IDENTITY

Zygmunt Bauman (1996) posits that the very notion of identity is a modern preoccupation. Only when human beings begin to be dis embedded from traditional spaces and relationships, long-accepted rhythms of time and well-established activities of survival, do we begin to ask such questions as "Who am I?" and "Where do I belong?" The notion of constructing a self makes sense, he argues, only when the materials for such construction have had to be gathered from far and wide, piled up out of the deconstruction of existing social worlds. Only then do we begin to worry - either existentially or theoretically - about the coherence of our biographical narratives or the bases for our group memberships (Giddens 1991).

John Hewitt (1989), by contrast, points out that the tenuousness of personal identity is simply part of the human condition. All identities include elements of continuity (being the same person over time), integration (being a whole person, not fragments), identification (being like others), and differentiation (being unique and bounded). And every human situation, not just modern ones, places identity in jeopardy. Most basically, no situation is every fully routine; there are always surprises. Every situation gives others the opportunity to evaluate whether we are who we have been believed

210

Nancy T. Ammerman

to be, whether our actions fit the roles we have assumed. And every situation carries a tension between assuming those roles, fitting in, declaring our identification with the group, and, on the other hand, doing something that emphasizes our uniqueness, our differentiation. Whether because our actions arouse doubts in others or because we ourselves seek to declare our independence or because the situation challenges existing assumptions, human society has never allowed identity to be unproblematic. Modern society is different in the number of roles and communities available for the choosing, but not different in these basic dynamics of identification and differentiation.

More than a generation ago, Goffman (1959; 1967), Garfinkel (1967), and Berger and Luckmann (1966) began the task of theorizing how persons construct, present, and conspire to protect the fragile stability of each other's selves. Their work began to lay out th~ w~ys in which each social situation calls for the creative work of its participants, each picking up the strands of the drama as it unfolds. Players take roles that make sense to and of themselves and others (Mead 1934), aligning their actions with scripts and categories that will be recognized and can be responded to by the other players. More recently, Hall, among others, has pointed to the ways in which we identify with and "perform" the positions to which we are assigned, talking our way into ongoing stories that are always partial and incomplete (Hall 1996). The ability to align our actions with the actions of others, mutually defining and working within a recognized script, marks us as sane and competent members of our society. To break character or to challenge the basic story line of the script, these theorists taught us, is to risk insanity or to incite revolution. ~lthough scripts and characters are constantly remade by the small dramas of everyday life, those dramas are also the agents that keep existing social structures in place."

In. the generation since, the "postmodern" fragmentation of everyday life has prompted many to speculate about the increasing complexity of identity construction, emphasizing the incoherence of the scripts, rather than their solidity. Even before adding relationships built in cyberspace to the mix, many have posited a fluidity of identity that makes coherence seem obsolete." Bauman and others argue that the notion of any "core" self is impossible, that we are tourists and vagabonds, rather than pilgrims with a sense of destination (Bauman 1996). We have no core itinerary guiding our movement through the world. A tentative step in the direction of order is taken by the French theorist Michel Maffesoli, who describes our postmodern situation as a new "time of tribes" (Maffesoli 1995). He argues that "we [social scientists] have dwelled so often on the dehumanization and the disenchantment with the modern world and the solitude it induces that we are no longer capable of seeing the networks of solidarity that exist within" (p. 72). Leaving aside the traditional institutions that are presumed to hold SOCiety together and define its citizens, he turns his focus to the solidarity created in everyday gatherings. Sounding often like Durkheim (1912/1976), he looks for the affective force of sociality and custom (a "religion of humanity") that binds people together in ever-shifting gatherings. Local face-to-face groups, as seemingly anonymous as the passengers on a bus, constitute, he proposes, a "neo-tribalism characterized by

7 Their insistence on the power of the scripts is echoed in Pierre Bourdieu's notion of "habitus" a set of practical dispositions or master patterns into which we are socialized so that our actions in any situation are exactly suited to our position in that field of interaction. See Swartz (1998).

8 This is a form of community and identity that needs much more attention. See Cerulo and associates (1992) for an excellent treatment of the subject.

Religious Identities and Religious Institutions

211

fluidity, occasional gatherings and dispersal" (p. 76). Faced with the fluidity of boundaries that brings ever-changing arrays of people together, we use theatrical displays of clothing and body art to found and reconfirm communities and recognize ourselves in them.

His is an attempt to find a new way of understanding the order that still exists in the midst of the seeming chaos, a chaos that appears to leave each of us to invent a new self for each new situation and each group to an arbitrarily defined fight for recognition. While not everyone is so sure that emerging "tribes" are potentially benign, Maffesoli is not alone in pointing to fluidity of boundaries and to the strength of sociality and custom. Neither selves nor groups are utterly reconstituted with each new encounter. Some continuity clearly prevails at the same time that a complex society continually challenges that continuity.

The tension between order and chaos, between continuity and revision, is reflected in differing emphases in thinking about identity." Some focus on fluidity and agency, on the ways in which each new encounter leaves the world or the identity slightly (or radically) changed. Others, following especially in the footsteps of Bourdieu (e.g. 1987), focus on the ways in which every interaction is structured by and reinforces patterns of difference, hierarchy, and domination, especially through categories of class, race, and gender (Lamont and Fournier 1992).

