The Meaning of Race in the Post-Genome Era
President Shirley M. Tilghman
March 9, 2010
Baldwin Lecture, Presented at Princeton University (Posted as Prepared)
Some may wonder why a man who never went to college should be associated with a lecture at an institution that would almost certainly have discouraged his attendance had he applied for admission in 1942. But this is 2010 and much — albeit not enough — has changed since Mr. Baldwin famously said, "To be a Negro in this country and to be relatively conscious is to be in a rage almost all the time." Reflecting on his childhood, Mr. Baldwin noted, "I didn't know how I would use my mind, or even if I could, but that was the only thing I had to use." Well, use it he did. His novels, essays, plays, and other work reflect an intellectual power, a moral independence, and a capacity to invest the written word with force and meaning that are truly remarkable. And that, after all, is the essence of our mission as a University community: to seek and to speak the truth and to see our world, ourselves, and most importantly others with a clearer eye.
With the establishment of the Center for African American Studies at Princeton, we committed this University to studying and teaching the meaning of race and its impact on the past, the present and the future throughout the entire curriculum. We purposefully rejected a plan that would have relegated race to one or two isolated departments or programs. Only by spreading conversations about race throughout the entire University can we be sure that every student who passes through our gates will encounter and grapple with one of the most vexing issues facing this country, and indeed the rest of the world. Whether these encounters happen in a sociology class on inequality, or upon reading the novels of Toni Morrison, or by examining historical documents on the Civil War, or while participating in a jazz ensemble, or acting in a play by August Wilson, or studying African art in the museum, or upon confronting the strands of American religious thought, or in discussing global immigration policy, or even — indeed — in a genetics course on human diversity, our students need to examine and refine their own ideas about identity if they are to be fully educated men and women — true cosmopolitans, to use Anthony Appiah's word. I am deeply grateful to those who created this vision for Princeton — Anthony Appiah and the members of his committee — and to Valerie Smith and Eddie Glaude, the extraordinary leaders who have realized that vision so brilliantly, and in such a short period of time.
The Baldwin Lecture, which was inaugurated in 2006 with Anthony's lecture on the Cosmopolitanism of W.E.B. Du Bois, is very much in the spirit of the mission of the Center for African American Studies. It invites faculty members from across all the divisions of the University to reflect on "the issue of race and American culture." I am very honored to be asked to follow in the footsteps of Anthony Appiah, Leonard Barkan, Bonnie Bassler and Tony Grafton — surely four of the brightest jewels in the Princeton crown. Honored, and a little intimidated.
The topic I have chosen for my lecture is "The Meaning of Race in the Post-Genome Era." Of course, I am already off to a rocky start with that title. Just as there are deep differences of opinion about whether the election of President Obama ushered in a new post-racial era in the United States, there is disagreement among scientists as to whether the publication of the first draft sequence of the human genome in 2001 constituted a turning point in biology that is worthy of being called a new era. Without taking a stand on that particular question, I would like to explore with you whether the sequencing of the human genome, and the many studies that have followed in its wake to collect sequence information from humans across the globe, have provided any new insight into the meaning of race.
This is not the first time that scientists have been asked to weigh in on the existence and significance of racial categories. I am sorry to have to say that too often when science has been brought to bear on the issue, the outcome has not yielded enlightenment. One of the earliest scientists to opine about racial categories was the great 18th-century Swedish taxonomist Carl Linnaeus. (See slide above.) Trained as a physician and motivated by a religious conviction that there was a natural order to all things in the universe, Linneaeus is best known for his system of classification of plants and animals. He did not stop with the animal world, but offered a classification system for humans that divided them into five categories that were based on a combination of geography and skin color — Homo sapiens Africanus, Americanus, Asiaticus, Europeanus and — Monstrosus. He described Native Americans as red, choleric and combative; Africans as black, cunning and negligent; Asians as yellow, melancholic and stingy and Europeans as white, sanguine and inventive, and inclined toward tight clothing. In general he did not seem to have a high opinion of any race save his own. As for Homo sapiens Monstrosus, they included dwarfs, giants, troglodytes and lazy Patagonians.
