[Paleopsych] TLS: Jerry A. Coyne: Legends of Linnaeus

[Premise Checker]

Jerry A. Coyne: Legends of Linnaeus
The Times Literary Supplement, 4.2.25

[This is a very good review that does not skirt the issues of the reality 
of race. It is similar to what Armand Leroi did on the Op-Ed page of the 
New York Times. Coyne's standard of proof goes way, way up, however, when 
the subject wanders into psychological differences among human races. As 
equality-inequality gets replaced with pluralism-universalism as the 
prinicpal left-right political axis during the 21st century, this 
jacking up the standard of proof will diminish. I'm writing a meme, called 
"Racial (Not Racist) Science," in which I'll be arguing that raced-based 
education will start emerging five years after the Republicans have left.]

     When "Europeans were governed by laws, Asians by opinions and Africans
     by caprice"

     RACE. The reality of human differences. By Vincent Sarich and Frank
     Miele. 287pp. Basic Books. £19.99. - 0 8133 4086 1

     When I apply to a United States government agency for a research
     grant, I'm asked to tick a box specifying whether I'm American Indian
     / Alaska Native, Asian, Black /African American, Native Hawaiian /
     Other Pacific Islander, or White. While there are people whose mixed
     ancestry puts them somewhere outside these boxes, in general one's
     race - even as categorized on government forms - is self-evident.

     Nevertheless, immediately below the boxes there's a disclaimer: "The
     categories in this classification are social-political constructs and
     should not be interpreted as being anthropological in nature". On the
     basis of biological indicators such as ancestry or skin colour, I've
     specified my race, but now I'm told that all I've done is specified
     membership in a particular "social-political construct".

     Presumably lost on the bureaucrats who create such forms is the irony
     that America's geneticists - experts on the biological basis of
     differences among groups - are being assured that the self-assessment
     they make on biological grounds is, in fact, not biological. This
     contradiction captures the tension beneath much of the current debate
     on race. It also highlights the peculiar American slant on this
     debate. While my colleagues in the United Kingdom must also specify
     their ancestry when applying for government funds, they are not
     subjected to double-speak disclaimers. Racism in the US, courtesy of a
     society built largely on slavery, has an especially long and sordid
     history, and remains a divisive hot-button issue. In many discussions
     of American social policy, racism is the proverbial elephant in the
     living room: recognized but carefully ignored.

     Is race a real biological phenomenon or merely a social-political
     construct an illusion about nature that is ungrounded in genetic
     reality and based solely on our need to categorize? While these two
     views seem mutually exclusive, they are simultaneously embraced by a
     popular liberal stance in the US. Many who favour affirmative action
     for disadvantaged minorities deny that race is biologically real, even
     though affirmative action involves preferential treatment of
     individuals belonging to specified biological groups. Ultimately, the
     reality or non-reality of race is an empirical biological question.
     But biologists - who are quite happy to study racial variation in
     other species - become anxious when they forsake squirrels and
     salamanders and turn instead to humans. Biologists are just as
     squeamish about discussions of human race as the rest of us.

     Perhaps more so: scientists are sensitive to accusations that bringing
     the authority of their field to bear on race can stigmatize minorities
     or legitimize racism.

     Despite its practitioners' ambivalence, the modern science of human
     genetics has accumulated a great deal of data on race. What, then, do
     these data tell us? As a recent issue of the prestigious journal,
     Nature Genetics, devoted to human race reveals, the story is
     apparently still murky. While one paper asserts that "traditional
     'racial' designations in humans are not bounded discrete categories
     but are fluid, socially defined constructs", another claims, "The
     emerging picture is that populations do, generally, cluster by broad
     geographic regions that correspond with common racial classifications
     (Africa, Europe, Asia, Oceania, Americas)".

