Ethnicity is a salient feature of numerous societies throughout the world. Few societies are ethnically homogeneous, even when they proclaim themselves to be. Consequently, ethnicity has been a preoccupation of sociologists since the early days of the discipline (although more so in the United States than elsewhere).
Yet there is not complete agreement on how the subject should be defined. In the past, it was common to highlight cultural difference as an essential feature of ethnic distinctiveness (see, e.g., van den Berghe 1967). Recently, this has lost favor on the grounds that cultural differences may vary from one setting to another and from one historical period to another. Following an approach attributed to Frederik Barth (1969), recent definitions have therefore focused on the existence of a recognized social boundary. But still among the most useful definitions is the classic one of Max Weber ([1922] 1968): An ethnic group is one whose members ”entertain a subjective belief in their common descent because of similarities of physical type or of customs or both, or because of memories of colonization and migration” Weber adds insightfully, ”It does not matter whether or not an objective blood relationship exists” (p. 389).
Despite definitional disagreements, there is general recognition that a number of characteristics appear as hallmarks of ethnicity; not all of them will be present in every case, but many will be. They include features shared by group members, such as the same or similar geographic origin, language, religion, foods, traditions, folklore, music, and residential patterns. Also typical are special political concerns, particularly with regard to a homeland; institutions (e.g., social clubs) to serve the group; and a consciousness of kind, or sense of distinctiveness from others (for the full listing, see Thernstrom et al. 1980, p. vi).
There is controversy over whether race should be viewed as a form of ethnicity. In this context, ”race” should not be understood as a bundle of genetically determined traits that of themselves generate social differences—a view that has been repudiated by the vast majority of social scientists—but as a kind of social classification used by members of a society. Many scholars distinguish between ethnicity and race. For example, Pierre van den Berghe (1967) defines race as a social classification based on putative physical traits and ethnicity as a classification based on cultural ones (see also Omi and Winant 1994). The contrast between the two can also be formulated in terms of volition vs. external constraint, with racial categories seen as more imposed by outsiders and ethnic ones as more claimed by the group members themselves (see, e.g., Waters 1990). But equally commonly, race is seen as a variant of ethnicity: A racial group is, then, an ethnic group whose members are believed, by others if not also by themselves, to be physiologically distinctive. This is the approach adopted in this article for several reasons. Not only do racial groups typically have the characteristics of ethnic groups (e.g., cultural distinctiveness), but many seemingly nonracial ethnic groups may also be believed to possess some distinctive physical features (e.g., the olive skin tone of Italians). The distinction between the two types of groups is therefore not hard and fast, a conclusion that is underscored by the historical transmutation of some racial groups into nonra-cial ones, for example, the Irish (Ignatiev 1995; Roediger 1991).
Sociologists recognize that the imprint of history on the contemporary ethnic relations of any society is deep, and this gives rise to another distinction that is potentially central to any discussion of ethnicity. It pertains to the mode of entry of a group into a society, and it has been formulated by Stanley Lieberson (1961; see also Blauner 1972) in terms of the situation that obtains just after contact between an indigenous group and one migrating into an area. One possibility is that the migrant group dominates, typically through conquest (often aided by the introduction of new diseases)—this is exemplified in the contacts between indigenes and European settlers in Australia and the United States. The other is that the indigenous group dominates, as occurred during the century of mass immigration (1830-1930) to the United States. The crux of the matter here is whether a group is incorporated into a society through force or through more or less voluntary migration. Lieberson argues that a group’s mode of entry is fateful for its trajectory of development in a society, and this is amply borne out in the literature on ethnicity.
Stated in very broad terms, three approaches have dominated the sociological study of ethnicity; a fourth and much newer one appears to hold the promise of eventual parity with them. Of the older approaches, one, the assimilation perspective, focuses on social processes and outcomes that tend to dissolve ethnic distinctions, leading to the assimilation of one ethnic group by another or by the larger society. The second approach could be labeled as stratification. As the name implies, it addresses the origins and consequences of inequalities of various kinds among ethnic groups. The third approach focuses on ethnic-group resources, and encompasses processes, such as mobilization and solidarity, by which the members of ethnic groups attempt to use their ethnicity to compete successfully with others. The newest approach is a social constructionist one. It stems from the recognition that ethnic boundaries are malleable and is concerned with the ways by which such boundaries are created, maintained, and transformed.
