by David Reich
Future studies of sex-biased mixture, though, will take full advantage of the power of the whole genome. Whole-genome studies can compare the thousands of independent genealogies recorded on chromosome X to the tens of thousands of independent genealogies carried in the rest of the genome. Comparison of genetic variation on the X chromosome to genetic variation elsewhere in the genome should in theory give far more statistical resolution, but while some studies of this sort have been revealing, the accuracy of their estimates has so far been disappointing, which may be because of intense bouts of natural selection that have affected chromosome X more than other chromosomes and that make interpretations of its patterns more difficult. So while many major mixture events—such as those of steppe pastoralists and farmers in Europe, or of Neanderthals and Denisovans with modern humans further back in human prehistory—may well have been sex-biased, detecting them by comparing fractions of ancestry on chromosome X to the rest of the genome is currently challenging.41 But our present problems with making precise estimates of sex bias based on the X chromosome are to a large extent technical, reflecting the limitations of the statistical techniques currently available. New methods that will be developed over the coming years will unleash the full power of comparison of the X chromosome to the rest of the genome. I hope that these improved methods, along with direct ancient DNA data from people who lived at the times mixture was happening, will enable new genomic insights into the nature of inequality in the deep human past.
The genomic evidence of the antiquity of inequality—between men and women, and between people of the same sex but with greater and lesser power—is sobering in light of the undeniable persistence of inequality today. One possible response might be to conclude that inequality is part of human nature and that we should just accept it. But I think the lesson is just the opposite. Constant effort to struggle against our demons—against the social and behavioral habits that are built into our biology—is one of the ennobling behaviors of which we humans as a species are capable, and which has been critical to many of our triumphs and achievements. Evidence of the antiquity of inequality should motivate us to deal in a more sophisticated way with it today, and to behave a little better in our own time.
11
The Genomics of Race and Identity
Fear of Biological Difference
When I started my first academic job in 2003, I bet my career on the idea that the history of mixture of West Africans and Europeans in the Americas would make it possible to find risk factors that contribute to health disparities for diseases like prostate cancer, which occurs at about a rate 1.7 times higher in African Americans than in European Americans.1 This particular disparity had not been possible to explain based on dietary and environmental differences across populations, suggesting that genetic factors might play a role.
African Americans today derive about 80 percent of their ancestry from enslaved Africans brought to North America between the sixteenth and nineteenth centuries. In a large group of African Americans, the proportion of African ancestry at any one location in the genome is expected to be close to the average (defining the proportion of African ancestry as the fraction of ancestors that were in West Africa before around five hundred years ago). However, if there are risk factors for prostate cancer that occur at higher frequency in West Africans than in Europeans, then African Americans with prostate cancer are expected to have inherited more African ancestry than the average in the vicinity of these genetic variations. This idea can be used to pinpoint disease genes.
To make such studies possible, I set up a molecular biology laboratory to identify mutations that differed in frequency between West Africans and Europeans. My colleagues and I developed methods that used information from these mutations to identify where in the genome people harbor segments of DNA derived from their West African and European ancestors.2 To prove that these ideas worked in practice, we applied them to many traits, including prostate cancer, uterine fibroids, late-stage kidney disease, multiple sclerosis, low white blood cell count, and type 2 diabetes.
In 2006, my colleagues and I applied our methods to 1,597 African American men with prostate cancer, and found that in one region of the genome, they had about 2.8 percent more African ancestry than the average in the rest of their genomes.3 The odds of seeing a rise in African ancestry this large by accident were about ten million to one. When we looked in more detail, we found that this region contained at least seven independent risk factors for prostate cancer, all more common in West Africans than in Europeans.4 Our findings could account entirely for the higher rate of prostate cancer in African Americans than in European Americans. We could conclude this because African Americans who happen to have entirely European ancestry in this small section of their genomes had about the same risk for prostate cancer as random European Americans.5
In 2008, I gave a talk about my work on prostate cancer to a conference on health disparities across ethnic groups in the United States. In my talk, I tried to communicate my excitement about the scientific approach and my conviction that it could help to find genetic risk factors for other diseases. Afterward, though, I was angrily questioned by an anthropologist in the audience, who believed that by studying “West African” or “European” segments of DNA to understand biological differences between groups, I was flirting with racism. Her questions were seconded by several others, and I encountered similar responses at other meetings. A legal ethicist who heard me talk on a similar theme suggested that I might want to refer to the populations from which African Americans descend as “cluster A” and “cluster B.” But I replied that it would be dishonest to disguise the model of history that was driving this work. Every feature of the data I looked at suggested that this model was a scientifically meaningful one, providing accurate estimates of where in the genome people harbor segments of DNA from ancestors who lived in West Africa or in Europe in the last twenty generations, prior to the mixture caused by colonialism and the slave trade. It was also clear that the approach was identifying real risk factors for disease that differ in frequency across populations, leading to discoveries with the potential to improve health.