But either such view of identity seems to me inadequate. I am unwilling to discard the possibility that persons seek some sense of congruence within the complexity of their lives. Nor do I believe that structured categories exist untouched by the actions and resistance of the actors who inhabit them. What seems essential is to move beyond the notion that any single category of experience - even race, class, or gender - defines identity or action. Identity is not an essential, core, category, nor is it well-conceived in binary either/or terms.'? To be feminine does not preclude being also masculine, nor does being "American" preclude being also "Irish" or "Hispanic." What we need is a way to talk about who we are and how we behave without reducing ourselves either to a single determining structural essence or to complete chaotic indeterminacy. While the realities of the late modern situation make analysis (and life itself) immensely complex, any adequate account of identity needs an account of the ongoing coherence that is constructed by human consciousness and the solidarity that is created by social gatherings, however temporary. In Giddens's words, "The reflexive project of the self ... consists in the sustaining of coherent, yet continuously revised, biographical narratives" (Giddens 1991: 5). Both the coherence and the revision are central to the process. This task is made challenging by the pluralization of our life contexts and the diversity of authorities and power present in any society, but neither the life project nor the analytical task can be set aside in the face of complexity.

IDENTITY AS A PROBLEM OF AGENCY AND STRUCTURE

At its root, differences over fluidity and constraint in the formation of identity grow out of different understandings of agency and structure. To what extent and in what ways

9 Cerulo (1997) calls these two camps the "constructionists" and the "postmodernists."

10 Minow (1997) is especially helpful in examining the political difficulties of insisting on this middle ground between essentialism and constructionism.

i' I,

212

Nancy T. Ammerman

do we understand the human person to be an agent in the creation of her or his own persona? Are groups free to define themselves, or are they defined by powerful others? The answer to those. ~uestio~s begins with the recognition that social action is guided by. pa:terned re.gulantles, social-constructed categories that organize our experience and thmkI~g. We SImply res??nd to the world in terms of what we think we already know about It. There are cognitive and psychological reasons, as much as social ones, for the fundamental way in which human thinking depends on socially constructed categories (DiMaggio 1997).

. Agency is located, then, not in freedom from patterned constraint but in our ability to m~o~e. those pa~ter~~ in .nonp~escribed ways, enabled in large measure by the very multiplicity of solidarities m WhICh we participate. Sewell (1992) locates agency in the fact that actors always occupy multiple structures and can import resources and schemas ("rules" or categories of understanding) from one to another - what he calls transposability. The rules that tell me who I am at work are not the same rules that gUi~~ my be~avior"at home or at church. Minow observes similarly that all identities a~e Intersectional, that we are always many things as once - female, white, Catholic, disabled, daughter, and the like (Minow 1997: 38ff). Indeed, part of the experience of e?ucation is :0 gain access to the schemas of cultures in distant times and places, addmg. other VOIces t~ the conversation about how life should proceed.

Emirbayer and Mische (1998) locate agency in the play of structures across time as well as across insti.tutions and space. They point to the human ability to bring past: present, and future into play at any given moment and to choose which "past" is the relevant one. They call this the "iterational element" of action. It is located in our abi~ity to categorize (if ~his is an X, then I do Y) and in our necessary formation of habits, WhICh a~e not automatic but are shaped into "settled dispositions." These theorists take very. seriously, then, the real power of existing schemas and their ability to produce ?redICtable "strategies of action" (Swidler 1986), but the equally real ability of actors to invoke those strategies in unpredictable ways.

The m.ovement across institutions and time is not, of course, done on a perfectly level playing field. Some actors have a disproportionate ability to mobilize human ~ymboli.c, and mater~al resour~es in the service of perpetuating or altering patterns of I~tera.ctlon. Sewell, like Bourdieu, points out that some actors can simply manipulate sItuatIOn: and conversations to their own symbolic and material advantage (Sewell 1992). Still, because we do not live in an enclosed world with only one pattern of ~esource allocation, no single situation is fully determined by itself. We constantly Import rules from one situation into another new or unfamiliar one. Identities, then, need to be understood as structured by existing rules and schemas constrained by eXistin~ distributi?ns .of resources and power, but also malleable in th~ everyday reality of movmg across Institutional contexts and among symbolic worlds.

Wha: each of these theorists has provided is explication for the dynamic nature of e~ch social encounter. We never arrive on the scene as a single identity, but always carry :VIth ~~ the multiple entanglements of our past and present. The very multiplicity of our Identities makes agency possible (ct. Coser 1991). Acting within and between structures across time and space, we cumulatively build up a persona and collectively shape the solidarities of which we are a part. Those personas and solidarities are themselves :hen, b~th .structures that constrain future action and sites for continuous revision and improvtsatton.

Religious Identities and Religious Institutions

213

IDENTITY AS A NARRATIVE CONSTRUCTION

What is already implied in these discussions of action and agency is the way in which "narrative" may prove a helpful metaphor for understanding the nature of identities. Studies of identity have long taken conversation and language as key sites for analysis. Indeed, the ability to use a group's language is basic to what we mean by membership and identity. To participate in the "discourse" of the group is to enter the social world that the group has constructed (Brown 1993). Our understanding of ourselves, including our incorporation of categories that keep us in dominated positions, is worked out in communication and language. As George Herbert Mead (1934) suggested, identity construction can be viewed in terms of the words we use - words that categorize, words that imply relationships (and often the unequal power inherent in them).

It is, however, critical to move past the words themselves. What narrative analysis offers us is attention to the relationships and actions that give words their meaning. If we are to understand the nature of identity in a complex world that involves multiple solidarities that both constrain and are continually reconstructed, we need a dynamic mode of analysis that moves beyond categorizing words and analyzing syntax. "(A)ll of us come to be who we are (however ephemeral, multiple, and changing) by being located or locating ourselves (usually unconsciously) in social narratives ... ," claims Margaret Somers (1994: 606). Narrative, she goes on, renders an event understandable by connecting it to a set of relationships and practices - historically and spatially, particular people doing socially patterned things.