Linnaeus' profound error was to conflate race with character, making sweeping generalizations about the traits of categories of peoples based on prejudice rather than careful observation or measurement. This unscientific leap that attached moral values and behavioral characteristics to geographical and skin color differences would persist for centuries, and survive at least in part because it hid behind the mantle of scientific credibility.
Linnaeus' classification system of humans lacked quantitative rigor, to say the least. The generations of scientists that followed attempted to bring measurement into the study of human classifications, but they, like Linnaeus, were unable to escape their biases and prejudices. (See slide above.) Phrenology, a field of inquiry founded by the Viennese physician Franz Joseph Gall in the late 18th century, argued that the brain was the organ of the mind — recall that there was no consensus on this for much of recorded history. Gall went further to suggest that the brain is composed of multiple distinct regions that house a variety of functions, such as sight, hearing and memory. Up to this point even a 21st century neuroscientist would agree with Gall. Where Gall began to go seriously wrong was in claiming that both moral and intellectual capabilities of the brain are innate, and that human characteristics such as love of one's offspring, courage, murderousness, metaphysical sensibility, witticism, poetical ability and obstinacy can be mapped to specific regions of the brain. Even more problematically, he proposed that by examining the shape, size and unevenness of the skull, one could discover how well each of these "faculties" was developed in an individual. A bump on the forehead foretold a well-developed organ of a generous person; one at the back of the head was the sure sign of a murderous personality.
Phrenology spread throughout 19th-century Europe, and traveled across the Atlantic as well, with practitioners profiting royally from skull readings of the gullible, including such luminaries as Johann Wolfgang von Goethe and the British Prime Minister David Lloyd George. Even a figure as scientifically sophisticated as Francis Galton, the founder of eugenics, consulted a phrenologist before venturing on a journey to southern Africa. His phrenologist advised him that he possessed a "sanguine temperament with considerable self-will, self-regard, and no small share of obstinacy." Presumably hearing those qualities gave him courage to set forth on his travels.
Phrenology might have passed into history as an amusing and largely harmless fad — a 19th-century relative of astrology — had it not been for its inevitable use as a tool to discriminate against the Irish in Britain, and across the Atlantic to justify the institution of slavery. Two of the most influential proponents of the American school of phrenology were the brothers Orson and Lorenzo Fowler, whose books were widely read throughout the middle of the 19th century. Based on "readings" of head shape, they concluded that: "The European race possesses a much larger endowment of the organs (here they were referring to the frontal and coronal positions of the head and brain) than other human species. Hence their intellectual and moral superiority over all other races."
In making such assertions, the phrenologists betrayed the most fundamental principle of the scientific method, as did the craniometrists who used brain size rather than skull shape and surface structure to draw sweeping conclusions about innate qualities of groups of humans. They began with their conclusion — the superiority of the European or Caucasian people — and then set about seeking data that would confirm that conclusion. As the evolutionary biologist Stephen Jay Gould demonstrated in his debunking of craniometry in his classic book "The Mismeasure of Man," the data were fudged to fit the prejudice, such as by selecting a preponderance of large skulls while making measurements about favored races and a preponderance of small ones when measuring those thought to be of lesser ability.
The next chapter in the history of applying scientific understanding to the meaning of race was the eugenics movement in 19th-century Britain. (See slide above.) Its founder, Francis Galton, was trained as a physician, but was a true polymath who became a pioneer in the use of statistics to quantify human traits. A cousin of Charles Darwin, Galton was deeply influenced by the publication of "On the Origin of the Species" and the evolutionary theory of natural selection. He became fascinated with the possibility of improving the human race by encouraging what geneticists called assortative mating — marriages between ever more fit individuals, leading over time to the elimination of the weak and infirm from the population. In his book "Hereditary Genius," published in 1869, Galton argued that just as farmers had been improving the stock of their crops and domesticated animals for millennia by selective breeding programs, it was feasible to improve the human species by applying those same principles to individuals of high social standing and demonstrated talent.
