摘要:AbstractWe inquire into the possibilities for racial and ethnic change during the next quarter century in the US. The argument proceeds in three stages. First, we review the accounts of past racial and ethnic shifts and argue that they are incomplete. Based on social-boundary theory, we claim that non-zero-sum mobility combined with the ability to convert socioeconomic mobility into social proximity to mainstream whites were essential to the successful assimilation of previous racial/ethnic outsiders such as Irish Catholics. Next, we ask whether the condition of non-zero-sum mobility might exist in the present and near future. Our positive answer relies on data that show increasing minority penetration into the top tiers of occupations in the US and the prospects that diversity will grow there as the largely white baby-boom cohorts retire. In the final stage, we ask what contingencies that might affect the extent to which these opportunities for change are fulfilled.Social scientists take for granted that racial and ethnic origins play a critical, though hardly the exclusive, role in determining the life chances of Americans, whether these are a matter of where they live, how much education they get, what kinds of jobs they do, or whom they marry. An enormous literature establishes that this assumption is generally warranted. For some of the most salient racial divides in the US, such as that between blacks and whites, that literature also demonstrates that numerous differentials in life chances have remained stable for decades. A salient instance concerns residential life chances, which not only involve who the neighbors will be, but what the “quality” of the neighborhood is, where this may be reflected in the risk of criminal victimization, the adequacy of the schools, or other ways. The research on segregation reveals the stability of black-white differences in these respects for at least half a century (Massey and Denton, 1993; Logan et al. 2004). In this sense, one can say that there exists in the US a crystallized racial/ethnic order, with whites occupying the top position and African Americans at the bottom, with others somewhere in between.There are sound reasons to think that this order influences the chances for success of the second generation. The segmented-assimilation theory about the incorporation of new immigrant groups asserts as much (Portes and Zhou, 1993); and empirical research sustains the view that the children of immigrants find themselves in a society where their options are constrained by the way other Americans view their racial/ethnic membership, even when this does not coincide with the perceptions and beliefs their parents have brought from their societies of origin. Thus, Mary Waters's (1999) research on the West Indian second generation reveals how its members struggle against the racist views that white Americans generally have of individuals with visible African ancestry. The results from the New York second-generation project also demonstrate the importance of group memberships for the trajectories of individuals through early adulthood (Kasinitz et al., forthcoming).Yet, looking at the past, we also know that the emergence of a massive second generation can unsettle racial and ethnic relations and lead to a reshuffling of the hierarchical order among groups. Enormous literatures, on assimilation and on whiteness, address these changes in the past. They suggest that we ought at least to entertain the notion of some degree of racial and ethnic change in the next few decades, with the arrival of a large and upwardly striving second generation on the scene (see also Alba and Nee, 2003; Bean and Stevens, 2003). This is what we would like to do in this paper on the basis of some results concerning racial and ethnic shifts among the incumbents of highly ranking occupations. We will argue that the shifts already evident, resulting probably from demography and affirmative action, combined with those that can be anticipated as a result of the departure of the baby boom cohorts of whites from the labor market, indicate an opportunity for important changes to the racial/ethnic order in the next two decades. If this opportunity is realized, something that is certainly not assured, it will likely usher in a period of unsettlement, when established assumptions about relations between individuals derived from their categorical memberships lose their certainty. This is the justification for the appearance of the word “liminality” in the title, for it refers to a period of transition, when previously fixed identities are suspended. In the final part of the paper, We will consider some contingencies that, based on our knowledge about the incorporation of earlier immigrant groups, will affect whether this opportunity is realized or not.