No one of these approaches could be described as preeminent; each is a major presence in contemporary research on ethnicity and has shown that it has something to contribute. Other approaches are possible but are not as theoretically and empirically developed as these four. One other approach seeks a basis for ethnicity in sociobiology, viewing it as a form of genetic nepotism, a generalization of the universal tendency among animals to favor kin. van den Berghe (1981) has been an exponent of such an approach, but as yet no body of evidence has been developed to distinguish it from more sociological approaches; other sociologists have not followed his lead. Ethnicity has also been viewed as ”primordial,” deriving from deeply seated human impulses and needs that are not eradicated by modernization (Isaacs 1975). But this viewpoint has not led to sociologically interesting research, and it has lacked exponents in recent decades. Another attempt, stemming from ”rational choice theory” (Banton 1983; Hechter 1987), seeks to explain ethnic phenomena in terms of the efforts of individuals to maximize their advantages (or, in technical language, utilities). Research using rational choice theory is still too immature to draw up a meaningful balance sheet on it.
The assimilation approach has deep roots in classical social theory as well as in American sociology, where it is often traced to Robert E. Park’s 1926 formulation of a race relations cycle of ”contacts, competition, accommodation, and eventual assimilation” (quoted from Park 1950, p. 150; cf. Lal 1990). The canonical statement of the assimilation approach is by Milton Gordon (1964; for an updated revision, see Alba and Nee 1997). Although Gordon was addressing the role of ethnicity in the United States, his formulation is so general that it can be readily applied to other societies. At the heart of his contribution is the recognition that assimilation is a multidimensional concept. He distinguished, in fact, among seven types of assimilation, but the critical distinction lies between two of them: acculturation and structural (or social) assimilation. Acculturation means the adaptation by an ethnic group of the cultural patterns of the dominant, or majority, group. Such acculturation encompasses not only external cultural traits, such as dress and language, but also internal ones, such as beliefs and values. Gordon (1964) theorized that acculturation is typically the first of the types of assimilation to occur and that the stage of ”’acculturation only’ may continue indefinitely” (p. 77)—hence the importance of the second assimilation type, structural assimilation. Structural assimilation is defined by Gordon to mean the entry of an ethnic group’s members into close, or primary, relationships with members of the dominant group (or, at least, with ethnic outsiders). The cardinal hypothesis in Gordon’s (1964) scheme is that structural assimilation is the key that unlocks all other types: ”Once structural assimilation has occurred … all of the other types of assimilation will naturally follow” (Gordon 1964, p. 81). Once structural assimilation occurs, the way is open to widespread intermarriage, an abating of prejudice and discrimination, and the full participation of ethnic-group members in the life of a society.
Gordon discussed certain models, or theories, of the assimilation process (they might also be described as ideologies because of their value-laden character). Although these were, again, developed for the U.S. context, Gordon’s discussion is so lucid that the models have passed into more general application. One is labeled as ”Anglo-conformity” by Gordon, and it describes an assimilation that is limited to acculturation to the behavior and values of the core ethnic group, taken to be Protestants with ancestry from the British Isles in the American context. A second model is that of the ”melting pot.” It envisions an assimilation process that operates on cultural and structural planes. One outcome is a culture that contains contributions from numerous ethnic groups and is adopted by their members. A parallel outcome on a structural plane is a pattern of widespread marriage across ethnic lines, in which the members of all ethnic groups participate and which leads ultimately to population made up of individuals of quite intermixed ancestry. The melting-pot idea corresponds with some popular notions about U.S. society, but so does the last model explicated by Gordon—namely, ”cultural pluralism.” Cultural pluralism corresponds with a situation in which ethnic groups remain socially differentiated, often with their own institutions and high rates of in-group marriage, and retain some culturally distinctive features. It is, in fact, an apt description of many societies throughout the world.