Galton's ideas regarding fitness were more about social class than race. He believed that genius and talent were hereditary traits, basing his conclusion on the observation that distinguished Victorians, such as statesmen, military commanders, scientists, poets and jurists were more likely than not to be related to one another. Although he would at times acknowledge the impact that societal and environment conditions could have on a person's success in life, he rejected the notion that the strict class system in Britain could explain the familial relatedness of the professional elite, asserting that men (and they were inevitably men) of great distinction must have had the benefit of a richer genetic inheritance. To his credit he never proposed the kinds of draconian steps that were ultimately taken up in the United States or Germany to improve the human stock. Instead he believed that the gradual improvement of the human species could be achieved by the state providing financial incentives to encourage early marriages and many offspring among people of high rank.
When the new science of eugenics crossed the Atlantic, it acquired a much nastier veneer. Eugenics was used between the First and Second World Wars to justify state-sanctioned prohibitions against reproduction on the part of the physically disabled and what were then termed the "feebleminded," as well as immigration and miscegenation laws to preserve the "purity" of the white race.
One of its earliest and most influential proponents was Charles Davenport, a biologist who founded the Cold Spring Harbor Laboratory in 1910 and established the Eugenics Records Office there. (See slide above.) Like Galton, Davenport was interested in quantifying human traits of all kinds, but unlike Galton, he had the great advantage of exposure early in his career to the re-discovery of Mendel's laws of inheritance. Unfortunately there is little evidence that he benefited from that exposure, as he consistently "deeply comingled science and social values," with "race as the touchstone," as the geneticist Maynard Olson pointed out a recent article entitled "Davenport's Dream." The second of Mendel's laws of inheritance states that genes assort independently of one another — that a trait such as skin color is not inherited in the same pattern in multi-generational family pedigrees as another trait such as height. Yet Davenport not only ignored this well-established principle in his studies, but also went one step further by claiming that complex human traits such as high intelligence or personality characteristics such as slovenliness followed simple inheritance patterns that were linked to the color of one's skin. He was blinded by his racial biases, attributing lack of fitness to low social status and what he termed poor racial origins. In defending his view that immigration policies should bar people of low hereditary history, Davenport wrote: "The idea of a melting pot belongs to a pre-Mendelian age. Now we recognize that characters are inherited as units and do not readily break up." Davenport ignored his own data which clearly showed that the very traits he abhorred — such as "imbecilic, criminalistic, insane, epileptic, alcoholic and sexually immoral" — were not inherited as simple units, displayed no simple pattern of inheritance and were greatly influenced by environmental factors. As the historian Dan Kevles writes in his very fine book on the subject, "In the Name of Eugenics," "…although [eugenics] was advanced with the authority and prestige attendant on one of America's most powerful biology directorships, it proceeded from science that, even by the standards of his own day, was usually dubious and often plain wrong."
Davenport had his scientific detractors, including Francis Galton himself and Thomas Hunt Morgan of Columbia University, the founder of modern genetics in America. Yet the influence of Davenport and his acolytes was widespread, with eugenics used as a justification for state-sanctioned sterilization of the "unfit" in over half the states in the union through much of the first half of the 20th century. Likewise, the Immigration Act of 1924, written primarily to exclude those of eastern and southern European background, was clearly inspired by eugenic considerations. President Calvin Coolidge publicly declared at the time of its passage: "America must be kept American. Biological laws show…that Nordics deteriorate when mixed with other races." Repugnance over the genocide practiced by Nazi Germany during World War II, which had deep eugenic roots, finally turned the public tide against such racially-motivated policies.
I have dwelled on this history of the scientific study of racial classifications for a reason — it highlights the perils that confront scientists when they venture into a terrain where one's objectivity, or more accurately lack of objectivity, can play a role in the pursuit of scientific meaning. What Linnaeus, Gall, Galton and Davenport have in common — and recall that several of them were considered scientific giants in their time — was deep-seated racial prejudice that biased the way in which they framed their questions, designed their studies and analyzed their data. It would be imprudent for us to think that such biases cannot creep into our thinking about race in the post-genome era.