Narrative takes an event and makes it part of a plot, that is, an action-account.

The event cannot do this for itself, but must be "emplotted" by the actors who must evaluate the various possible scenarios available to them.'! The events that become part of a narrative are selected from all that we know of the world. They are placed in a temporal order that implies causation and provides closure. And they are placed in a structure of relationships. As Ewick and Silbey (1995) point out, the process of emplotment is an inherently moral exercise, giving meaning at the same time that it creates explanation and order. This process of emplotment need rarely be conscious: internalized narratives guide most action through habit. Nor are narratives grand stories that explain the world. They need only be unspoken accounts that take an event and give it meaning by making it part of an implied episode or chapter, accounts that identify the characters in the event as part of a larger cast and that situate the event in a meaningful setting.

Among the narratives at play in identity construction are, according to Somers (1994), four types. What she calls "ontological narratives" are the socially constructed stories that are carried by the individual actor as a way of orienting and emplotting the actor's own life. This is her way of reinstating some notion of "core" or "coherence" in the face of arguments about the self as vagabond. To avoid the presumptions of immutability contained in the notion of an "ontological" self, however, I would prefer to capture this idea as "autobiographical narratives," instead. Choices about how to act

11 Emirbayer and Mische's (1998) notion of agency is very compatible with a narrative analysis.

Every action, they claim, contains, in addition to the "iterative" (past patterns), an imagined future, and an improvised present; and creative selection is involved in all three dimensions. The "imaginative element" in agency is the human ability to generate future trajectories of action (plots), to imagine what may happen as a result of my action.

214

Nancy T. Ammerman

Religious Identities and Religious Institutions

depend as much on the internal themes and plots of this autobiographical narrative as on the situation and cultural plots we imagine to be in play. The core self is constantly being negotiated in the various social contexts of a life, but it retains certain themes against which new events and episodes are weighed. Persons understand themselves as certain sorts of characters who are capable of acting in certain ways and incapable or unwilling to act in others.P An autobiographical narrative makes possible the predictability with which we respond to each other and imparts a certain trustworthiness and integrity to our action.P

It is important to note here that individual internal narratives may be at odds with the story projected to others. Persons are quite capable of acting strategically and/or without sincerity, creating a narrative more suited to what they think others will reward than to their own conscious autobiographical narrative. Likewise, those internal narratives may include characters and episodes that are never recognized by others as "real." Whether the voices heard by a schizophrenic or the visions of a mystic or the body images that tell an anorexic she is fat, autobiographical narratives may guide behavior in ways that do not include the "rational" assessment and critique of the larger community.

But much of identity is guided by those community assessments. In addition to autobiographical narratives, Somers posits the "public narratives" which are attached to groups and categories, cultures and institutions.l" Whether it is the court system or shopping malls, ethnic group or gender, these social institutions and categories provide recognized" accounts" one can give of one's behavior, accounts that identify where one belongs, what one is doing and why (Mills 1940; Scott and Lyman 1968). These are publicly constructed and shared, existing beyond the agency and consciousness of any single individual. Some have enormous strength and widespread recognition; others seem more malleable and/or more narrowly recognized. The strength of an institution can, in fact, be measured by the degree to which its narratives are available in the culture, the extent to which its stories are used to emplot actions across many settings.

Finally, Somers lists metanarratives, which are overarching cultural paradigms for how stories go - a narrative of progress or Enlightenment, for instance _ and" conceptual narratives," that is, those constructed by scientists for the sake of explanation. In making the determination about how to emplot an event, then, we evaluate possible story lines according to whether they fit with existing themes - both internal and external _ that guide those plots. That process is not utterly free, of course, and is often constrained by the power of certain actors to keep dominating stories in place.

Narrative theories posit that action proceeds, then, from the specific place and time in which it is Situated, including thereby all of the available culturally constructed stories in that place. It proceeds, as well, from the relationships embedded in the situation,

12 Teske's (1997) work on the construction of activist identities makes clear that it is possible for individuals to construct a schema to describe themselves that can then shape the action they perceive as inevitable and necessary.

13 The moral dimensions of the human construction of a self are taken up by Shotter (1984), Niebuhr (1963), and others. Much of "virtue" or "character" ethics has these issues as a central concern.

14 These public narratives reside in what Bourdieu would call "fields," the operative arena that determines which forms of cultural capital and which habitus will come into play. See Swartz (1998).

215

including the specific institutional context of rules and practices in which it is locat.ed (Lewin 1996). And it proceeds from the individual (but socially constructed) autobiographical narratives of the actors. Action takes place in a rela:ional setti~g, whi~h is composed of institutions (recognized, patterned structural relations), public narratives, and social practices, all of which are both patterned and contested - constructed and constrained.

Somers and other narrative theorists go a long way toward providing the sort of dynamic and layered mode of analysis needed in understanding identities, but at least one more layer remains. While they acknowledge the way in which narratives are situated in particular places and times, they often forget that they are also enacted by actual physical bodies in material environments. The metaphor of narrative ru~s the risk of allowing us to reduce social action to texts and words, when the habits that guide us, as well as the experiences that disrupt those habits, are often carried by affect more than thought, by deeply sensual memories and impulses as much as by plot lines. I am convinced that embodied practices are crucial. Gestures, postures, music, and movements tell the story and signal our location in it. There has been a good deal of attention to the way social situations define bodily meaning and experience (Collins 1992; Giddens 1991; Young 1989), but less attention to the physical self as agent in defining identity and membership. Here students of ritual may have something to contribute to the analysis of other forms of social interaction (Comaroff 1985; Soeffner 1997).