     Some aspects of race are not controversial: black couples produce
     black children and white couples white children, regardless of where
     they live. But do these readily visible indicators of race - skin
     colour, hair type, etc - serve as markers for more profound
     differences, including behaviour and intelligence? This is the key
     question. With the human genome project completed, and human genetics,
     fuelled by new data, steaming ahead, science can surely now address
     (and, one hopes, answer) this question.

     To biologists, races are sets of populations within a species that are
     both geographically separated and differ genetically in at least one
     trait. Sometimes, depending merely on whether a biologist is a
     taxonomic lumper or splitter, a race gets its own name, and is
     designated a "subspecies".

     There are plenty of animal races, including mouse populations that
     differ only in coat colour or number of chromosomes, and sparrow
     populations that differ in size and song.

     Using this standard biological definition, Homo sapiens clearly has
     races, differing in constellations of traits such as skin and eye
     colour, and colour and form of the hair. Nobody has trouble
     distinguishing a Japanese from a Finn, even though recent mixing of
     humans has produced individuals who are intermediate and
     unclassifiable. The golfer Tiger Woods, for example, is often
     identified as black, but most of his genes are from East Asian
     ancestors.

     Well, if races exist, how many are there?

     We can't answer this question because human physical variation occurs
     in nested groups, and where one draws the lines is arbitrary. In his
     notorious book The Origin of Races (1962), the anthropologist Carlton
     Coon opted for five: Caucasoid (European and Middle Eastern), Capoid
     (Bushmen), Congoid (black African), Australoid (Australian aborigines)
     and Mongoloid (Asians). But these groups can themselves be subdivided:
     Mbuti Pygmies, for example, could be considered a race on the basis of
     height, and Swedes clearly differ from Bedouins, though both are
     "Caucasoid".

     What happens when we forsake skin colour, nose shape and hair type,
     and look instead at less visible forms of genetic variation -
     differences at the DNA or protein level - in standard racial groupings
     (Coonian or otherwise)? Remarkably, virtually all the variation
     uncovered using various high-tech molecular techniques fails to
     correlate, or correlates only weakly, with the classical physical
     variation that determines race.

     The first of these studies was done using the relatively crude
     techniques available in 1972. The evolutionist R. C. Lewontin
     calculated that 85 per cent of the genetic variation in our species
     occurs among individuals within an average population, another 8 per
     cent among populations within a race, and only 6 per cent among
     morphologically defined races. More recent DNA studies have confirmed
     this result. Reviewing these data, the geneticist Luca Cavalli-Sforza
     and his colleagues note: "The differences in human groups, even very
     distant ones and no matter whether the groups are defined on a racial
     or on a geographic basis, represent only a small fraction of the
     global genetic diversity of our species".

     The discrepancy between physical "racial" traits and the variation in
     most genes can be seen by measuring skin pigmentation in humans. Here
     the result is reversed: 88 per cent of variation in skin colour occurs
     among races. Clearly there is a difference between the geographical
     distributions of variation in genes coding for the physical traits
     that distinguish human groups, and of variation in the genes that do
     other things.

     The dichotomy between DNA and morpho-logy, however, is completely
     understand- able given the evolutionary history of Homo sapiens. The
     key aspect of human races is that they are young. Although our lineage
     split off from that of chimpanzees about 6 million years ago, all
     modern humans descend from a relatively small group that spread out of
     Africa, beginning about 60,000 years ago. It's an evolutionary fact of
     life that large differences between groups can arise only when they
     have, at most, limited contact with each other, and thus are not
     capable of exchanging genes. There has thus been only a very short
     time for racial differences to have arisen between human isolates:
     differences must have necessarily arisen in 60,000 years, only 1 per
     cent of the interval since we diverged from apes. This is roughly
     3,000 generations, a mere eye-blink in evolutionary time.

     The disparity between "racial" and overall genetic variation tells us
     two profoundly important things. First, the genes for physical
     differences are not representative of the genome as a whole; that is,
     the degree to which skin colour differs between two people is not a
     good indicator of their overall genetic difference. Second, no matter
     how one demarcates human races, the total genetic difference among
     them is trivial.