To turn now to the current era, the idea of sequencing the human genome was first suggested in the mid-1980s. The rationale for doing so combined high purpose with pragmatism. The high purpose was often described in florid language — the genome was the "blueprint of life"; the "instruction manual for a human being." Once all those As, Gs, Cs and Ts were properly interpreted, it would reveal the parts list that makes up a human — the genes that are embedded in the DNA of our 23 pairs of chromosomes. Those genes encode the workhorses of the body: the thousands of proteins and RNAs that are the molecules of life. And the sequence would be a treasure trove of information for biomedical scientists intent upon using genetic data to find treatments and cures for human disease. On the pragmatic side, many scientists around the world were already working on the sequencing of the human genome, slowly, labor-intensively and very expensively, gene by gene, in small individual investigator-led laboratories. At the going rate in the 1980s, the human genome would have taken hundreds of years and many billions of dollars to sequence. It was argued that with a concerted effort that focused on improving technology, speeding up the process and bringing the costs down dramatically, it was conceivable that we could have the entire sequence in 10-15 years. And that is precisely what happened. With the infusion of significant federal dollars to develop better sequencing technology, and a decade spent practicing on organisms with smaller genomes like bacteria, yeast, worms and fruit flies, the sequencing of the human genome became a reality.
But whose genome would be sequenced? At the outset no one thought about the question very much, and the first DNA templates that were readied for sequencing came from a small number of graduate students at Caltech, where the DNA libraries were being prepared. When it dawned on those of us overseeing the project in the late 1990s that the identities of those students and their DNA sequences would be in the public domain, and therefore available to anyone, including their insurance companies for analysis, it was decided that those DNA libraries would have to be scrapped and reconstructed using a collection of DNAs from anonymous volunteers. Critics immediately pointed out that because the volunteers came from Buffalo, New York, where the new libraries were being prepared, they would represent only a subset of the genetic diversity in the human species, and therefore the knowledge gained would privilege one group over another. There were compelling technical reasons to sequence as few individuals as possible, but in the end the argument was resolved when it was pointed out that this was going to be the first not the last genome to be sequenced, and this has indeed proven to be the case.
This was not the first time, however, that race became a contentious issue for the human genome project. In 1991 a group of American human geneticists proposed a global effort to sample the degree of genetic diversity of all human populations — which they called the Human Diversity Project. (See slide above.) Their goal was to collect DNA samples from indigenous populations throughout the world, and to analyze them to understand the evolution of the species, to map historical migration patterns, and to begin to catalog the relationships between disease susceptibility and genetic make-up — all laudatory goals. Nevertheless, from the outset this project was met with blistering criticism, with accusations of racism, "genetic colonialism" and intellectual property theft, which took the organizers completely by surprise. As Kenneth Kidd of Johns Hopkins, one of the founding members of the project, said in Science magazine at the time, "We're not trying to exploit people; we're trying to include them. It's racist to avoid the totality of humans." The opponents were deeply offended by what they perceived as the paternalism of the scientists, and suspicious that the project was simply a new eruption of racism and Western-style capitalism at work. In the end the project was never funded, but the episode provided scientists with a glimpse of the enormous complexity that would attend the study of race in the context of the genome.
The final push to sequence the human genome was a short one, thanks to the dramatic improvements in technology during the 1990s, and on June 26, 2000 (see slide above), President Bill Clinton called a hasty press conference to announce the completion of the draft sequence by two separate groups, a private effort at Celera Genomics, led by Craig Venter, and a publicly-funded international project, represented in this picture by Francis Collins, then leading the National Human Genome Institute.
It was a euphoric moment for all who had participated in the project, but even at that moment of celebration, there was a sense that the genome could potentially open up a proverbial Pandora's Box of issues, particularly surrounding the issue of race. (See slide above.)
