The Western concept of race must be understood as a classification system that emerged from, and in support of, European colonialism, oppression, and discrimination. It thus does not have its roots in biological reality, but in policies of discrimination. Because of that, over the last five centuries, race has become a social reality that structures societies.9
For Jared Diamond, author of the bestselling Guns, Germs, and Steel, “[t]he reality of human races is another commonsense ‘truth’ destined to follow the flat Earth into oblivion.”10
It is in this context that Part II sets out to convince you that the orthodoxy about race is scientifically obsolete.
6
A Framework for Thinking About Race Differences
Of necessity, Part II is organized radically differently from Part I. In Part I, I could assume that most readers came to the discussion accepting that the sexes have major genetic differences with regard to sexual function and reproduction, and thus are open to the idea that those known genetic differences could spill over into effects on personality, abilities, and social behavior. My task was first to describe some phenotypic differences of interest, then show how genetic and neuroscientific findings are linking up with phenotypic differences.
MY MODAL READER
Readers of Human Diversity presumably span the range in their opinions on these issues, but I’m writing for what I conceive to be the statistically modal type: reasonably open-minded but also accepting the intellectual received wisdom. My sense of that received wisdom regarding sex differences is that a strict view of “gender is a social construct” is seen as too extreme—it is accepted that gender is largely a social construct, but not entirely so. My experience with race has been different. The intellectual received wisdom seems to be that significant racial differences in cognitive repertoires are known to be scientifically impossible.
In the case of race, there is no equivalent to the Y chromosome and no equivalent reason to assume that significant racial genetic differences are plausible. That being the case, it would be pointless for me to begin with evidence about phenotypic differences in cognitive repertoires across races in the same way that I presented them for the sexes. The logical reaction from many readers would be, “So what? They’re not genetic.”[1]
My goal in Part II is to get past the first hurdle in thinking about race differences: to lay out the evidence that it is evolutionarily reasonable to expect that phenotypic differences among races in cognitive repertoires could be at least partly genetic and that expanding knowledge about genetic variants supports that expectation. I also want to convey that this is not some new, fringe position, but the result of accumulating knowledge about genes and race that goes back almost 30 years. That’s why this presentation takes a historical approach. Recall the analogy with archaeological digs that I offered in the introduction. When it comes to race differences, science has identified a promising site, mapped it, and has a plan for next steps, but the actual excavation of the site is in its early stages. You are going to come away from the discussion in Part II with (I hope) an appreciation of where things stand and curiosity about what comes next—curiosity, not dread. The most likely scenario is that we will find many interesting but usually small distinctions.
My confidence that such distinctions will be found is based on three developments over the last 30 years, concentrated in the years since the genome was sequenced:
It was discovered that human populations are genetically distinctive in ways that correspond to self-identified race and ethnicity.
Advances in the ability to date evolutionary changes have revealed that evolutionary selection pressure since humans left Africa has been extensive and mostly local to the different continents.
Raw race differences in genetic material related to cognitive repertoires are common, not exceptional.
Each of these developments has its own chapter.
What the Orthodoxy Gets Right
As we set out, let me specify what the orthodoxy gets right. Franz Boas and Ashley Montagu were right to say that many nineteenth-century conceptions of race were caricatures divorced from biological reality. Richard Lewontin was right that race differences account for only a small fraction of the biological variation existing among humans. Stephen Jay Gould was right to reject the once widely held belief that humans evolved independently in Europe, Asia, and Africa for hundreds of thousands of years. The orthodoxy is not wrong altogether but goes too far when it concludes that race is biologically meaningless. We have before us an exercise in modifying our understanding of race, not resurrecting nineteenth-century conceptions.
The orthodoxy is also right in wanting to discard the word race. It’s not just the politically correct who believe that. For example, I have found nothing in the genetics technical literature during the last few decades that uses race except within quotation marks. The reasons are legitimate, not political, and they are both historical and scientific.
Historically, it is incontestably true that the word race has been freighted with cultural baggage that has nothing to do with biological differences. The word carries with it the legacy of nineteenth-century scientific racism combined with Europe’s colonialism and America’s history of slavery and its aftermath.
Scientifically, it is an error to think of races as primordial. Part of the story I will tell describes the repeated cycles of mixing, isolation, and remixing that have gone on among the populations that left Africa. Such cycles have also gone on within the populations that remained in Africa, not to mention remixing by populations that revisited Africa. As you will see, the number of groups into which people can be sorted genetically is fluid and depends on how much genetic information is brought to bear on the sorting.
The combination of historical and scientific reasons makes a compelling case that the word race has outlived its usefulness when discussing genetics. That’s why I adopt contemporary practice in the technical literature, which uses ancestral population or simply population instead of race or ethnicity, throughout the rest of Part II.