Far from being extremists, my questioners were articulating a mainstream view about the danger of work exploring biological differences among human populations. In 1942, the anthropologist Ashley Montagu wrote Man’s Most Dangerous Myth: The Fallacy of Race, arguing that race is a social concept and has no biological reality, and setting the tone for how anthropologists and many biologists have discussed this issue ever since.6 A classic example often cited is the inconsistent definition of “black.” In the United States, people tend to be called “black” if they have sub-Saharan African ancestry—even if it is a small fraction and even if their skin color is very light. In Great Britain, “black” tends to mean anyone with sub-Saharan African ancestry who also has dark skin. In Brazil, the definition is different yet again: a person is only “black” if he or she is entirely African in ancestry. If “black” has so many inconsistent definitions, how can there be any biological meaning to “race”?
Beginning in 1972, genetic arguments began to be incorporated into the assertions that anthropologists were making about the lack of substantial biological differences among human populations. In that year, Richard Lewontin published a study of variation in protein types in blood.7 He grouped the populations he analyzed into seven “races”—West Eurasians, Africans, East Asians, South Asians, Native Americans, Oceanians, and indigenous Australians—and found that around 85 percent of variation in the protein types could be accounted for by variation within populations and “races,” and only 15 percent by variation across them. He concluded: “Races and populations are remarkably similar to each other, with the largest part by far of human variation being accounted for by the differences between individuals. Human racial classification is of no social value and is positively destructive of social and human relations. Since such racial classification is now seen to be of virtually no ge
netic or taxonomic significance either, no justification can be offered for its continuance.”
In this way, through the collaboration of anthropologists and geneticists, a consensus was established that there are no differences among human populations that are large enough to support the concept of “biological race.” Lewontin’s results made it clear that for the great majority of traits, human populations overlap to such a degree that it is impossible to identify a single biological trait that distinguishes people in any two groups, which is intuitively what some people think of when they conceive of “biological race.”
But this consensus view of many anthropologists and geneticists has morphed, seemingly without questioning, into an orthodoxy that the biological differences among human populations are so modest that they should in practice be ignored—and moreover, because the issues are so fraught, that study of biological differences among populations should be avoided if at all possible. It should come as no surprise, then, that some anthropologists and sociologists see genetic research into differences across populations, even if done in a well-intentioned way, as problematic. They are concerned that work on such differences will be used to validate concepts of race that should be considered discredited. They see this work as located on a slippery slope to the kinds of pseudoscientific arguments about biological difference that were used in the past to try to justify the slave trade, the eugenics movement to sterilize the disabled as biologically defective, and the Nazis’ murder of six million Jews.
The concern is so acute that the political scientist Jacqueline Stevens has even suggested that research and even emails discussing biological differences across populations should be banned, and that the United States “should issue a regulation prohibiting its staff or grantees…from publishing in any form—including internal documents and citations to other studies—claims about genetics associated with variables of race, ethnicity, nationality, or any other category of population that is observed or imagined as heritable unless statistically significant disparities between groups exist and description of these will yield clear benefits for public health, as deemed by a standing committee to which these claims must be submitted and authorized.”8
The Language of Ancestry
But whether we like it or not, there is no stopping the genome revolution. The results that it is producing are making it impossible to maintain the orthodoxy established over the last half century, as they are revealing hard evidence of substantial differences across populations.
The first major engagement between the genome revolution and anthropological orthodoxy came in 2002, when Marc Feldman and his colleagues showed that by studying enough places in the genome—they analyzed 377 variable positions—it is possible to group most people in a worldwide population sample into clusters that correlate strongly to popular categories of race in the United States: “African,” “European,” “East Asian,” “Oceanian,” or “Native American.”9 While Feldman’s conclusions were broadly consistent with Lewontin’s in that his data also showed more variation within groups than among them, his study defined clusters in terms of combinations of mutations instead of looking at mutations individually as Lewontin had done.
Scientists were quick to respond. One was Svante Pääbo, who eight years later would go on to lead the work to sequence whole genomes of archaic Neanderthals and Denisovans. Pääbo came to the debate about the nature of human population structure as a founding director of the Max Planck Institute for Evolutionary Anthropology in Leipzig, which was set up in 1997 in an effort to return Germany to a field in which it had played a leading role before the Second World War but that it had largely abandoned due to anthropologists’ central contribution to developing Nazi race theory.