Urban ecology, dating back to the origins of the Chicago School of American sociology, is quite compatible with the assimilation approach. A core tenet of this tradition is that the spatial separation of one group from another mirrors the social distance between them and changes as this does. This ultimately implies a model of spatial assimilation (Massey 1985), according to which residential mobility follows from the acculturation and socioeconomic mobility of ethnic-group members and is often an intermediate step on the way to more complete—that is, structural—assimilation. The model envisions an early stage of residential segregation, as the members of ethnic groups— typically, immigrants and their children—are concentrated in urban enclaves, which frequently results in the displacement of other groups. But as the members of an ethnic group acculturate and establish themselves in the labor markets of the host society, they attempt to leave behind less successful co-ethnics and to convert socioeconomic and assimilation progress into residential gain by ”purchasing” residence in places with greater advantages and amenities. This process implies, on the one hand, a tendency toward dispersion of an ethnic group, opening the way for increased contact with members of the ethnic majority, and, on the other hand, greater resemblance in terms of residential characteristics between successful ethnic-group members and their peers from the majority group.
The assimilation perspective has been successfully applied to American ethnic groups derived from European immigration. In a review of the evidence, Charles Hirschman (1983) documents the abating of ethnic differences in the white population in terms of socioeconomic achievement, residential location, and intermarriage. To cite some representative research findings, Stanley Lieberson and Mary Waters (1988), comparing the occupations of men of European ancestry in 1900 and 1980, find a marked decline in occupational concentrations, although these still show traces of the patterns of the past. These authors and Richard Alba (1990) also demonstrate the great extent to which interethnic marriage now takes place within the white population: Three of every four marriages in this group involve some degree of ethnic boundary crossing. The assimilation perspective as applied to European Americans has not been without its critics; Andrew Greeley (1971) has done the most to assemble evidence of persisting ethnic differences.
There can be no doubt that African Americans do not exemplify the patterns expected under the assimilation perspective (e.g., Massey and Denton 1993). A question that now motivates much debate and research is the degree to which the assimilation patterns will be found among contemporary immigrants and their descendants (see Alba and Nee 1997; Portes and Rumbaut 1996). It is too early to answer the question, but reflecting on the potential differences between past and contemporary immigrations, Portes and Zhou (1993) have added a new concept to the ethnicity arsenal—namely, segmented assimilation, which acknowledges that the strata of American society into which individuals assimilate may vary considerably. In their view, the children of dark-skinned, working-class immigrants who grow up in the inner city are at great risk of assimilating into the indigenous lower class and thus experiencing assimilation with little or no upward mobility.
Much of the evidence on assimilation and ethnic change is derived from cross-sectional studies rather than those done over time; the latter are difficult to conduct because of the limited availability of comparable data from different time points. Cross-sectional analyses involve some dissection of an ethnic group into parts expected to display a trajectory of change. One basis for such a dissection is generational groups. Generation here refers to distance in descent from the point of entry into a society. (By convention, generations are numbered with immigrants as the ”first,” so that their children are the ”second,” their grandchildren are the ”third,” etc.) Generally speaking, later generations are expected to be more assimilated than earlier ones. Another basis for dissection is birth cohorts, defined as groups born during the same period. Cohort differences can provide insight into historical changes in a group’s position. Both kinds of differences have been used to study ethnic changes in the United States (for an application of the generational method, see Neidert and Farley 1985; for the cohort method, see Alba 1988).
The second major approach to the study of ethnicity and race, labeled above as ”stratification,” is considerably less unified than the assimilation approach, encompassing quite diverse theoretical underpinnings and research findings. Yet there are some common threads throughout. One is an assumption that ethnic groups generally are hierarchically ordered: There is typically a dominant, or superordinate, group, which is often described as the majority group (even though in some societies, such as South Africa, it may be a numerical minority of the population). There are also subordinate groups, often called minorities (although they may be numerical majorities). Second, these groups are assumed to be in conflict over scarce resources, which may relate to power, favorable occupational position, educational opportunity, and so forth. In this conflict, the dominant group employs a variety of strategies to defend or enhance its position, while minority groups seek to challenge it. Often, the focus of the stratification approach is on the mechanisms that help preserve ethnic inequalities, although there has also been some attention to the means that enable minorities successfully to challenge entrenched inequality.