INGREDIENTS FOR UNDERSTANDING IDENTITY

We may understand identities as emerging, then, at the everyday intersec~ions of autobiographical and public narratives. We tell stories about ourselves (both lI.terally and through our behavior) that signal both our uniqueness and our membership, that exhibit the consistent themes that characterize us and the unfolding improvisation of the given situation. Each situation, in turn, has its own story, a public narrative shape~ by the culture and institutions of which it is a part, with powerful persons and prescribed roles establishing the plot, but surprises and dilemmas that may create gaps in t~e s~r~pt or cast doubt on the proffered identity narratives of the participants. Both the indlvidual and the collectivity are structured and remade in those everyday interactions.

We are situating the study of identity, then, in the socially structured arenas of interaction present in everyday life. IS Those everyday arenas have two key characteristi~s w_e must recognize. First, they are both structured and constructed. Our mutual storytel~mg.Is both patterned and improvised. Entrenched habits and powerful actors may mamtam existing templates for action, reinforcing the reality of social categories that define us. Nevertheless, stories and characters are constantly being revised. An adequate understanding of both personal and communal identity requires attention to the reality of

both agency and structure, both revolution and hegemony. . .

It also requires attention to the intersectionality of the situations out of WhICh Identities are constructed. Actions arise out of the multiplicity of public narratives available to modern actors. Because no situation is rigidly bounded, multiple public narratives

IS These are Marx's "social relations of production," the occasions for socially constructed actions and ideas that constitute the basis for society (Marx 1844/1964).

216

Nancy T. Ammerman

are always present, and no institutional field is defined utterly in its own terms. All situations are characterized by a fluidity of boundaries and the presence of story lines gleaned from the multiple contexts in which modern and postmodern persons live. While some visible signals, such as race, class, or gender, may act as powerful narratives across settings, in our own minds and in the actions of others toward us, no single story and no single context is an adequate account of an identity. All identities are intersectional, oriented toward the multiple stories of which they are a part.

LOCATING RELIGIOUS IDENTITIES

If we are to understand religious identities, then, we must begin by attending to episodes of social interaction (whether face-to-face or mediated) that are emplotted in a religious narrative - one in which "religious" actors, ideas, institutions, and experiences playa role in the story of who we are and who I am. An interaction takes on a religious character when it directly or indirectly invokes the co-participation of transcendence or Sacred Others, invoking a narrative in which they playa role." Action may directly reference the words, actions, or presence of a Sacred Other, but the religious narrative may also be more Implicit. Once experiences of transcendence have been institutionalized in rituals, stories, moral prescriptions, and traditions, those practices are then recognized as religious, whether or not the participants experience them as direct encounters with the Sacred (or even believe Sacred Others to exist). Participating in practices that have been handed down through a religious tradition (lighting Sabbath candles, for instance) invokes thereby religious narratives, whether or not the participants understand their action to directly involve a Sacred Other. When I say I am a Baptist, you recognize that as a religious identity (with more or less accurate expectations about how Baptists behave) simply because of the implied connection to religious institutions and traditions I am invoking. Here the distilled and institutionalized symbols of religious experience evoke religious narratives, whether or not particular individuals believe in or experience them. Likewise, within institutionalized religious contexts, given episodes of social interaction will be governed by accepted strategies of action that mayor may not directly involve transcendent ideas or experiences, mayor may not invite direct participation by Sacred Actors. Religious narratives - the building blocks of individual and collective religious identities - are activated, then, by settings in which they are implied and by actions into which they have been distilled, as well as by overt experiences and direct references.

In modern, functionally differentiated societies, religious experiences of any sort have been assumed to be confined either to a recognized religious institution or to the privacy of one's own ecstasy. Religious institutions have become the sole social repository of mystery, according to this view, keeping it safely domesticated and out of public view. I would argue, however, that this is a very incomplete inventory of the presence of religion in society.'? If we take structured-yet-improvised episodes of social

16 Berger (1974) argues for a substantive definition of religion that depends on the presence of a socially recognized Sacred Other. This is basic to his disagreement with Luckmann, who uses a functional definition. However, Luckmann (1991) also recognizes the role of "great transcendences," the sorts of extra-empirical actors referenced here.

17 In what follows I am seeking to expand the modern social territory seen as potentially religious.

Berger (1992) makes a similar move in expanding the modern cognitive territory for religion.

Religious Identities and Religious Institutions

217

interaction as our basis and recognize the necessary iniersectionality of all such episodes, there is no a priori reason to assume that religious episodes will only happen in religious institutions or in private seclusion. If it is true that all social contexts contain multiple narratives, that schemas from one social arena can be transposed onto another, then it must be true that under certain conditions religious narratives may appear in settings outside officially religious bounds. No matter what the presumed functional arena, narratives of transcendence might intervene.

Rather than making assumptions of religious absence based on the meta-narrative of secularization, or assuming that religious narratives can only be plausible if they have no competition, our task as social scientists ought to be the examination of ordinary episodes of social interaction to determine the presence or absence of religious narratives and practices (Ammerman 1994). If we do not begin with a conceptual narrative that assumes a radical functional differentiation between religious and nonreligious (or between "public" and "private"), we may be able to ask important questions, then, about the circumstances under which religious narratives of identity come into play. Once having removed our conceptual blinders we can begin to ask more basic questions about the social organization of religious identities, analyzing them as potentially part and parcel of the multiple narratives that shape all of social life. Situations where religious identities seem to clash with other identities (e.g., gay evangelicals) or where identities are being remade in new contexts (e.g., immigrants) remain theoretically interesting, then, not because they are anomalies, but because they are exemplars. They provide models that can inform the study of religious identities of a more common sort.

