Who We Are and How We Got Here

by David Reich

  What’s Next for Understanding the African Story

  Some of the most striking examples of the complexity of human population structure in Africa are the patterns of natural selection on the continent. People of West African ancestry today have a high rate of sickle cell disease, conferred by a mutation that changes the blood protein hemoglobin, the molecule that more than any other is responsible for ferrying oxygen around the body. This mutation has risen to substantial frequency under the pressure of natural selection in several places in Africa: in far West Africa (e.g., Senegal), in west-central Africa (e.g., Nigeria), and in central Africa (whence the mutation spread to eastern Africa and southern Africa via the migrations associated with the Bantu expansion). The reason this mutation has risen to such a high frequency in each of these populations is that if a person carries one copy of the mutation from either of his or her parents, it protects against the infectious disease malaria. Malaria is so dangerous that the protection provided to the approximately 20 percent of the population who carry one copy of the sickle cell mutation is balanced in evolutionary terms with the cost that the approximately 1 percent of the population has to pay in carrying two copies of the mutation and suffering from sickle cell disease, which kills in childhood without treatment. Strikingly, the mutation has arisen independently in each of three locations in Africa, which we know from the fact that the sequences on which it resides are all different. From a naive perspective this seems surprising, as one would think that a mutation like this would be so advantageous to the people who carry it that once it arose it would spread around the vast malaria zone of Africa propelled by a tailwind of natural selection if there was even a small rate of interbreeding among neighbors.42 A similar pattern is seen for the mutations in the lactase gene that confer an ability to digest cow’s milk into adulthood. The genetic basis for lactase persistence is completely different in North Africans and in the Fulani of West Africa than it is in the Masai of Sudan and Kenya, who carry different mutations, albeit in the same gene.43

  As Peter Ralph and Graham Coop have shown, the multiple origins in Africa of sickle cell mutations and of mutations that allow people to digest cow’s milk imply that the rate of migration among these populations—even in parts of sub-Saharan Africa less than a couple of thousand kilometers from each other—has been extraordinarily low since the need for these mutations arose. As a result, the most efficient way for evolutionary forces to spread beneficial mutations has often been to invent mutations anew rather than to import them from other populations.44 The limited migration rates between some regions of Africa over the last few thousand years has resulted in what Ralph and Coop have described as a “tessellated” pattern of population structure in Africa. Tessellation is a mathematical term for a landscape of tiles—regions of genetic homogeneity demarcated by sharp boundaries—that is expected to form when the process of homogenization due to gene exchanges among neighbors competes with the process of generating new advantageous variations in each region. The size of the regions where the same sickle cell mutation or same lactase-persistence mutation prevails reflects the rate of gene exchange among neighboring populations in Africa over the last thousands of years.

  Our understanding of African population history is still in its early stages, but it is already clear that the story is complicated, with separations within major lineages such as East African Foragers and South African Foragers dating back deep in time, and layers of mixture beyond the most recent ones that have arisen due to the spread of agriculture. Eventually, by obtaining many more samples of ancient DNA from Africa, we will be able to comprehend the range of human variation in Africa in the last tens of thousands of years and make meaningful reconstructions of population structure.

  What we can already be sure of is that in Africa, as in every region that has yielded ancient DNA, the model of an evolutionary tree in which today’s populations have remained unchanged and separate since branching from a central trunk is dead, and that instead the truth has involved great cycles of population separation and mixture. What we can be sure of, too, is that in Africa, as in every world region that has yielded ancient DNA, the data will disprove many commonly held assumptions. The implications of this complexity for society, and for the way we need to rethink who we are, is the theme of part III of this book.

  Part III

  The Disruptive Genome

  10

  The Genomics of Inequality

  The Great Mixing

  The American melting pot began to swirl almost as soon as Christopher Columbus arrived in 1492. European colonists, their African slaves, and the indigenous Americans were from populations whose ancestors had been isolated from each other for tens of thousands of years. Within a few years of meeting they began mixing, founding new populations that today number in the hundreds of millions.

  Martín Cortés “el Mestizo” belonged to one of the first of those populations. He was born within four years of the start of the 1519 military campaign in which his father, Hernán Cortés, led just five hundred soldiers to overthrow the Aztec Empire that dominated Mexico. His mother, “La Malinche,” was one of twenty female captives given over to the Spanish after a battle and she first served as an interpreter before becoming the mistress of Hernán Cortés. The Spanish quickly invented a term for the people of combined European and Native American ancestry who emerged from unions like this. “Mestizo” comes from the Spanish word mestizaje, which in English means miscegenation—the mixing of different “racial” types. To maintain their status in the social hierarchy, the Spanish and Portuguese set up a casta system in which people of entirely European ancestry (especially those born in Europe) had the highest status, while people who had even some non-European ancestry had lower status. This system collapsed under the demographic inevitability of admixture; within a few centuries people of entirely European ancestry were either an extreme minority or gone, and it was no longer feasible to limit power to those with entirely European ancestry. Following the independence movements of the nineteenth and early twentieth centuries, mixed ancestry became a source of pride in South and Central America. In Mexico, it defines national identity.1