One tradition in ethnic stratification research has looked to mechanisms of inequality that are rooted in ideologies and outlooks that are then manifested in the behavior of individuals. This is, in fact, a common meaning for the word racism. A longstanding research concern has been with prejudice, which is generally defined as a fixed set of opinions, attitudes, and feelings, usually unfavorable, about the members of a group (Allport 1954). Prejudice is frequently an outgrowth of ethnocentrism, the tendency to value positively one’s own group and denigrate others. It can lead to discrimination, which is a behavior: the denial of equal treatment to a group’s members, exemplified by the refusal to sell homes in certain neighborhoods to minority-group members. The investigation of prejudice was one of the early testing grounds for survey research. In the United States, this research uncovered a dimension of social distance, expressing the specific gradations of social intimacy the majority is willing to tolerate with the members of various ethnic groups (Bogardus 1928). Recent research has revealed a paradoxical set of changes: on the one hand, a secular decline in traditional prejudice, most notably the prejudiced attitudes and beliefs held by whites about blacks; on the other, little increase in support for government policies that implement principles of racial equality (Schuman et al. 1998). This divergence has led many scholars to theorize about the emergence of modern forms of prejudice, exemplified by the concept of symbolic racism (Kinder and Sears 1981).
However persuasive as explanatory factors prejudice and discrimination may appear to the layperson, sociologists have in recent decades more and more neglected them in favor of institutional, or structural, mechanisms of inequality (see, e.g., Bonilla-Silva 1997). One reason for this shift has been skepticism that prejudice and individual-level discrimination by themselves are adequate to account for the depth and durability of racial and ethnic cleavages in industrial societies, especially since these purported explanatory factors have seemed to decline in tandem with rising educational levels. (However, the emphasis on structural mechanisms can itself be faulted for neglecting the ideological component in racism.)
One expression of the focus on structural factors has been the notion of institutional racism (Blauner 1972). According to it, inequality among racial and ethnic groups depends not so much on individual acts of discrimination as it does on the workings of such institutions as the schools and the police, which process and sort individuals according to their racial and ethnic origins and ultimately impose very different outcomes on them. An assumption of this approach is that this sort of discrimination can occur on a wide scale without equally widespread prejudice. Indeed, it may even be possible without any discriminatory intent on the part of individuals in authority; an example would be educational tracking systems, which sort students according to racial background based on culturally and socially biased cues that are presumed by teachers and administrators to be related to intellectual ability. Studies deriving from the notion of institutional racism have in fact provided some compelling analyses of the perpetuation of inequalities (on education, see Persell 1977), although they also can easily descend into controversy, as when any unequal outcome is declared to indicate the operation of racism.
A crucial arena in which both institutional and individual forms of racism operate to produce inequality is that of residence. Residential segregation is probably the most prominent indicator of the persisting importance of race in the United States; and it is critical to many other inequalities because life chances, especially for children, vary sharply across neighborhoods, which differ in many ways, from the qualities of the housing and schools they contain to the risks that their residents will be victims of crime. Decades of research with highly refined data and measures, such as the well-known Index of Dissimilarity, have shown that levels of neighborhood segregation by race are quite high and, at best, moderating very slowly (Farley and Frey 1994; Massey and Denton 1993). The pattern in many American cities can be described as one of hypersegregation, in the sense that the overwhelming majority of African Americans reside in large, consolidated ghettos containing virtually no whites and few members of other groups (Massey and Denton 1993). Residential segregation is not much explained by the economic inequalities between whites and blacks. So-called audit studies, involving matched pairs of white and black housing applicants, reveal considerable outright discrimination in the housing market. While whites are now more willing than in the past to accept blacks as neighbors, it appears that their tolerance is usually limited to small numbers. Contemporary segregation is also the consequence of government policies, past and present. The policies of the Federal Housing Administration, which effectively led in the 1930s to the modern mortgage instrument but were biased against mortgage lending in areas with many black residents, have had an enduring impact of American residential patterns, reflected in the division between largely white suburbs and largely nonwhite cities. (For a thorough discussion of the mechanisms behind segregation, see Massey and Denton 1993.)
A major theme in the stratification approach is the often complicated relationship, or interaction, between ethnicity and social class. One viewpoint is that ethnicity is, to some degree at least, a manifestation of deeply rooted class dynamics. This has led to analyses that emphasize the economic and material foundations of what appear superficially to be cultural and ethnic distinctions. Analyses of this type have sometimes been inspired by Marxism, but they are hardly limited to Marxists. For example, Herbert Gans ([1962] 1982), in an influential analysis of second-generation Italians in a Boston neighborhood, argued that many of their distinctive traits could be understood as a function of their working-class position, which was not greatly changed from the situation of their southern Italian ancestors. In a related vein, Stephen Steinberg (1989) argues that cultural explanations of ethnic inequalities, which impute ”undesirable” characteristics to some groups and ”desirable” ones to others, are often rationalizations of economic privilege.