RELIGIOUS ORGANIZATIONS AND NARRATIVES OF IDENTITY

That conceptual turn should not, however, lead us to neglect explicitly religious organizations, places where the society has indeed institutionalized an expectation that religious interaction will take place. Religious organizations are important sites for. religious experience and for the constructing of religious identities. They are suppliers of "public narratives," accounts that express the history and purposes of a cultural or institutional entity (Somers 1994: 619). These organizations create widespread social arenas in which religious action can occur, and they supply structured religious biographical narratives - the saved sinner, the pilgrim - within which the actor's own autobiographical narrative can be experienced.

Religious organizations establish such narratives through elaborate sets of roles, myths, rituals, and behavioral prescriptions that encourage participants to perceive Sacred Others as their coparticipants in life. They establish a "grammar" for the stories people tell about the world (Lindbeck 1984), a grammar that extends to the body, as well as to language (Hervieu-Leger 1993). As Warner points out, music, posturing, rhythmic movement, and eating are human experiences that create community, defi~e boundaries and identities, but also sometimes allow the bridging of those boundaries (Warner 1997).18 Simple melodies and the deep resonance of sound, he argues, create an

18 Although Bartkowski (2000) focuses primarily on discourse, he also has paid attention to the use of space, physical contact, and gesture, and other ways in which Promise Keepers have remade male identities.

218

Nancy T. Ammerman

"

experience beyond words and ideas that is inherently communal and identity defining. Similarly, rhythmic common movement is a powerful bonding force that creates community and establishes practices that become part of a member's repertoire of action (see Bellah, Chapter 3, this volume). By supplying and reinforcing habitual gestures and actions, religious organizations orient their participants toward the sacred dimensions of experience.

While religious organizations generate and sustain powerful narratives, the intersectionality of identities and the permeability of modern institutional boundaries guarantee that these narratives will not remain singular or untouched. Even institutional religious participation is not always limited to a single organization or tradition. Nancy Eiesland describes one such multiple-religious family, residents of an Atlanta exurb (Eiesland 2000). While they are members of the local United Methodist Church, the wife attends meetings of a "Grief Relief" support group at the nearby Baptist megachurch. She has siblings who are Presbyterian and Catholic, respectively. Her husband grew up with little attachment to any faith, and neither of them had been part of a Methodist church before joining this one. The religious narratives in which they participate include elements from all these ties at once. It would be a mistake to say that they "are" Methodist. They are constructing religious identities that weave together stories from all these experiences of religious community and faith.

Given that members participate in multiple public narratives, from both religious and secular institutional sources, we can ask which religious institutions supply the most robust and portable plot lines. The narratives supplied by religious organizations may be more or less richly nuanced, allowing them to address wider or narrower ranges of human existence. They may also be more or less able to incorporate counter-narratives, making sense of the very events that would seem to challenge their plausibility. 19 Part of the analyst's job is to assess the degree to which any given religious organization is generating, nurturing, and extending the language, grammar, gestures, and stories that are capable of surviving in the everyday practical competition among modern identity narratives.

Over the last forty years, for instance, liberal Protestant traditions have notoriously neglected their unique narratives, creating a time of "vanishing boundaries" (Hoge, Johnson, and Luidens 1994). Higher education has led to increasing knowledge about multiple religious traditions and to increasing contact (including intermarriage) with persons from those traditions (Wuthnow 1988). The typical period of youthful exploration has extended well into adulthood, and increasing numbers of liberal Protestant youth have simply never returned. Whatever religious accounts they may have learned as children are now buried beneath layers of new experience that mayor may not extend those childhood stories. Even their parents are hard-pressed to give an account of their religious identity that extends beyond an attempt to "do unto others as you would have them do unto you" (Ammerman 1997b).

Our recent research found, for instance, that barely one-third of the members of the Episcopal and United Church of Christ congregations we surveyed had grown up as Episcopalians or Congregationalists (or in the other denominations out of which the merged UCC was formed), respectively. Not surprisingly, persons who are

19 Christian Smith (1998) argues that it is precisely this ability to explain its enemies that has rendered American evangelicalism so robust.

Religious Identities and Religious Institutions

not maintaining a lifelong religious tradition are less likely to describe their current denominational identification as important to how they think about themselves. All the church attenders we surveyed - from the Church of God members to the Presbyterians and Lutherans - chose, on average, "spiritual person" and "devout Christian" as more important to them than their particular denominational identity. But for noncradle members the margin was much wider than for cradle members, and "spiritual person" was a more popular self-designation than "devout Christian." Having been exposed to numerous religious narratives, they have developed a less particular way to describe themselves. While "religious seeker" is not the term they most often chose, their journey has nevertheless been incorporated into an autobiographical narrative more "spiritual" than "religious" (Roof 1999ai Wuthnow 1998). In turn, congregations in which "switchers" dominate are less likely to describe themselves as strongly attached to their denomination's traditions. Congregations full of "switchers" often report that they have given up on maintaining the narratives of the denominational tradition, emphasizing a more generic Christian story (Sikkink 1999).

Some switcher congregations, however, have adopted a different narrative strategy.

