"...for he has no right to give names to objects which he cannot define."
Do "races" exist as meaningful biological categories? Physical anthropologists and human biologists have been studying race (e.g., blacks vs. whites, or Europeans vs. Asians) for centuries. For most of that time, they subscribed to the perspective that race was a taxonomic category, and they sought to identify the biological characteristics (such as cranial shape or skin color) that characterized and defined these different groups. This perspective assumed that the differences between racial categories were biological, and that these categories were predictive of other traits (such as ancestry, temperament, intelligence, or health).
But it gradually became clear that this understanding was not scientifically sound. Groupings of people by skin color did not produce the same result as groupings of people by skull shape or blood type. Furthermore, as scientists began to study human variation with the tools of genetics (in the process creating my fields, anthropological genetics and human population genetics), it became apparent that human genetic variation does not divide humans into a few discrete groups. There are virtually no sharp boundaries, either with physical features or with patterns of genetic diversity, that show where one population "ends" and the next "begins."
These observations have led the majority of physical anthropologists, human biologists, and human geneticists in recent decades to conclude that the racial groups we recognize are social categories constructed in a specific cultural and historical setting, even if we consider physical features when categorizing people. These social categories can have biological consequences. (For example, someone who experiences the stress of racism may be more likely to develop high blood pressure and hypertension than someone who does not.)
However, according to former New York Times science writer Nicholas Wade, we should never have stopped thinking of race as a taxonomic category. In his new book, A Troublesome Inheritance: Genes, Race and Human History, Wade claims that the latest genomic findings actually support dividing humans into discrete races, and that the genetic makeup of different races contributes to behavioral and economic disparities.
Others have already critiqued this book, particularly Wade's storytelling in chapters 6 through 10 ("a much more speculative arena," as he puts it), in which he explains that English populations have a genetically based "willingness to save and delay gratification," and that Jews are adapted for capitalism in a manner analogous to the Eskimos' adaptation to survival in an Arctic environment (p. 214) -- assertions unsupported by scientific evidence, to put it mildly.
I'm far more interested in the central premise of Wade's argument, which is passing unchallenged by all but a few reviews. To support his claim, Wade relies heavily on a 2002 paper (by Rosenberg et al.) that used a program called structure to group people based on similarities in markers distributed across the genome. He notes that the program identified five major clusters in this 2002 study, which corresponded to the major geographic regions (Africa, Eurasia, East Asia, Oceania, and America) of the world. Therefore, Wade argues, these results clearly show that humans are divided into racial categories that match continents.
But structure didn't simply identify five clusters. It also identified two, three, four, six, and seven clusters. (Rosenberg et al. 2002 actually identified up to 20 divisions, but 1 through 7 are the primary ones they discussed. They also divided their worldwide sample into regions and then ran structure within those regions to look at more fine-scale population structure.)
Why? Researchers using structure have to define the number (K) of clusters in advance, because that's what the program requires. The program was designed to partition individuals into whatever pre-specified number of clusters the researcher requests, regardless of whether that number of divisions really exists in nature. In other words, if the researcher tells structure to divide the sampled individuals into four clusters, structure will identify four groups no matter what. Structure's results are extremely sensitive to many different factors, including models used, the type and number of genetic variants studied, and the number of populations included in the analysis (Rosenberg et al. 2005).
Figure 1 from Rosenberg et al. 2002 showing structure runs at two, three, four, five, and six populations. Each population is separated by a black line. Each thin vertical line represents one person. Ancestry groupings inferred from the program on the basis of genetic similarity are represented by different colors, so that a thin vertical line that is approximately 60-percent purple and approximately 40-percent orange indicates a person who was inferred to have 60-percent ancestry from the "purple" genetic cluster and 40-percent ancestry from the "orange" genetic cluster.
So when Rosenberg et al. 2002 told structure to use K=6, they got six clusters, with the sixth corresponding to a northwestern Pakistani group, the Kalash. Does this make the Kalash a separate race? Wade doesn't think so. When they told structure to use K=3, they got three clusters back, corresponding to Africa, Europe/Middle East/South Asia, and East Asia/Oceania/Americas. So are Native Americans and Australians not separate races? Rosenberg et al. never published any statistical evidence that justifies picking five races instead of seven, or four, or two (although such methods do exist; see Bolnick et al. 2008). Wade seems to like K=5 simply because it matches his preconceived notions of what race should be:
It might be reasonable to elevate the Indian and Middle Eastern groups to the level of major races, making seven in all. But then many more subpopulations could be declared races, so to keep things simple, the five-race, continent-based scheme seems the most practical for most purposes. (p. 100)
Practical. Simple. So by his own admission, Wade wants us to cut up human diversity into five races not because that's what the statistical analyses show but because thinking about it as a gradient is hard.
I've focused this review on technical details because I think that it's important that non-geneticists understand how Wade is distorting the results of recent research on genome-wide human variation. I won't speculate on whether this distortion is deliberate or a result of simple ignorance about genetics, but it is serious. Wade's book is all pseudoscientific rubbish, because he can't justify his first and primary point: his claim that the human racial groups we recognize today culturally are scientifically meaningful, discrete biological divisions of humans. This claim provides a direct basis for the whole second half of the book, in which he makes speculative arguments about national character. In other words, the entire book is a house of cards.
Bolnick DA. 2008. Individual ancestry inference and the reification of race as a biological phenomenon. In: Koenig BA, Lee SS-J, Richardson SS, editors. Revisiting Race in a Genomic Age. New Brunswick, N.J.: Rutgers University Press. pp. 77-85.
Rosenberg, NA., et al., 2002. Genetic structure of human populations. Science 298: 2381-2385.
Rosenberg N.A., et al. 2005. Clines, Clusters, and the effect of study design on the inference of human population structure. PLoS Genetics 1, 660-671.