Third Interlude: Genetic Terms You Need to Know to Read the Rest of the Book
I cannot discuss any of the propositions of Part II without using technical terms regarding genetics. Hence it’s time for another interlude, a refresher course in this area of biology.
Genome. A genome consists of two strands of DNA (deoxyribonucleic acid), intertwined in the famous double helix and found in the nucleus of almost every cell of every organism. In humans, each strand of DNA consists of a string of more than three billion occurrences of one of four chemicals: adenine, thymine, guanine, and cytosine, usually referred to by their first letters, A, T, G, or C. Each occurrence in each strand is lightly linked by hydrogen bonds to a corresponding occurrence in the other strand. They are called base pairs. Stretched out, the three billion base pairs in the nucleus of just a single cell would be about six feet long. But the strands are curled up into a space about six microns across—six millionths of a meter.
Site is a synonym I will use for base pair. In the technical literature, locus is often used for this purpose, but with the advent of genome-wide association studies, locus also is often used to designate a larger stretch of the genome.2 I always use site to refer to a single base pair and locus when both sites and regions are involved.
WHAT DOES “SEQUENCING OF THE HUMAN GENOME” MEAN?
DNA, like some other complex molecules, has a specific physical sequence of bits. DNA has a sequence of nucleotides arranged in base pairs. To sequence a segment of DNA is to determine the base pair residing at each address (which also requires that you have an exact address for the beginning and end of the segment relative to the rest of the genome). It is like a map of New York’s Broadway showing every building and its unique address, but not what any of the buildings are used for. Sometimes you will read about “mapping” or “decoding” the genome. They are probably referring to the same thing as “sequencing.”
In the early 1970s,
British biochemist Frederick Sanger, who had already won a Nobel Prize for sequencing insulin, joined another British biochemist, Alan Coulson, to invent a method for sequencing up to 80 nucleotides at a time, far more than anyone had been able to do previously. In 1977, they published an improved, faster version that was the foundation for subsequent generations of sequencing technology.3
During the 1980s and 1990s, limited segments of the human genome were sequenced, but it was left to the Human Genome Project, begun in 1990, to undertake the huge collaborative effort needed to sequence and stitch together the identities of the three billion base pairs making up the human genome.
Drafts of the complete genome were released starting in June 2000. On April 14, 2003, the National Human Genome Research Institute announced that the Human Genome Project had been successfully completed and published the full sequence.4 In the rest of this book, I will sometimes refer to 2003 as the date for the sequencing of the genome, with the understanding that work on the human genome had already begun using the earlier drafts.
Chromosome. All the base pairs are contained in chromosomes. A chromosome is a long chunk of DNA, with the helix usually packed in a tight structure. Humans have 46 chromosomes arranged in 23 pairs, one of which is the pair of sex chromosomes that figured so prominently in the discussion of sex differences. Under a powerful microscope, chromosomes can be visually differentiated from one another, and they are also functionally differentiated. The adjective autosomal refers to the 44 nonsex chromosomes, their genes, and SNPs.
SNP. Now we come to a term that you will encounter many times: SNP, pronounced “snip,” the term for sites that consist of different pairs of letters in different people, and thereby are the major source of variation among human beings. A SNP is one of several kinds of genetic variants. The letters stand for single nucleotide polymorphism. To qualify as a SNP, a given combination at a given site must occur in at least 1 percent of the genomes of whatever species is being studied.
SNV. This term, which is unpronounceable, stands for single nucleotide variant. It includes sites that show variation without any implication for frequency (all SNPs are also SNVs, while a site with a minor allele frequency of less than .01 is a SNV but is not a SNP). I will sometimes use SNP loosely to refer to all nucleotide variants when I think that switching back and forth between SNP and SNV would be unnecessarily confusing.
Allele. The letters involved in a SNP are called alleles, emphasis on the second syllable. A large majority of SNPs have only two alleles. Such SNPs are biallelic. It is customary to call the allele with the lower frequency the minor allele, and to use the frequency of the minor allele as the default when talking about a given site.
Microsatellite. SNPs are not the only bits of genetic material that can show variation among individuals. Sometimes, tandem base pairs are repeated (e.g., CACACACA), but with varying numbers of repeats in different individuals. Unlike SNPs, which usually have only two variations, the number of tandem repeats, sometimes abbreviated VNTR (variable number of tandem repeats), goes from two to dozens. A VNTR is called a microsatellite if it involves no more than five tandem base pairs. A larger VNTR is called a minisatellite.
Genotype. The genotype is an organism’s genetic material with regard to a given trait. Writ large, it refers to all of the loci that have a causal effect on the trait. Writ small, it refers to the combination of alleles present on the pair of chromosomes at a given site.5 In a biallelic site, there can be three such combinations, which are generically denoted as AA, Aa, or aa (respectively indicating two copies of one of the DNA letters, one of each, and two of the other DNA letter).