They emphasize practices intended to introduce new adherents to the stories and traditions of the denomination. They teach newcomers their distinctive modes of worship, introduce children and adults to denominational ideas and stories through Christian education programs, and tell tales of the great deeds done through the cooperative efforts of the churches that share their denominational identity. As a result, in these churches the tie between the congregation's identity and that of the denomination remains strong in spite of the mixture of religious stories represented by those in the pews (Ammerman 2000). Theirs is an active process of narrative construction, of bringing individual stories into a new communal context at the same time that a tradition is being passed on and thus modified (Bass 1994). Within some religious organizational contexts, then, religious identities are being constructed in rather intentional ways out of longstanding narratives. Tradition becomes more a verb than a noun (Calhoun 1991), supplying and introducing accounts and characters to new cohorts of religious actors. By telling the stories, practicing the rituals, and celebrating the heroes, these congregations consciously keep a genre of denominational public narratives alive.'?

It is important to note that the narratives derived from religious tradition are not static. Sacred stories, no less than any others, are both structured and improvised, determined by tradition and created out of human appropriation of that tradition. Indeed, primal religious narratives that involve episodes of transcendence are inherently unstable, disrupting existing scnpts." "Sacred Others" are notoriously unpredictable. If we recognize religious identities as both structured and emergent, then one of the most interesting questions we may ask is about the conditions under which religious episodes emerge in surprising ways, redefining the expectations of the actors in them. To use

20 Hervieu-Leger (2000) argues that posttraditional religious institutions must mobilize a combination of emotional belonging and rational appeals to an "ethicocultural heritage." For example, pilgrimages involve the experience of a long journey, the exhilaration of being part of a large throng, recognition by international media, rituals in which potent symbols (like the Pope) are mobilized, exposure to sites in which traditional stories are embedded, and participation in didactic efforts to pass on those stories.

21 Berger's (1967) discussion of "exstasls" and "dealienation" is a particularly provocative suggestion of the way in which religious experience can threaten established orders.

219

220

Nancy T. Ammerman

Weber's (1925/1978) terms, when does "charismatic" authority trump "rational-legal" or "traditional" rules? A variety of students of religious ritual have attempted to assess the ability of ecstatic experiences to alter the narratives participants take with them into the more mundane world." Others have noted that religious experience has its own ordered "flow" (Neitz and Spickard 1990). A deeper understanding of religious identities would surely take up the question of these tensions between everyday order and transcendent chaos. How is that everyday order maintained, and when are glimpses of transcendence allowed to intruder-" While religious organizations are primary sites for locating religious narratives, they are by no means passive repositories.

RELIGIOUS NARRATIVES BEYOND RELIGIOUS BOUNDARIES

A given autobiographical narrative may contain plot lines derived from numerous religious organizational contexts and from both structured traditions and emergent experience. But it is important to look for religiously oriented narratives in other social contexts, as well. There are enormous numbers of opportunities for encounters with transcendence and equally pervasive religious plot lines available in contexts as varied as mass media, small study groups, voluntary social service activity, even corporate retreats.P' Popular music, television programs, and movies often use religious images and stories, both borrowing from existing traditions and inventing new ones. Incorporated into the telling of stories about love and life, writers and artists invoke sacred actors and images.

In addition, myriad religious sources beyond official institutions supply us with signals by which we can recognize religious coparticipants. So-called New Age practices make their way through a loose network of bookstores and conventions, movies and Internet sites. But New Age is only one small stream within the eclectic flow of religious products and experiences present in every corner of late modern culture. Far more pervasive - but also largely outside the bounds of traditional congregations and denominations - are the narratives supplied by conservative Christian preachers, family advisors, clothing manufacturers, event producers, broadcasters, politicians, and missionaries. But, within every religious tradition, entrepreneurs in the cultural marketplace offer prescriptions and exhortation on how to live out a properly religious life.

These extrainstitutional religious producers are often just that - producers of goods and services that create a material world that supports and expresses the narratives of those who inhabit it. Whether it is a New Age t-shirt or a Conservative Christian coffee mug, clothing and props are used to signal religious identities to whatever community or potential community may observe them. In mass culture, jewelry and bumper stickers can tell a story that signals the membership of some and the exclusion of others.v

22 See, for example, Alexander (1991), Neitz (2000), McRoberts (Chapter 28, this volume), and Nelson (1997) for recent analyses of the way religious experience constructs reality.

23 Berger's more recent musings on these subjects can be found in A Far Glory: The Quest for Faith in an Age of Credulity (1992).

24 On mass media, see Hoover (1997); on small groups, see Wuthnow (1994); on volunteering, Wuthnow (1991); and on religion in business, Nash (1994).

2S Maffesoli (1995), Soeffner (1997), and others have paid attention to "punk" bodily displays, but few have noted the way Christian clothing and jewelry functions analogously to create an implied community of evangelicals within public spaces. An exception is McDannel! (1995). Read and Bartkowski (2000) pay attention to the role of clothing for Muslim women.

Religious Identities and Religious Institutions

221

The interactions of those who thereby recognize each other as coparticipants in a story extends and elaborates that same story.

Religious clothing is one example of the ways in which religious narratives and practices cross institutional lines. Privatized religious identities may, of course, be at work in any setting. Individuals for whom religious narratives playa central role may weave religious accounts together with the experiences of everyday life. Recall Neitz's study of converts to charismatic Catholicism (Neitz 1987). As they experience the stresses and strains of everyday work and family life, they "tryon" the accounts provided by the charismatic community. Those who finally identify with the prayer group are those for whom everyday autobiographical narratives and public religious narratives begin to be consonant. It is not just that they have learned to experience God's presence in weekly prayer meetings, but that they have learned to see God's hand at work in the most mundane of everyday events, whether or not other participants in those events see the story in a religious light. While their conversion is obviously encouraged and shaped by a religious organization, the stories it engenders cross institutional boundaries - at least by way of the private experiences of participants.