It is sometimes argued that inequalities that once rested on an ethnic basis now rest primarily on one of class. An important, if controversial, instance is William J. Wilson’s (1978, 1987) claim of a ”declining significance of race” for American blacks. One part of Wilson’s argument focuses on an increasing socioeconomic division within the black population. This is held to result from the increasing opportunities available to young, well-educated African Americans since the 1960s. However, while improvements have been registered for a minority of the group, the lot of the black poor has not improved—it has even worsened. Wilson describes their situation as one of an underclass, which he defines in terms of isolation from the mainstream economy and society. His explanations for the emergence of the underclass are structural, not individualistic, and include the spatial concentration of the black poor in rundown urban neighborhoods, which have been stripped of their institutional fabric and middle-class residents, and the exodus of suitable job opportunities from central cities to suburbs and Sunbelt areas. In opposition to Wilson, others have argued that the emergence of underclass ghettos is better understood as a consequence of racism, as exemplified in residential segregation (e.g., Massey and Denton 1993).
An economic approach has also been used to explain ethnic conflict, which is seen as an outgrowth of the conflicting material interests of different ethnic groups. An exempler is provided by the theory of the ethnically split labor market (Bonacich 1972). Such a labor market develops when two ethnically different groups of workers compete (or could compete) for the same jobs at different costs to employers. It is typical in such situations for the higher-priced group of workers to have the same ethnic origins as employers, and therefore for the lower-priced group to be ethnically different from both. Nevertheless, it is in the interest of employers to substitute lower-priced workers for higher-priced ones wherever possible, despite the ethnic ties they share with the latter. Intense ethnic conflict can therefore develop between the two groups of workers, as the higher-priced group seeks to eliminate the threat to its interests. Two strategies may be employed: exclusion of the lower-priced group (for example, through legal restrictions on the immigration of its members) or creation of a caste system, that is, the limitation of the lower-priced group to a separate sphere of undesirable jobs. Split-labor-market theory has been applied to black-white relations in South Africa and the United States.
Yet, even in terms of a strictly economic approach, the precise genesis of the conflict between different ethnic groups of workers is open to question, and the theory of segmented labor markets draws another picture (Piore 1979). This theory divides the economies and labor markets of advanced capitalist societies into a primary sector, which contains relatively secure, well-paid jobs with decent working conditions and the opportunity for advancement, and a secondary sector, made up of insecure, dead-end jobs at low wages. Regardless of their class position, workers from the dominant group prefer to avoid jobs in the secondary sector, and usually they can manage to do so. Even unemployment may not be sufficient to force them into the secondary sector, since the benefits and resources available to most members of the dominant group, such as relatively generous unemployment compensation and seniority rights, enable them to wait out periodic unemployment. Hence, there is a need for another supply of workers, typically drawn from minorities and immigrants, who have no alternative but to accept employment in the secondary sector. Immigrants, in fact, are often willing to take these jobs because, as sojourners, they find the social stigma attached to the work to be less meaningful than do the native-born. In contrast to the theory of the split labor market, which takes the existence of an ethnic difference among workers as a given, segmented-labor-market theory can explain why ethnic differences, especially between natives and immigrants, are so prevalent and persistent in the industrial societies of the West.
An economic explanation of ethnic differences is sometimes placed in a context of worldwide colonialism and capitalist exploitation (Rex 1981). Indeed, ethnic inequalities within a society are sometimes seen as the consequence of international relations between colonizers and the colonized. The notion that subordinate groups form economically exploited internal colonies in Western societies is an expression of this view (Blauner 1972). This notion is compatible with a hypothesis of a cultural division of labor, according to which positions in the socioeconomic order are assigned on the basis of cultural markers and hence ethnic origin (Hechter 1975).
