How has growing knowledge about human genetics affected how American textbooks present race? This article analyzes 80 biology textbooks published from 1952 to 2002 to reveal that U.S. biology texts have pursued the topic of race with renewed vigor in recent years. Moreover, textbooks have redefined race as genetic without furnishing empirical evidence for this framing. The textbooks' transformation sheds light on the broader relationship between race and science in the United States, where claims about racial difference have not only drawn instrumentally and selectively from empirical research, but at times forgo scientific grounding altogether. As the textbooks show, both the tight and the loose linkage of race to science can preserve the cultural authority of the race concept. The texts also make clear that race is not a one-time construct or a relic of centuries past. Instead, it is continually remade -- and is being reworked today -- suggesting its dynamic adaptation for ongoing use as a fundamental tool of social stratification.

Despite their longstanding interest in race, American sociologists have conducted little empirical research on sociodemographic patterns or longitudin

What Brubaker [2016. “The Dolezal Affair: Race, Gender, and the Micropolitics of Identity.” Ethnic and Racial Studies 39 (3): 414–448] referred to in these pages as “the Dolezal affair”, though widely and intensely scrutinized, was only one example of a broader series of recent public controversies over an individual's “true” racial identity. This article argues that in the early-twenty-first century, claims of race-group membership are being complicated by technological developments in genetics and in cosmetics, as well as by new respect for subjective self-identification. As a result, there are more paths than ever to claiming and demonstrating racial belonging. In particular, I suggest that four new types of race-group member are emerging: genetic, cosmetic, emotive, and constructed. Should these types come to be widely accepted as genuine race members, racial groups will become more heterogeneous, resembling kaleidoscopic arrays of core and peripheral members who differ in terms of how many qualifications for belonging they may legitimately claim.

Shiao, Bode, Beyer, and Selvig argue that the theory of race as a social construct should be revisited in light of recent genetic research, which they interpret as demonstrating that human biological variation is patterned in “clinal classes” that are homologous to races. In this reply, I examine both their claims and the genetics literature they cite, concluding that not only does constructivist theory already accommodate the contemporary study of human biology, but few geneticists portray their work as bearing on race. Equally important, methods for statistically identifying DNA-based clusters within the human species are shaped by several design features that offer opportunities for the incorporation of cultural assumptions about difference. As a result, Shiao et al.’s theoretical distinction between social race and biological “clinal class” is empirically jeopardized by the fact that even our best attempts at objectively recording “natural” human groupings are socially conditioned.

Estimates of the size of the multiracial population in the United States depend on what prompts people to report multiple races on censuses and surveys. We use data from the 2015 Pew Survey of Multiracial Adults to explore how racial self-identification is shaped by the generational locus of an individual’s multiracial ancestry—that is, the place in one’s family tree where the earliest interracial union appears. We develop the theoretical rationale for considering generational heterogeneity and provide its first empirical demonstration for U.S. adults, by estimating what shares of the population identify multiracial ancestry in their parents’ or grandparents’ generation, or further back in their family tree. We find that multiracial generation is related to—and likely confounded with—the ancestry combinations that individuals report (e.g., white-Asian, black–American Indian). Finally, we show that later generations are less likely than their first-generation counterparts to select multiple races when they self-identify. Consequently, we argue that generational locus of multiracial ancestry should be taken into account by demographers and researchers who study outcomes for multiracial Americans.