  Migration of Africans to the Americas after 1492 occurred on a similar scale as migration of Europeans. All told, an estimated twelve million enslaved Africans were forced to make the journey, jammed into the holds of ships before being sold at auction.2 Slave traders from Spain, Portugal, France, Britain, and the United States made great fortunes by satisfying the colonialists’ need for manual labor. African slaves worked in the silver mines of Peru and Mexico and raised crops such as sugarcane and eventually tobacco and cotton. Africans were less affected by Old World diseases than Native Americans and easier to exploit than indigenous people, as they were far from home and scattered among a population that did not speak their languages. Deprived of their cultural points of reference, slaves had little ability to organize or resist. Most were sold in South America or the Caribbean, where they were often worked to death. Around 5 to 10 percent were brought to what became the United States. Following the first recorded sale of slaves by Portuguese traders in 1526, the rate of importation into the New World increased, plateauing at around seventy-five thousand per year until the trans-Atlantic slave trade was outlawed—in the British colonies in 1807, in the United States in 1808, and in Brazil in 1850.

  Today there are hundreds of millions of people in the Americas with African ancestry, the largest numbers in Brazil, the Caribbean, and the United States. The mixing of three highly divergent populations in the Americas—Europeans, indigenous people, and sub-Saharan Africans—that began almost five hundred years ago continues to this day. Even in the United States, where European Americans are still in the majority, African Americans and Latinos comprise around a third of the population. Nearly all individuals from these mixed populations derive large stretches of their genomes from ancestors who lived on different continents fewer than twenty generations ago. A small percentage of Eur
opean Americans have large stretches of African or Native American DNA as well, the legacy of people who successfully “passed” themselves off into the white majority.3

  A 1973 science-fiction novel, Piers Anthony’s Race Against Time, envisions a future in which the mixing of populations initiated by European colonialism reaches its inevitable conclusion, and by the year 2300 nearly all humans belong to a “Standard” population.4 In that year, only six unmixed people are left: one pair of “purebred Caucasians,” one pair of “purebred Africans,” and one pair of “purebred Chinese.” These “purebreds” are being raised in human zoos by foster parents and are being groomed to breed with the only remaining individual of similar ancestry to sustain humanity’s diversity, a diversity that is viewed by the “Standard” population as a resource of irreplaceable biological value on the verge of being lost. The premise of the novel is that the centuries after 1492 were a uniquely homogenizing time in the history of our species, a period of unprecedented mixing of previously separated populations enabled by transoceanic travel, which brought together groups whose ancestors had not been in contact with one another for tens or hundreds of thousands of years.

  But this premise is mistaken. The genome revolution has shown that we are not living in particularly special times when viewed from the perspective of the great sweep of the human past. Mixtures of highly divergent groups have happened time and again, homogenizing populations just as divergent from one another as Europeans, Africans, and Native Americans. And in many of these great admixtures, a central theme has been the coupling of men with social power in one population and women from the other.

  Founding Fathers

  Not long after the Constitutional Convention of 1787, the man who would become the United States’ third president, Thomas Jefferson, began a sexual relationship with his slave Sally Hemings. Jefferson owned a large plantation in the state of Virginia, where some 40 percent of the population was enslaved.5 Sally Hemings was an African American slave with three European grandparents. But her mother’s mother was a slave of African descent, and under Virginia law the status of a slave was maternally inherited. Jefferson and Hemings had six children together.6

  The Jefferson-Hemings relationship has been disputed by some who have suggested that Jefferson—who is America’s greatest Enlightenment thinker and the author of the U.S. Declaration of Independence—would not have maintained an illegitimate family. However, a genetic study published in 1998 revealed a Y-chromosome match between the male-line descendants of Eston Hemings Jefferson, the youngest son of Sally Hemings, and the male-line descendants of Jefferson’s paternal uncle.7 The genetic findings could in theory be explained if a male relative of Jefferson was the father rather than Jefferson himself. But there is no historical evidence for this possibility, and there is a credible nineteenth-century account of the Hemings-Jefferson relationship from Madison Hemings, another son of Hemings. A study by the Thomas Jefferson Memorial Foundation in 2000 concluded that, with high probability, the story was true.8

  According to the account of Madison Hemings, his mother had a chance at freedom because she joined Jefferson in France, where slavery was illegal, but she agreed to return as a slave to the United States with Jefferson under the condition that their children would eventually be set free. Hemings was thirty years younger than Jefferson, and in France, where she began her relationship with him between the ages of fourteen and sixteen, she was dependent on him. She was also the half sister of Jefferson’s wife, Martha Randolph, who had died of complications of childbirth several years earlier and whose father had a secret relationship with the mother of Sally Hemings.9