The stratification approach need not focus exclusively on socioeconomic differences. Some scholars, in fact, prefer to see inequalities of power as more fundamental (Horowitz 1985; Stone 1985). This is a very general perspective on ethnic stratification, which is quite compatible with such fundamental notions as dominant and subordinate groups. According to it, social-class relations are but one instance, no matter how important, of the institutionalized inequalities between ethnic groups. Equally important, ethnic dominance cannot be reduced to, or explained solely in terms of, social-class mechanisms. (An implication is that class analysis of ethnic relations can be reductionist, an attempt to explain away ethnicity’s causal independence.) Thus, the antagonism and sectarian violence between Catholics and Protestants in Northern Ireland is not comprehensible solely in terms of a social-class analysis, even though aggregate class differences between the groups exist as a result of centuries-long Protestant domination. This domination, the legacy of the colonial treatment of Ireland by the British, is manifest in a number of areas—in separate residential neighborhoods and schools, in social relations between members of the two groups, and in the political system. In short, domination encompasses much more than social-class privilege and gives even working-class members of the Protestant group a sense of status and superiority.
Distinguishing empirically between ethnic stratification based on power and that rooted in economic structure has proven difficult. In one attempt, Hubert Blalock (1967) formulated a power threat hypothesis, to be contrasted with one derived from economic competition between groups. These two hypotheses can be tested in the relationship between discrimination and the size of a minority group. In particular, threats to the power of the dominant group are expected to result in discrimination that rises sharply with increases in the size of a minority; the same is not true for economic competition. So far, this test has been applied mainly to the American South (Tolnay and Beck 1995).
Theories concerning power differentials among ethnic groups border on the third major approach to the study of ethnicity, with its focus on ethnic-group resources. This approach, like the preceding one, takes its point of departure from the inequalities among groups. However, its vision is less one of the domination of some groups over others than it is of a more balanced competition that is affected by characteristics of the groups, such as their numbers, their solidarity, and their ability to form separate ethnic subeconomies. Such characteristics can give the group and its members relative advantages, or disadvantages, in this competition. Insofar as advantages are conferred, there may incentives for individuals to maintain their attachments to a group rather than assimilate. In a sense, theories of ethnic-group resources can be seen as counterarguments to assimilation theories.
This is certainly clear in Nathan Glazer and Daniel Patrick Moynihan’s ([1963] 1970) politically based explanation for the continuing importance of ethnicity in the United States. These authors acknowledge that immigrant cultures fade quickly under the impact of the assimilation process; assimilation is accomplished to this degree. However, ethnicity comes to coincide with differences in American circumstances, such as residential and occupational concentrations, which are similarly affected by government policies and actions. Hence, ethnicity takes on importance in the political sphere: Ethnic groups become ”interest groups,” reflecting the interests of many similarly situated individuals. This role breathes new life into what might otherwise languish as an Old World social form. Glazer and Moynihan give many examples of the working of such interest groups in New York City.
Others have argued that ethnicity has become ”politicized” in many contemporary societies, including many industrialized ones, and this leads to an unanticipated ethnic ”resurgence.” Daniel Bell (1975) states one basis for this point of view, claiming that politics is increasingly replacing the market as the chief instrument of distribution and that politics recognizes only group claims, thus enhancing ethnicity’s political import. (The direction of change seems to have shifted since then, however.) In a more general fashion, Francois Nielsen (1985) contends that ethnicity offers a wider basis for political recruitment than the chief alternative, social class, and therefore ethnic-based movements have a greater chance for success. Addressing the situation in Third World societies, Donald Horowitz (1985) sees the ethnic political conflict that troubles many of them as originating in part in colonial policies and then being intensified by the anxieties of groups over their status in the postcolonial order.
Students of ethnic politicization have focused especially on the phenomenon of ethnic mobilization, which is epitomized in separatist movements in modern states, as in the Congo, Quebec, and Tibet (Olzak 1983). Mobilization can be regarded as one manifestation of the state of ethnic solidarity, a core concept in the literature on ethnicity. Ethnic groups marked by solidarity can be defined as self-conscious communities whose members interact with each other to achieve common purposes, and mobilization occurs when members take some collective action to advance these purposes. Recent research on ethnic movements appears to demonstrate that they are not generally interpret-able in modern polities as the vestiges of traditional loyalties that have yet to be submerged by the modernization processes attendant upon development; rather, such movements can be outcomes of these processes and thus increase as economic development proceeds. The specific causes of this linkage are disputed, however (Olzak 1983).