Pääbo took seriously his moral responsibility as head of an ambitious German institute of anthropology, and wondered whether the truth about human population structure could be more like the anthropologist Frank Livingston’s suggestion that “there are no races, there are only clines”—a view in which human genetic variation is characterized by gradual geographic gradients that reflect interbreeding among neighbors.10 To explore this possibility, Pääbo investigated whether the clusters the Feldman study found appeared sharply defined because the analyzed populations had been sampled in a nonrandom fashion across the world. To understand how nonrandom sampling could contribute to this result, consider the United States, which harbors extraordinary diversity, but where genetic discontinuities among groups such as African Americans, European Americans, and East Asians are sharper than in the places from which immigrant populations came because the United States has drawn its immigrants from a subset of world locations. For example, in the United States, most of the African ancestry is from a handful of groups in West Africa,11 most of the European ancestry is from northwest Europe, and most of the Asian ancestry is from Northeast Asia. Pääbo showed that such nonrandom sampling could account for some of the effects Feldman and colleagues observed. However, later work proved that nonrandom sampling could not account for most of the structure, as substantial clustering of human populations is observed even when repeating analyses on geographically more evenly distributed sets of samples.12
Another flurry of discussion followed a 2003 paper led by Neil Risch, who argued that racial grouping is useful in medical research, not just to adjust for socioeconomic and cultural differences, but also because it correlates with genetic differences that are important to know about when diagnosing and treating disease.13 Risch was convinced by examples like sickle cell disease, which occurs far more often in African Americans than in other populations in the United States. He argued that it was appropriate for doctors to be more likely to think of sickle cell disease if the patient is African American.
In 2005, the U.S. Food and Drug Administration lent support to this way of thinking when it approved BiDil, a combination of two medications approved to treat heart failure in African Americans because data suggested it was more effective in African Americans than in European Americans. But on the other side of the argument, David Goldstein suggested that U.S. racial categories are so weakly predictive of most biological outcomes that they do not have long-term value.14 He and his colleagues showed that the frequencies of genetic variants that determine dangerous reactions to drugs are poorly predicted by U.S. census categories. He acknowledged that the reliance on racial and ethnic categories is useful given our poor present knowledge, but predicted that the future will involve testing individuals directly for what mutations they have, and doing away altogether with racial classification as a basis for making individualized decisions about care.
Against this backdrop of controversy emerged work like mine, focusing on methods to determine population origin not just of our ancestors but also of individual segments of our genomes. The anthropologist Duana Fullwiley has written that the development of what she calls “admixture technology” and the language of “ancestry” that geneticists like me have adopted is a reversion to traditional ideas of biological race.15 She has pointed out that in the United States, the “ancestry” terms that we use map relatively closely to traditional racial categories, and her view is that the population genetics community has invented a set of euphemisms to discuss topics that had become taboo. The belief that we have embraced euphemisms is also shared by some on the other side of the political spectrum. At a 2010 meeting I attended at Cold Spring Harbor Laboratory, the journalist Nicholas Wade described his resentment of the population genetics community’s “ancestry” terminology, asserting that “race is a perfectly good English word.”
But “ancestry” is not a euphemism, nor is it synonymous with “race.” Instead, the term is born of an urgent need to come up with a precise language to discuss genetic differences among people at a time when scientific developments have finally provided the tools to detect them. It is now undeniable that there are nontrivial average genetic differences across populations in multiple traits, and the race vocabulary is too ill-defined and too
loaded with historical baggage to be helpful. If we continue to use it we will not be able to escape the current debate, which is mired in an argument between two indefensible positions. On the one side there are beliefs about the nature of the differences that are grounded in bigotry and have little basis in reality. On the other side there is the idea that any biological differences among populations are so modest that as a matter of social policy they can be ignored and papered over. It is time to move on from this paralyzing false dichotomy and to figure out what the genome is actually telling us.
Real Biological Difference
I have deep sympathy for the concern that genetic discoveries about differences among populations may be misused to justify racism. But it is precisely because of this sympathy that I am worried that people who deny the possibility of substantial biological differences among populations across a range of traits are digging themselves into an indefensible position, one that will not survive the onslaught of science. In the last couple of decades, most population geneticists have sought to avoid contradicting the orthodoxy. When asked about the possibility of biological differences among human populations, we have tended to obfuscate, making mathematical statements in the spirit of Richard Lewontin about the average difference between individuals from within any one population being around six times greater than the average difference between populations. We point out that the mutations that underlie some traits that differ dramatically across populations—the classic example is skin color—are unusual, and that when we look across the genome it is clear that the typical differences in frequencies of mutations across populations are far less.16 But this carefully worded formulation is deliberately masking the possibility of substantial average differences in biological traits across populations.