Genetic marker. A genetic marker is usually a SNP or a microsatellite. Genetic markers are useful because they take different forms—different alleles in the case of SNPs, different numbers of repeats for microsatellites—but they typically do not have a known function. On the contrary, analysts often take pains to select genetic markers that are thought to be nonfunctional. The utility of a genetic marker is that it is neutral with regard to natural selection, enabling the researcher to analyze patterns independently of natural selection’s confounding effects.
Noncoding. Sites can do a number of things, but a majority of the three billion sites don’t seem to have any effect on anything. They used to be called “junk DNA,” but that term is falling out of use as subtle functions are discovered, especially involving gene expression. Coding is used to describe a site that is part of a region encoding a protein. Noncoding refers to all other sites. Some noncoding sites are located in regions with important regulatory functions, but many of them still have no known function.
Gene. That leaves the word you have known since elementary school, gene. Its meaning has gotten more complicated as geneticists have learned how DNA works. If you are in your fifties or older, you probably still think of gene in the Mendelian sense: the genetic unit that determined a trait, with dominant and recessive versions—the gene for eye color, for example. In modern genetics, the term refers to a contiguous region spanning many sites whose expression or transcription leads to the production of something useful—usually a protein. The number of protein-coding genes is around 20,000.
A mistake to be avoided: People who grew up with Mendelian genetics often still talk about “having a gene for” some trait—a gene for intelligence, for example. It is now known that extremely few traits are determined by a single gene (even using the word in its modern sense). Variation on a relatively simple trait can be determined by hundreds or thousands of SNPs located in many genes. To use the technical term, almost all heritable traits are polygenic.
Genetic drift. You are probably already aware that the mother and father each contribute about 50 percent of their genes to their child. Which parent contributes which genes is mostly a matter of figurative coin flips.6 Suppose (I’m simplifying for the illustration) that for a given SNP with two alleles, A and a, 80 percent of each sex carry the AA combination and 10 percent carry the aa combination. Given random mating, 64 percent (.8 × .8) of the couples who produce the children both carry the AA combination, and they are guaranteed to pass on the AA combination to their offspring. Another 4 percent (.2 × .2) will both carry aa and are guaranteed to pass on the aa combination. The other 32 percent of the couples carry at least one copy of the a allele between them. Whether one or both of the offspring’s chromosomes carry the a allele is a matter of 50/50 coin flips. For example, suppose that the father is Aa and the mother is AA. The mother will certainly contribute an A allele. The odds that the father will contribute an A allele are just 50 percent. It makes no difference that 80 percent of the population carries the A allele; in the process of meiosis, when one chromosome from each parent is contributed to the offspring, it’s a coin flip.
Repeated coin flips can produce odd results, so it is unlikely that the next generation has exactly 80 percent of AA combinations. It could be somewhat higher or somewhat lower. The percentage drifts, for purely statistical reasons. The smaller the population, the greater the expected drift. If you flip a fair coin a million times, the percentage of heads is going to be almost exactly 50 percent. Flip it ten times, and it will often be far from 50 percent. Now suppose that for several generations in a row, the coin flips happen to increase the incidence of allele A. Given small populations and enough generations, sometimes the percentage will hit 100—allele A will “go to fixation” or become “fixed.” It hasn’t been a process of natural selection, but of coin flips. Nonetheless, whatever effect allele A has is now universal in this particular population. And, by necessity, whatever effect allele a may have had is also gone from the population.
Enough background. It’s time to recount how human ancestral populations can be defined not by the color of their skin or any other visible characteristic, but by the profiles of their DNA.
7
Genetic Distinctiveness Among Ancestral Populations
Proposition #5: Human populations are genetically distinctive in ways that co
rrespond to self-identified race and ethnicity.
A good place to start is by understanding a simple truth that was predicted theoretically many decades ago and has since been validated by empirical evidence: Any human population becomes genetically distinctive by the mere fact of separation from others who share the same ancestry.
Suppose a few centuries from now humans invent the warp drive and a coalition of countries—East Asian countries, let’s say—launches a ship with 100 humans to colonize another planet. For mysterious reasons, the coalition does not screen the colonists for qualities like intelligence, sociability, physical attractiveness, levelheadedness, or fertility. Instead, it employs a sophisticated algorithm that randomly picks 100 East Asians.[1]
Even as the door of the spaceship closes on the 100, they will already be genetically distinctive from the East Asians who remain on Earth even though the random selection procedure was perfect. The reason is that the 100 colonists will carry with them only a portion of the genetic diversity among all East Asians.
In the database for Phase 3 of the 1000 Genomes Project, the East Asian sample has about 19,257,000 sites with two alleles.2 Almost two-thirds of them have a minor allele frequency among East Asians of less than .01.3 It can be expected that most of those rare variants will not be carried by even one of the 100 colonists. They’re gone forever in that subgroup of East Asians (unless, by a fantastic coincidence, they reappear through mutations).
Human Diversity Page 17