Culture is another domain in which the search for group resources has been carried out. The group-resources approach is compatible with the cultural-pluralist description of society, described earlier. More commonly in the past than today, the relative success of ethnic groups has been explained in terms of cultural traits. Quite often, the advantages these give have been analyzed in social-psychological terms. A well-known attempt along these lines was that of Bernard Rosen (1959), who matched American ethnic groups against the profile of the ”achievement syndrome,” a configuration of values that was presumed to predispose individuals to success. Included was an orientation to the future rather than the past and a downplaying of fatalism. In Rosen’s analysis, the presence or absence of these traits in the culture of a group was explained according to the group’s history and experience, and frequently in terms of the culture of the society from which it came. This sort of analysis, presuming stable cultural traits and rooting socioeconomic success in social-psychological prerequisites, has fallen into disfavor of late. In fact, it is often seen more as popular myth than as social science (Steinberg 1989). Cultural explanations, however, are not limited to the social-psychological realm. As one example, Ivan Light (1972) has devised an intriguing partial explanation for the entrepreneurial proclivities of different groups— in terms of the extent to which their cultures sponsor mechanisms that generate capital for the start-up of small businesses. Light argues that the business success of some Asian groups can be understood in part as an outcome of the rotating-credit association, a traditional social form imported from their home societies. Nevertheless, sociologists recently have stressed the malleability of culture and have tended to view it more as an adaptation to, and hence outcome of, socioeconomic position than as a cause of it. Consequently, cultural interpretations currently play only a minor role in the study of ethnicity. This may be a neglect engendered by cyclical intellectual fashion—in the future, they may loom larger, especially in the analysis of the immigrant groups proliferating in many societies (see, e.g., Zhou and Bankston 1998).
A focus of intense interest in recent years has been on the economic resources some groups are able to attain and thus on the advantages and opportunities adhering to ethnicity for their members. This interest is expressed in somewhat divergent research streams, running along distinctive conceptual channels. One has been carved out around the linkage that appears in many societies between minority-group status and entrepreneurial activity. This linkage makes sense in terms of the disadvantages borne by many minority groups in mainstream economies, where they may suffer from various forms of discrimination and be channeled into low-status positions. Entrepreneurial activity, then, represents an attempt to evade these economic disadvantages. But, as many observers have pointed out, this hypothesis alone cannot account for the wide variation in entrepreneurial levels and forms among groups (Light 1984). An early attempt at a general theory is that of middleman minorities, which begins from the observation that a few groups, such as the Chinese and Jews, have occupied entrepreneurial niches in numerous societies (Bonacich 1973). The theory views entrepreurialism, especially in commercial forms, as consistent with the sojourner status and interstititial position of these groups, between elites and masses. Yet this theory is too specific to account for contemporary immigrant entrepreneurialism, which frequently, for instance, involves nonsojourner groups. Therefore, Ivan Light (1984) has turned toward broader conceptual accounting schemes, seeing a range of ethnic and class resources as behind entrepreneurialism. Very intriguing here is the notion of ethnic resources, which include the established ethnic networks that train newly arrived immigrants and set them up in business. This notion helps to account for the clustering that is so obvious in ethnic small business, such as the concentration of New York’s Koreans in greengroceries and dry cleaning.
A second stream of research builds partly on the immigrant proclivity toward entrepre-neurialism. The theory of the ethnic enclave, developed by Alejandro Portes and various collaborators, holds that some groups are able to form ethnic subeconomies, which shelter not only entrepreneurs but also workers from the disadvantages they would face in the secondary sectors of the mainstream economy (Portes and Bach 1985). Enclave theory is particularly interesting because it provides a material motive for resisting assimilation. As exemplified in the case of the Cubans of Miami, enclave-forming groups typically contain a high-status stratum composed of individuals with professional occupations, capital, or both, along with lower-status strata of workers seeking employment. A key to a full-fledged enclave is the establishment of networks of firms in interrelated economic sectors. The success of these firms is predicated on ethnic loyalties to an important degree. Workers and bosses may find mutual advantage where they share the same ethnicity. Workers may be willing to work longer hours or for lower wages, thus enhancing the profitability of a business, because they are able to work in a culturally familiar environment (usually speaking their mother tongue, for example). Workers may also have the opportunity to learn about running a business, and some eventually graduate to become entrepreneurs themselves. Enclave economies may be sustained in part by servicing the needs of their own ethnic communities, but if they are to be truly robust, they must also plug into the mainstream economy. This is the case, for instance, with the many ethnic firms in the American garment industry.