     So does this mean that we can ignore human race? No. These conclusions
     do not mean that races do not exist, nor that their differences are
     uninteresting. Racial differences can tell us about the evolutionary
     factors that have driven differentiation. Sickle-cell anaemia, for
     example, is most common in blacks whose ancestors came from Equatorial
     Africa. Because carriers of the sickle-cell mutation have some
     resistance to falciparum malaria, it is likely that the high frequency
     of this mutation in African and African-derived populations resulted
     from natural selection in response to malaria in tropical Africa.
     Ancestry in this and other cases can also help diagnose genetically
     based disease.

     What about the major "racial" traits: skin colour, hair type, etc? Why
     are these differentiated geographically? Skin colour is the easiest to
     explain in terms of natural selection. The darker skin of tropical
     groups probably protects them from intense ultraviolet light and
     lethal melanomas, while the pale skin of higher- latitude groups
     allows penetration of the skin by light necessary for the synthesis of
     essential vitamin D.

     Yet many differences between groups are not so easily explained as
     adaptations to the environment. How, for example, can we account for
     the eye folds of Asians, or the longer noses of Caucasians? Here,
     "sexual selection" is a possibility. In this process, a trait's
     advantage comes not from enhancing one's ability to survive (the
     purview of natural selection), but from enhancing one's access to
     members of the opposite sex. A famous example is the elaborate tail of
     the male peacock, a species in which females prefer males with more
     ornate tails. Perhaps the big noses of whites, and curly hair of
     blacks, are human versions of the peacock's tail. Darwin (who
     conceived the idea of sexual selection) was the first to suggest that
     such selection could explain the differences among human groups, based
     on different preferences in different places.

     Sexual selection operates most effectively on observable traits. The
     peacock's tail, or its equivalent, must be readily visible. If sexual
     selection played a powerful role in human racial differentiation, we
     would expect those genes that affect appearance to be more prone to
     racial differentiation than genes that do not. And that, you'll
     recall, is exactly what geneticists have found. Moreover, sexual
     selection is an attractive explanation for "racial" features for two
     other reasons.

     First, through language and other vehicles of culture, humans have a
     remarkable ability to disseminate ideas and opinions. Such cultural
     evolution can occur much faster than genetic evolution: an idea can
     spread through an entire population in a matter of hours.

     (When I was young, the rumour that Paul McCartney's supposed death was
     encrypted in a Beatles song swept the US within a single day.) A new
     mutation, however, can take many generations to spread through a
     population, even if it is strongly favoured by natural selection. Now
     imagine that the idea or fad in question involves the preferred
     appearance of one's mate. For example, an empress in China might
     prefer men with straight black hair and almond-shaped eyes. By
     creating a fashion, her preference spreads culturally to all her
     female subjects and, lo and behold, in the next generation the
     curly-haired and round-eyed individuals will be largely replaced by
     individuals with straight black hair and almond-shaped eyes. It is
     this interaction between cultural transmission and genetic evolution
     that makes the idea of sexual selection especially appealing.

     Second, sexual selection can in principle act incredibly fast (its
     rate depends on the strength of the preference and the rate of its
     cultural spread), making it an ideal candidate for effecting the rapid
     evolutionary differentiation of physical traits that has occurred
     since our migration from Africa.

     However, the most controversial aspects of race centre not on DNA or
     hair texture, but on behaviour. Many assume that physical indicators
     of race are mirrored by equally striking genetic differences in
     temperament and intelligence. On the other hand, many scientists feel
     that, given the results of Lewontin and others, the population
     genetics of our species militates against this possibility. Because
     most of the genetic variation in our species does not correlate with
     race, we would not expect a priori that particular traits would be
     differentially distributed among the races. However, it is still
     possible that some variation that matters in our society - say,
     variation underpinning differences in intelligence - does in fact lie
     within that small proportion of variation that correlates with race.
     We do not know the genes underlying such traits, and therefore it is
     impossible to resolve the issue of whether there is a genetic
     correlation of intelligence or personality with race. Thus some have
     used an alternative (and far less satisfactory) method: ignore
     genetics, measure the traits themselves in individuals from different
     groups, and ask whether such variation correlates with race.