But sometimes religious narratives and practices cross institutional boundaries in much more publicly accessible ways. Both Mary Pattillo-McCoy (1998) and Richard Wood (1999) have offered persuasive accounts of the ways in which religious idioms can enable social movement activity. Prayer, hymn singing, and biblical storytelling can exist alongside economic and political rhetoric in attempts to mobilize citizens for action. In so doing, the activist identity that is constructed is infused with religious meaning. The symbols and rituals of "civil religion" are less oriented toward change, but they, too, offer a transcendent account of collective identity (Bellah 1967). Similarly, businesses of all sorts may tell religious stories about their founding and purpose, encouraging religious identification among their workers and customers (Bromley 1998b).

Even when the organization itself does not claim any sort of religious narrative, units within it may be dominated by coreligionists who establish an environment in which they carryon a religious narrative about who they are and what they are doing. At the church I call Southside Gospel Church, several members recounted their successful efforts to get church friends hired at their workplaces (and/or to convert coworkers), resulting in a "Christian" workplace in spite of the secular structures in which it was lodged (Ammerman 1987). Woven throughout the activity of producing and selling commercial products was a narrative of God's activity in their lives, guiding and reflecting on those transactions, sometimes breaking into their conversations with outsiders, as well. A similar pattern is emerging in our recent research with social service providers. While some aspects of their organizations and interactions are defined by structures of governmental or economic necessity, other signals emerge, as well. Their stories of individual "vocation" and organizational "mission" are full of religious symbols, and their communities of solidarity and support are populated by religious actors."

It is not, however, always possible to bring religious narratives into play. In many settings, official or unofficial rules prohibit any but the most privatized engagement with religious experiences or ideas. Individuals may bring their faith to work, for instance,

26 Ongoing analysis from the "Organizing Religious Work" project, Hartford Institute for Religion Research, Nancy Ammerman, principal investigator.

222

Nancy T. Ammerman

but it is often prohibited from escaping their own private musings. As with any ather identity, we cannot understand the nature of religious identities without asking questions of institutional power and hegemony. We need to know what the existing rules are and what resources various actors bring to the task of identity construction and maintenance.

But religious narratives are also often excluded because they violate the metanarrative of rationality. Where social institutions depend for their legitimacy on a myth of reason, events and interaction defined as religious are unlikely and unwelcome. Under that meta-narrative of modern progress and Enlightenment, individuals and institutions have learned to separate episodes and chapters in their lives into separate narratives, submerging experiences that seemed to violate the larger narrative's prescriptions. When relationships with a Sacred Other threatened to intrude in contexts not deemed appropriate, those relationships were stuffed back into the closet. Indeed, as this metaphor suggests, the analysis of religious identities could learn a good deal from analysis of the ways in which gay identities have been suppressed (Butler 1990; Rahman 2000). Whether the mechanisms are psychological denial or subcultural seclusion, dominant cultures can suppress identity narratives that violate the basic rules by which power is distributed or orderly meaning maintained. Attention to all the ways in which cultural elites shape the available narratives is a critical project for those who wish to understand the formation of religious identities.

One of those elite sectors, of course, is located in the modern nation-state. Here we find that religious identities have been excluded (except as expressions of individual preference) because bitter experience has taught us the dangers of linking God to temporal powers that tax and kill (Casanova 1994). The particular history of negotiation between "church" and "state" in the Western world has framed a story that casts religion as a dangerous character to be avoided at all cost. Throughout the middle of the twentieth century, courts in the United States struggled with the ways in which religious identities could and could not be recognized in various public settings, ranging from schools and hospitals to zoning decisions and presidential politics. In the midst of the arguments, many in U.S. SOCiety came to perceive that all public shared spaces must be kept free of religious events, actors, ideas, and symbols. More recent arguments have begun to question and criticize those assumptions (Carter 1993). It is simply not clear when the power of the state can and should be brought to bear on the ability of persons and organizations to invoke religious narratives and rationales for their public behavior. Nor is it clear when or if public religious behavior violates necessary norms of civility. The meta-narratives of modern civility are being challenged and remade, and these meta-narratives playa powerful role in the ability to bring religious narratives to bear outside religious institutions.

CONSTRUCTING RELIGIOUS IDENTITIES

Every social interaction, then, provides an opportunity for the expression and elaboration of narratives that come from the variety of settings and memberships represented by the participants. The construction of religious identities is a multilayered exercise that takes place in specialized religious settings, but also in every other institutional context. Autobiographical narratives are constructed in a world where episodes of transcendence can occur anywhere; no interaction is utterly secular or utterly sacred. The

Religious Identities and Religious Institutions

223

permeability of boundaries and the intersectionality of identity require more subtle tools of analysis than the categorical checklists of old. It requires tools that will let us move beyond either/or assumptions about religious identity.

We might begin with a not-so-simple catalogue of religious narratives, looking for the chapters and themes that are most common in different social locations. To what extent does a person use various religious stories as organizing frames for the episodes of a life? Do those stories come from and resonate with specific religious traditions? What narratives occur most commonly as markers of membership in various religious collectivities? And how are religious narratives and social action implicated in each other across institutional boundaries? Both the cataloguing and the organizing are basic tasks mandated by the multiple arenas and permeable boundaries of the late modern world.