One of the most fundamental questions that the sequence was intended to answer is the genetic basis for the enormous variation within the human species. (See slide above.) When American Express brought these two extraordinary athletes together to take this picture, they had one goal in mind — to catch your attention as you flipped through a magazine. And I suspect the admen succeeded, for Wilt Chamberlin, arguably the greatest basketball player of all time unless you hail from Chicago, and Willy Shoemaker, certainly the greatest jockey of his day, capture a significant percentage of that variation — in height, weight, body mass index, hair color and texture, skin color, and taste in ties. Yet when a biologist looks at these two individuals, we are just as likely to see the marked similarities: the identical organization of the body plan, with the bilateral symmetry of the arms, legs, and eyes, and the asymmetric design of the heart. We see identical biochemical processes driving everything from metabolism to reproduction.
The tension between similarity and difference was also seen once we began to compare human genomes. Two human genomes chosen at random differ on average at only one in one thousand bases — in other words, at the level of the genome we are 99.9% identical to one another. That discovery was not greeted with enthusiasm in all quarters to be sure. (See slide above.) On the other hand, the human genome is composed of 3 billion base pairs, so 0.1% translates into two individuals differing by roughly 3 million bases — a non-trivial number of differences. Among the 3 million differences between Wilt and Willy are a relatively small number that are responsible for the dramatic phenotypic variation we see in the photograph. The majority are referred to as neutral changes — meaning they are thought to have no impact whatsoever on the phenotype of the individual.
Of the 0.1% of the genome that varies among individuals, we now know that 85-90% of the variation is shared among all humans, and only 10-15% define differences between populations. In other words, differences between individuals are significantly greater than differences between groups. This striking fact reflects the now well-documented history of our species, shown in this slide (above). The first humans are thought to have arisen in East Africa approximately 200,000 years ago, and to have spent the first 150,000 years on the African continent, accumulating in their genomes the majority of the variation that we observe today. When a small number of individuals migrated north out of Africa approximately 50,000 years ago, creating what is known as a genetic bottleneck, they brought with them a small subset of that variation, leaving Africans as the most diverse humans today. As the early humans moved both east and west over time, new variants arose in each population that were unique to that group. This post-Africa variation accounts for the 10-15% of the total variation between populations today.
As a consequence of the slow migration of Homo sapiens across the planet, it is possible to differentiate among humans from widely dispersed geographic regions by assessing the 10-15% of the genetic variation that is region-specific. In a study by Bamshad et al at the University of Utah (see slide above), DNAs were stripped of all ethnic and geographical identifiers, and assayed for 160 genetic markers that were known to be highly polymorphic (meaning different in at least 1 in 100 individuals) between populations. The results from each individual were mapped on a triangle, with the highest level of confidence of ancestry at the three corners of the triangle. On the upper left you can see that with just 160 markers, one can predict with great accuracy the geographic origin of today's Europeans, Africans and Asians. On the right, in contrast, you can see that the predictive power of this approach falls apart with South Indians. Why? Because South Indians fall into a continuum between Europeans and Asians. Not only were they on the path of migration east from Europe to Asia, because of their location between Europe and East Asia, but they were most likely to have interbred over long expanses of time with their neighbors to the east and west.
This result highlights the first important finding of the post-genome era that pertains to race: the degree of human variation is a continuum across the globe. Only when we compare populations that are geographically separated from one another and with whom little admixture has occurred do differences become sufficient to distinguish one group from another. The classic view of race, based on physical characteristics such as skin color and facial structure, would have placed South Asians in a distinct racial group, yet the genome analysis identifies them as a genetic amalgam. This is where the biological, as opposed to the cultural, notion of race does not hold up to close scrutiny.
The second finding is that genetic distinctions among individuals that we continue to define as members of different races based on physical and cultural characteristics are declining rapidly, as is evident when Americans of Asian, European and African descent are assayed in the same way. What is immediately apparent in the lower left panel is that for each of those groups, genetic analysis is less reliable in predicting their geographic origins than with indigenous populations, with the greatest decline occurring within African Americans. The points have generally moved away from the corners, reflecting less confidence in the assignment of the individual to a specific group, almost certainly because of inter-marriage over the last 300 years.