  Historians have attempted to quantify how widespread families like these were in the United States. Mixed-ancestry unions were often unrecorded, and when they were, children were categorized in different ways by different states. Genetics can help here. Although so far no one has analyzed DNA from African American graveyards to chart the emergence of a mixed-ancestry community in the United States, genetic studies of the present-day African American population are already enriching our understanding. Mark Shriver led a 2001 study that analyzed mutations that are extremely different in frequency between present-day Europeans and West Africans in order to study the African American populations of South Carolina. Shriver and his colleagues used these results to estimate the proportion of ancestors who lived in Europe a few dozen generations earlier.10 The highest proportion, around 18 percent, is found in the inland state capital, Columbia, a percentage at the low end of the range of cities in other U.S. states. They estimated about 12 percent European ancestry along the South Carolina coast, including in the slave port of Charleston, which they thought might reflect waves of slave importation keeping the African ancestry high. They estimated the lowest proportion of European ancestry, around 4 percent, on the Sea Islands off the coast, reflecting the history of isolation of the slaves who settled there, an isolation attested to by the fact that the Sea Islanders are the only African Americans still speaking a language, Gullah, with an African-derived grammar. Comparison of Y-chromosome and mitochondrial DNA types that are highly different in frequency between African Americans and Europeans also shows that by far the majority of the European ancestry in these populations comes from males, the result of social inequality in which mixed-race couplings were primarily between free males and female slaves.11

  The patterns in South Carolina are a microcosm of those in the United States as a whole. Katarzyna Bryc, at the personal ancestry testing company 23andMe, worked with me to analyze more than five thousand self-described African Americans in the company database, and found that the average European ancestry proportion was 27 percent in most of the genome but only 23 percent on chromosome X.12 Comparing proportions of ancestry on chromosome X and the other chromosomes can provide information about differences in male and female behavior during population mixture, because two-thirds of X chromosomes in the world are carried in females compared to only about half of all other chromosomes, so the X chromosome is relatively more influenced by female history. By computing the proportion of European male and female ancestors that would be necessary to produce the observed difference in European ancestry between chromosome X and the autosomes, Bryc was able to estimate the separate male (38 percent) and female (10 percent) proportion of European ancestors in African Americans. These numbers imply that the contribution of European American men to the genetic makeup of the present-day African American population is about four times that of European American women. When I discussed these findings with the sociologist Orlando Patterson, he pointed out that the fraction of the European ancestry in African Americans that came from males—which if different from half is called “sex bias”—must have been far greater during the time of slavery. Since the civil rights movement in the United States in the mid-twentieth century, cultural changes have caused the sex bias to reverse, with more coupling between black men and white women. If we carried out DNA studies of African American skeletons from a hundred years ago, there is every reason to expect an even greater sex bias.

  The genetic patterns suggest that the Thomas Jefferson–Sally Hemings model was replicated countless times by other couples. While this story is one we know about because it is close to us in time and involved famous people, there is every reason to think that sex bias has been central to the history of our species. The genome revolution makes it possible to measure sex bias dating to periods for which we have no records, and thus to begin to understand how inequality may have shaped humanity in deep time.

  The Genomic Signature of Inequality

  In humans, the profound biological differences that exist between the sexes mean that a single male is physically capable of having far more children than is a single female. Women carry unborn children for nine months and often nurse them for several years prior to having additional children.13 Men, meanwhile, are able to procreate while investing far less time in the bearing and early rearing of each child, a biological differe
nce whose effects are amplified by social factors such as the fact that in many societies, men are expected to spend little time with their children. So it is that, as measured by the contribution to the next generation, powerful men have the potential to have a far greater impact than powerful women, and we can see this in genetic data.

  The great variability among males in the number of offspring produced means that by searching for genomic signatures of past variability in the number of children men have had, we can obtain genetic insights into the degree of social inequality in society as a whole, and not just between males and females. An extraordinary example of this is provided by the inequality in the number of male offspring that seems to have characterized the empire established by Genghis Khan, who ruled lands stretching from China to the Caspian Sea. After his death in 1227, his successors, including several of his sons and grandsons, extended the Mongol Empire even farther—to Korea in the east, to central Europe in the west, and to Tibet in the south. The Mongols maintained rested horses at strategically spaced posts, allowing rapid communication across their more than eight-thousand-kilometer span of territory. The united Mongol Empire was short-lived—for example, the Yüan dynasty they established in China fell in 1368—but their rise to power nevertheless allowed them to leave an extraordinary genetic impact on Eurasia.14

  A 2003 study led by Christopher Tyler-Smith showed how a relatively small number of powerful males living during the Mongol period succeeded in having an outsize impact on the billions of people living in East Eurasia today.15 His study of Y chromosomes suggested that one single male who lived around the time of the Mongols left many millions of direct male-line descendants across the territory that the Mongols occupied. The evidence is that about 8 percent of males in the lands that the Mongol Empire once occupied share a characteristic Y-chromosome sequence or one differing from it by just a few mutations. Tyler-Smith and his colleagues called this a “Star Cluster” to reflect the idea of a single ancestor with many descendants, and estimated the date of the founder of this lineage to be thirteen hundred to seven hundred years ago based on the estimated rate of accumulation of mutations on the Y chromosome. The date coincides with that of Genghis Khan, suggesting that this single successful Y chromosome may have been his.