As with any other identity, however, we cannot understand the nature of religious identities without also asking questions of institutional power and hegemony. We need to know what the existing plot rules are and what resources various actors bring to the scene. Under what conditions, for instance, are glimpses of transcendence allowed to intrude on everyday, ordered, reality? How and where does the meta-narrative of rationality, progress, and Enlightenment, exclude accounts that reference sacred actors and experiences? How is the idea of a secular state being renegotiated to include (perhaps) new public arenas in which religious narratives can be voiced (Casanova 1994; Carter 1993)? Attention to all the ways in which cultural (and religious) elites shape the available narratives is a critical project for those who wish to understand the formation of religious identities. We need attention to the various ways in which mechanisms of culture and state make some narratives more available and permissible than others. Questions of power and domination are central to the construction of religious identities no less than to any other sort.

It is important to note that the structures that shape religious identity formation are not only those imposed by powerful secular authorities. They are also the very religious institutions that claim legitimate authority to determine who may give voice to their narratives. By the stories they tell and the people they valorize, religious institutions highlight some life plans and ignore or denigrate others (Nason-Clark 1997). Mostly these messages are carried by the routine activities and habits of the participants, but overt sacred authorities can step in, as well. Whether silencing a Southern Baptist woman who entertains the possibility of a clergy identity or excluding a Methodist man who constructs a story in which he and a partner live in a religiously blessed union, religious institutions intervene to control the stock of identity narratives available to their participants.

But even religious authority is not unchangeable. All narratives of identity - both individual and collective - are both constructed and constrained. We listen for the public narratives we recognize and tell the personal stories that have shaped us: And in the midst of those intersecting narratives, we continually recreate an autobiography that is "coherent, but constantly revised" (to return to Giddens's [1991] words). While powerful authorities keep existing stories in place, new narratives are constantly emerging. Ongoing stories are disrupted by unexpected events and deliberate innovation. Accounts from one arena are imported into another, as new participants carry plots from place to place. The study of religious identity is not the study of external assaults on an unchanging religious core. Rather, it is the study of religious narratives

224

Nancy T. Ammerman

Ii

that are themselves the product of ongoing interaction, both among the diverse human participants in the drama and between them and whatever unpredictable sacred experience they recognize in their midst.

If we posit that at least some individuals and some social settings can and do generate experiences of transcendence, then the study of religious identities should take place at that intersection where individual and social meet the sacred. Given the human propensity for ordering our world, we may expect such intersections to occur in patterned and institutionalized ways. But given the equal human propensity for imagination, invention, and disruption, we can also expect both internalized and externally structured religious narrative patterns to shift over time. The transcendent referent that makes an identity narrative a religious one is neither a fixed set of institutional symbols nor an utterly chaotic experience in which selves and situations are redefined by divine fiat. It is at once both structured and emergent.

Individuals improvise religious narratives out of past experience and interaction, the other times and places in which sacred actors and institutions have had a role. Their culture and its institutions create situations that are more or less open to religious action. From both the existing themes of an individual autobiography and the available themes in the situation, episodes emerge and are "emplotted." Describing religious identities is not a matter of asking a checklist of categorical questions, but a matter of analyzing a dynamic process, the boundaries of which cannot be assumed to fall neatly within private or personal domains. Intersectionality means that no situation or identity is ever utterly devoid of multiple narratives, both public and private, sacred and secular. People can signal the presence of religious ideas, symbols, story lines, and sacred coparticipants within a wide range of social contexts, both to themselves and to others, invoking religious narratives of widely varying scope and robustness. Wherever those religious signals are being generated and received, new narratives are being created and old ones retold. Understanding religious identities will require that we listen for stories in all their dynamic complexity, situating them in the multiple relational and institutional contexts in which contemporary people live their lives.

CHAPTER SEVENTEEN

Religion and the New Immigrants

Helen Rose Ebaugh

Changes in U.S. immigration laws in the past four decades have had far-reaching consequences for American religion. Even though the majority of the new immigrants are Christian (Warner and Wittner 1998; Ebaugh and Chafetz 2000b), the practices, symbols, languages, sounds, and smells that accompany the ethnically and racially diverse forms of practicing Christianity, brought by immigrants from Latin America, the Caribbean, the Philippines; China, Vietnam, India, Africa, and elsewhere challenge the various European practices of Christianity that have predominated in the United States since its founding. As Maffy-Kipp (1997) argues, rather than immigrants "de-Christianizing" religion in America, they have, in fact, "de-Europeanized" American Christianity. In addition, the new immigrants have brought religious traditions, such as Buddhism, Hinduism, Islam, Zoroastrianism, Sikhism, Vodou, and Rastafarianism, that were unfamiliar to Americans prior to the mid-1960s. Today many American neighborhoods are dotted with temples, mosques, shrines, storefront churches, Christian churches with foreign names, guadwaras, and botannicas.

THE HISTORICAL CONTEXT

The "new immigrants" refer to those who entered the United States after the passage of the Hart-Cellar Immigration Act of 1965. The abolition of the country-of-origin quotas established in 1924, and the dramatic increase in immigration visas provided to people from Asia and Latin America, in particular, significantly altered the racial and ethnic backgrounds of immigrants. For example, the number of Asian immigrants living in the United States rose from about 150,000 in the 1950s to more than 2.7 million in the 1980s, while the number of European immigrants fell by more than one-third. Likewise, during the 1950s, the six hundred thousand immigrants who came from Latin America and the Caribbean accounted for one in four immigrants, while three decades later, the 3.5 million immigrants who arrived from these areas accounted for 47 percent of all admissions (Miller and Miller 1996). Of the five million immigrants who arrived between 1985 and 1990, only 13 percent were born in Europe, Canada, Australia, or New Zealand, while 26 percent came from Mexico, 31 percent from Asia, and 22 percent from other parts of the Americas (Chiswick and Sullivan 1995: 216-17). In addition, per country limitations on legal flows have increased the national diversity

225