The third finding is that positive selection — genetic variation that is beneficial and therefore strongly selected when it arises in a population — can lead to marked changes in phenotype — appearance — without appreciable drift in genetic background. A good example of this are the Ainu people of northern Japan, whose light skin and body hair have defined them as culturally and racially distinct from the rest of the Japanese, despite the fact that they remain very closely related to other Japanese at the level of the genome.
Thus the sequencing of the human genome has revealed that the proxies we have historically used to define race, such as physical characteristics and geographical origins, are not irrelevant, and to a first approximation they are reflected in distribution of genetic variation in our species. Yet any classification system at the level of the genome needs to be much more nuanced, as the distribution of genetic variation is continuous across the globe, and admixture and positive selection make prognostications of genetic make-up based on outward physical characteristics imprecise at best.
The challenge ahead is to understand the role that the 10-15% of genetic variation that is group-specific has on human biology — solving the Wilt and Willy problem. While it is clear that individuals whose ancestors hail from different parts of the planet can be genetically distinguished from one another with high confidence, we know almost nothing about how those differences translate into differences in biologically important characteristics. The reason is simple — most of the complex human characteristics we might wish to understand, like susceptibility to common diseases such as heart disease or stroke, or traits such as athleticism, aggressiveness, and especially intelligence, are the consequence of the action of many genes acting in concert, not single ones, and the tools we need for identifying and studying those genes are proving to be very elusive.
There are scientists who foresee nothing but danger in the examination of complex human characteristics using genetic tools, and would prefer that we ban studies where one can foresee that potential for discrimination against one group or another. As a scientist I find this proposal to be unrealistic, and naïve. As Maynard Olson said in his article about Davenport, "History offers few examples in which scientists have been able to pluck selectively the bits of knowledge that would ultimately prove useful from a vast sea of ignorance…fearful science is typically bad science." Furthermore there are very good medical reasons for continuing to try to understand the consequences of genetic diversity in humans.
To give you one example: CCR5 is a protein that spans the membrane of T cells and serves as a co-receptor for the HIV virus. Variants in the gene encoding this protein have been shown to slow or even prevent the progression of HIV infection in those of European descent, but accelerate it in those of African descent. This is not a difference in the CCR5 gene itself, but rather a difference in how two different genetic backgrounds respond to the same mutation. In the past skin color would have been used as a surrogate to predict outcome; but given what I have just shown you about the imprecision of skin color, especially in the United States, as a proxy for genetic make-up, you can see why biomedical scientists might wish to use more informative genotyping to predict disease progression.
It is certain that debates within the scientific community about whether the historic descriptors of race or ethnicity are valid anymore, or indeed ever were, and whether they should be replaced with more robust information about genetic ancestry will continue. My own view is that attempts to stop the progress of science have been remarkably unsuccessful over time, and the benefits of understanding the impact of human variation outweigh the risks. But I say this with the specter in the back of my mind of someone speaking from this stage in a hundred years time, castigating the scientists of my era for their blindness to the ways in which their scientific findings were used to sustain prejudice and discrimination. How we proceed in this new era, and whether for good or ill, are now up to all of us, scientist and non-scientist alike. In this regard the Center for African American Studies is poised to play a critical role in the ongoing debate we must have about the meaning of race in the post-genome era. As a University that prides itself in being in the nation's service and the service of all nations, we need to be a leading voice in this debate, through the scholarship that we produce, the conversations that we stimulate and the future leaders whom we educate. As a community in which scientists, social scientists and humanists work and study in close proximity to one another, we can ensure that this discussion be as broad as possible. As the Reverend Dr. Martin Luther King Jr. observed, "Darkness cannot drive out darkness; only light can do that." At Princeton we are in the business of generating light.
Thank you for your attention, and thank you to my colleagues in the Center for African American Studies for the great honor of delivering the Baldwin Lecture.