Despite its attractiveness on theoretical grounds, the implications of the enclave economy are disputed. One criticism is that such an economy offers few benefits for workers; the economic gains accrue to ethnic entrepreneurs (Sanders and Nee 1987). In addition, there is growing recognition that fully formed enclave economies are uncommon (Logan et al. 1994). These problems have led investigators to reinvigorate a broader notion, that of the ethnic niche, which refers to any economic position where a group is sufficiently concentrated to draw advantage from it, typically by being able to steer its own members into openings (Model 1997; Waldinger 1996). Thus, one can speak of an entreneurial niche, such as the Koreans have established in various lines of small business, or of an occupational niche, exemplified by the former dominance of the Irish in municipal employment in numerous cities. The ethnic-niche idea implicates other notions, such as the operation of ethnic networks that assist group members in finding positions in a niche. The niche idea is also frequently linked to that of an ethnic queue, an hierarchical ordering among minority groups, and to demographic and socioeconomic shifts. The latter open up niches, as exemplified by the withdrawal of white ethnics from some lines of small business in American cities, and the former determine which other groups are best positioned to take advantage of the openings (Waldinger 1996).
Oddly, a relatively neglected dimension of the ethnic-resources perspective concerns ethnic communities themselves, despite an almost universal recognition among scholars that some ethnics prefer to live in communities where their fellow ethnics are a numerous, if not the predominant, element of the population. Ethnic neighborhoods continue to be a salient aspect of metropolitan life in the United States and connected to inequalities that affect the well-being and life chances of their residents; and now they are emerging on a substantial scale in suburban settings (Horton 1995). A longstanding idea is that the residents of ”institutionally complete” communities are less likely to assimilate (Breton 1964). Given the burgeoning of immigrant communities in the United States and elsewhere, it seems certain that this scholarly neglect will soon be repaired.
The final orientation, the social constructionist one, is so new that it cannot yet be said to have accumulated a substantial corpus of findings (for general reviews, see Nagel 1994; Omi and Winant 1996). This is not to deny its roots in the classical conception of ethnicity, as exemplified by the Weberian definition cited earlier, with its emphasis on the subjective view and, implicitly, on the relation of ethnic boundaries to social closure. As its name implies, this perspective stresses that ethnic boundaries and meanings are not preordained but malleable and thus that an understanding of ethnic distinctions must generally be sought in contemporary, rather than historical, circumstances. This perspective is especially attuned to the possible emergence of new forms of ethnicity, as some see, for instance, in the cultural hybridity associated with the rising frequency of mixed racial ancestry in the United States. There is also an emphasis on heightened ethnic fluidity, even the potential for revolutionary shifts, in the present because of the rise of the global economy and associated phenomena, such as large-scale migrations across borders, cultural diffusions, and transnational networks (Sassen 1988).
The most striking findings to emerge from this perspective concern historical shifts in racial boundaries, confounding the lay view that racial distinctions are determined by impossible-to-overlook physical differences among humans. Yet, historically, such boundaries have shifted in the United States within relatively brief intervals. During their immigrations, Irish Catholics, Italians, and Eastern EuropeanJews were all perceived as racially different from native-born whites, as nineteenth-century cartoon depictions of the Irish demonstrate. Today, the racial element in the perception of ethnic differences among whites has disappeared. The limited historical studies have not yet fully explicated the decades-long process by which this happened but make clear that it was not benign, as immigrants used violence and stereotyping, among other means, to create social distance between themselves and African Americans and thereby make plausible their candidacy for ”whiteness” (see Ignatiev 1995; Roediger 1991). Then, their acceptability to other white Americans grew as they climbed the social ladder and mixed occupa-tionally and residentially with other whites.
In conclusion, one must acknowledge that the literature on ethnicity remains unsettled in its theoretical core. The persistence—perhaps even the resurgence—of ethnic difference and conflict in societies throughout the world has attracted much attention from sociologists and other social scientists. But the paradoxes associated with ethnicity, evidenced in the United States by the assimilation of some groups and the continued sepa-rateness and even subordination of others, have yet to be resolved. They remain fruitful for sociology, nevertheless: The study of ethnicity has produced some of the discipline’s most striking findings and, no doubt, will continue to do so.