     This practice has a famously ugly history. It provides a convenient
     way, given a little bad science, to confirm one's prejudices. In his
     eighteenth-century classification of humans, the botanist Carolus
     Linnaeus noted that Europeans are "governed by laws", Asians "governed
     by opinions" and Africans "governed by caprice". Happily, these
     attitudes have changed, but claims that resurrect the spirit, if not
     the letter, of Linnaeus still surface regularly. The Bell Curve (1994)
     by Richard J.

     Herrnstein and Charles Murray, for example, argued that the higher IQ
     of whites than of blacks rested on genetic differences.

     Clearly, the time is ripe for an objective, scientifically informed
     book on human variation. Regrettably, Race: The reality of human
     differences, by Vincent Sarich and Frank Miele, is not that book. It
     is not that the authors are unqualified to write such a book. Sarich
     is a well-respected evolutionary geneticist and Miele is the senior
     editor of Skeptic magazine. Some of what they say about the biology of
     race is thoughtful and provocative. Unfortunately, this topic occupies
     only a third of the book. The rest is a confusing melange: a canned
     history of anthropological controversies, a tedious discussion of how
     to date human evolution using variation in molecules (a method
     pioneered by Sarich), and a dubious argument that an evolutionary view
     of race makes affirmative action impractical. Sarich and Miele
     eventually take the bull by the horns, but ultimately fail to make
     their case that races show substantial, genetically based differences
     in behaviour, temperament and intelligence.

     Reviewing the data, Sarich and Miele conclude that races are indeed
     genetically differentiated in many aspects of temperament and
     intelligence. They rely on four lines of evidence, but each of these
     is flawed:

     The argument from dogs. Sarich and Miele compare human races to dog
     breeds. Both show large differences in traits but small differences in
     DNA. And we all know that dog breeds have different temperaments:
     sheepdogs are obedient, chihuahuas irascible, and terriers pugnacious.
     In what they call the "canine comparison", Sarich and Miele argue that
     because dog breeds, like human races, arose by selection (although in
     dogs this selection was imposed by humans), one can draw conclusions
     about humans from dogs. Ergo, we expect large differences of
     temperament between races. But this argument is absurd. Many breeds of
     dogs were deliberately selected for behaviours useful to humans -
     sheepdogs to follow the shepherd's commands - but there is no evidence
     that human races have undergone selection for different temperaments.

     Differences in athletic performance. Sarich and Miele quote
     extensively from Jon Entine's Taboo (2000), which maintains that
     genetic differences between races explain differences in athletic
     abilities. On the basis of fragmentary evidence, Entine claims that
     blacks are superior runners because of special musculature, physiology
     and bone structure. It is clear that blacks dominate other groups in
     both long-and short-distance running, and it is equally clear that
     different African groups excel in different areas: East Africans in
     distance-running, and West Africans in sprinting. But even if Entine
     is right, he fails to explain the difference in athletic ability among
     populations of Africans. Moreover, it is arguable whether athletic
     ability is a socially important behaviour.

     Behaviour of newborn infants. To support inborn racial differences in
     temperament, Sarich and Miele cite the work of the American
     psychologist Daniel Freedman, who studied the behaviour of newborn
     infants. Less than forty-eight hours old, these babies had little or
     no contact with their mothers, so any differences in behaviour were
     presumed to be genetic. Indeed, Freedman found significant differences
     among races. Chinese and American Indian newborns, for example, were
     more placid and easier to calm than whites, while black newborns had
     superior head and muscle control. But even if we ignore the potential
     influence of prenatal environment on these differences, their social
     implications are uncertain. The different "behaviour" of black
     infants, for example, probably reflects an accelerated rate of
     development, which leads to stronger muscles. While Freedman's
     findings are provocative, their relevance to adult behaviour is
     completely unclear.

     Differences in IQ. Sarich and Miele make much of the thirty-point
     difference in IQ between whites and sub-Saharan blacks. Using their
     "evolutionary default hypothesis" - if races have diverged in physical
     traits, they must also have diverged in other traits - they imply that
     much of this difference is genetic. But there is no evidence for this
     supposition, nor does it follow that all traits must follow the
     pattern of pigmentation, as shown by the similarity of DNA among
     races.

     Moreover, differences between groups can be based on culture or
     environment rather than genetics, or a combination of these. Since the
     end of the Second World War, for example, the Japanese on average have
     increased in height nearly 5 inches, and are predicted to outgrow
     Americans within a decade. This gain of height, however, was caused
     not by evolution, but by improved standards of living.

     It is remarkable how little we really know about the genetics of human
     behaviour.

     But what is more disturbing than the weakness of these data is Sarich
     and Miele's willingness to extract from them such strong conclusions.

     Scientists should exercise special care when dealing with issues of
     social importance.

     While pigmentation may have diverged due to selection based on
     sunlight or temperature, and facial features by sexual selection, what
     do we expect for behaviour and intelligence?

     Is there any reason to think, for instance, that Caucasians
     experienced especially strong selection for high intelligence, or
     Asians for placid behaviour? One can make up ex post facto stories,
     but they are just that -stories. One can just as easily claim that
     high intelligence would be adaptive in all groups. Donald E. Brown's
     book Human Universals (1991) lists numerous traits found in all
     societies, including crying, gossip, jokes, music-making, fear of
     snakes and a preference for sweets. It is the unjustified assumption
     that physical differences are invariably accompanied by evolved
     behavioural ones that turns race from a biological reality into a
     flawed social construct.

     Sarich and Miele devote their final chapter to the social implications
     of race. They weigh three alternative societies: a meritocracy, in
     which race is ignored; affirmative action, in which members of some
     races are given preference; and "resegregation", the cultural
     separation of races. Unsurprisingly, they defend the meritocracy, as
     "it most clearly corresponds with what both of us were taught was 'the
     American way'", and "it has produced the greatest economic gain for
     the greatest number of people in the history of the world". Their
     evolutionary justification, of course, is that races are genetically
     unequal in different ways, and the meritocracy is best served by
     allowing the free expression of these inequalities: "Imposing equality
     requires the use of government force, thereby reducing individual
     freedom . . . it also hamstrings individual initiative and the
     intellectual and economic growth that comes with it". Others, of
     course, would disagree, noting that societies can also be judged by
     how humanely they treat the least advantaged. Old-age pensions,
     welfare and free medical care might be considered more important than
     new cars and iPods.

     Responding to Stephen Jay Gould's assertion that "human equality is a
     contingent fact of history", the authors argue: "No, we say. If
     there's one thing the evolutionary process cannot produce, it is
     equality". Sarich and Miele are wrong: some universal human traits are
     produced by the evolutionary process and are by definition "equal" in
     different people. The kind of divergent evolution Sarich and Miele see
     as rampant in humans typically takes time, great spans of it. The
     critical fact here is that the evolutionary youthfulness of human
     races has simply not allowed enough time for wholesale genetic
     differentiation among groups. Yes, certain traits - skin colour, hair
     type, etc - have diverged courtesy of natural and sexual selection,
     but these are special cases. Throughout the entire genome and across
     the entire planet, then, we see evidence supporting Gould's
     contention: each group - race - embraces virtually the same pool of
     genetic variation. Unlike some of my colleagues, who condemn any
     exploration of the genetics of race as pernicious, I applaud Vincent
     Sarich and Frank Miele for trying to bring rigour to a famously
     unrigorous area. Unfortunately, that rigour is sadly lacking when they
     try to buttress their claim that significant biological differences
     exist among human groups.