Pharyngula

That question of race

John Wilkins has an excellent linky post on the subject of race. My position on the issue is Richard Lewontin’s (seen here in a RealAudio lecture by Richard Lewontin), and more succinctly stated by Wilkins:

So, do I think there are races in biology as well as culture? No. Nothing I have seen indicates that humans nicely group into distinct populations of less than the 54 found by Feldman’s group (probably a lot more – for instance, Papua New Guinea is not represented in their sample set). And this leads us to the paper by the Human Race and Ethnicity Working Group (rare to see a paper that doesn’t list all the authors). They rightly observe that while there are continental differences in genetics, there is no hard division, and genetic variation doesn’t match up with cultural differences per se. There is a genetic substructure to the human population, but it isn’t racial.

Comments

  1. #1 thwaite
    April 22, 2006

    I predict there will be incoherence in this discussion using the word ‘race’ given its several definitions in Wilkins’ article, ranging from Blumenbach (old bad value-laden essentialism) to contemporary multi-allelic genetic clusters.

    And yet peoples everywhere recognize others easily as exemplars of one race or another, sometimes to high levels of specificity. Why? A clue arises in that the great majority of racial differences are literally only skin-deep, affecting appearances as “markers” – just the sort of traits sexual selection can favor. Darwin (no racist) proposed that human racial variation is due to sexual selection, and this was his main goal in his book DESCENT OF MAN AND SELECTION IN RELATION TO SEX although he also surveyed sexual selection generally. This proposal has had its ups and downs historically (as has sexual selection generally). It’s discussed by Jared Diamond, briefly and clearly in his 1992 book THE THIRD CHIMPANZEE (recommended). I posted an excerpt on my campus web site:
    “Sexual Selection and Origin of Human Races”
    http://online.sfsu.edu/~thwaite/Darwin/subSexSelRacesDiamond.html

    So yes race is quite visible – and we like what we see so much that we prefer to mate with it. This perpetuates and can amplify both the visible traits, and the preference for them.

  2. #2 Miguel
    April 22, 2006

    represent each individual’s genome as a point in a space of extremely high dimension, and define a race as a set of points whose distance from each other is less than some radius.

    I don’t know much about this stuff, but I wonder if the Curse of Dimensionality (high dimension = poor discrimination between nearest and furthest neighbour) is going to make the choice for the race-defining radius problematic.

  3. #3 plunge
    April 22, 2006

    He should be more specific and clear: the genetic substrate of the human race does not match up with the common racial categories. You can’t run around calling people “black” and expect it to have much of any consistency relating to aynone’s genetic similarities or disimilarities.

  4. #4 thwaite
    April 22, 2006

    … poor discrimination between nearest and furthest neighbour

    Yeah. So much for the mating acumen of any robot using such an algorithm for it’s mate-selection choices. Nature does it different, apparently.

    It occurs to me PZ didn’t discuss the role of sexual selection (which complements natural selection) in his reimagined humanity. Its effect is less predictable – SS leads to whimsical results like peacock tails rather than obvious optimizations for survival. But its potential is suggested by the last sentence of the Diamond excerpt I linked to above:
    “If humanity survives another twenty thousand years, I predict that there will be women with naturally green hair and red eyes -and men who think such women are the sexiest.”

  5. #5 BlueIndependent
    April 22, 2006

    I don’t really know if this is news to me. I would just say, from a entirely laymen’s point of view, that this is obvious overall. The differences we see are just functions of the different environments each sub-culture evolved in.

    Or perhaps that is where I misunderstand what is being discussed?

    Also, I think a lot of time has passed, that has allowed the human genetic clusters/genes to assimilate any previous differences, say those gathered from the intermingling of neaderthal and cro-magnon species.

    I really am not familiar with any of this material scientifically, so I apologize if I sound like an idiot.

  6. #6 JP
    April 22, 2006

    This is all I need to know about the origin of races:
    http://en.wikipedia.org/wiki/Nuwaubians#The_Races_and_Their_Origins

  7. #7 Jonathan Badger
    April 22, 2006

    Darwin (no racist) proposed that human racial variation is due to sexual selection, and this was his main goal in his book DESCENT OF MAN AND SELECTION IN RELATION TO SEX

    Darwin was horribly racist, at least according to our modern standards. It’s impossible to read “Descent of Man” today without wincing on nearly every page. Yes, he was an abolitionist and therefore more enlightened than many people of his day, but he was a still a product of his times — to him the world consisted of “civilized races” and “savages” — the latter are “much given to imitation” as are apes.

    It really wasn’t Darwin’s fault that he was a racist, any more than it was Newton’s fault that he was obsessed with the dimensions of Solomon’s temple — both men were brilliant and yet somewhat trapped by the culture they were in — as we are no doubt as well.

  8. #8 James Gambrell
    April 22, 2006

    Race never was a precise word, races defined families of characteristics, not exclusive catagories. All it is is an extension of the similarities we see between families and clans to a population level.

    First of all, the whole “more variation within than between” defense completely misses the point. Race is about phenotype not genotype. Most europeans can digest milk fairly well, most asians can’t, what percent difference is that? You can’t apply percentages to phenotypic differences that don’t have a natural zero point, and you certainly can’t come up with an aggregate phenotypic difference since there are an infinite number of phenotypic attributes you could measure, and there is no way of assigning weights to any of them.

    On the other hand, its misleading to pretend there is no genetic basis for racial differences, how could their not be? You can’t say typical racial characteristics (nose shape, skin color, hair type, etc) are “cultural”. Races have been roughly defined for a long time, and now what will happen is we will match up genes with races. From what I’ve read some biologists have already done this. There are already groups offering racial-makeup tests. They will never be perfect but I think they are better than relying on self-report for affirmative action programs, for example.

    I just don’t understand why biologists like Lewontin and I guess Myers want to resist the identification of genetics with race. Feldman’s findings of 54 genetic groups doesn’t disprove race, it just shows that there is evidence for 54 or more races. The fact that some races or groups of races have been singled out historically and socially doesn’t mean that race isn’t a valid concept. You can always find races within races within races. Ancestry is useful within species so long as there are groups that substantially self-segregate, as races do in many parts of the world. If groups self-segregate because of cultural differences for a long enough period of time, you could expect a new race to develop.

    The genetic study of race is incredibly important from social science standpoint. It will help us seperate genetically-based characteristics from more culturally based ones. It will inform social problems, are they race-based, or culture based? Do blacks and whites sit apart from each other at lunch because of genetics, or culture? It will inform all kinds of medical decisions as well. I think biologists need to help us answer these questions and not play the PC card. As long as we face problems with racial dimensions, we will continue to use the word race. Sure, the word doesn’t have a technically satisfying definition, but how many words do?

  9. #9 PZ Myers
    April 22, 2006

    As Lewontin says, race is real…as a social construct. It is not real biologically. There are prominent markers that we use to recognize “race”, but they tend to be superficial and not really relevant to the substantive interactions between people.

    Those 54 groups don’t really fit the idea of race that well. They represent families of people who have some recognizable social identity. It does not mean that they have significant unique biological properties.

    I think the biological contribution here should be simple statements along the lines of, “you can mate with them, and their ancestors mated with your ancestors. We’re all human.”

  10. #10 Jonathan Badger
    April 22, 2006

    Feldman’s findings of 54 genetic groups doesn’t disprove race, it just shows that there is evidence for 54 or more races. The fact that some races or groups of races have been singled out historically and socially doesn’t mean that race isn’t a valid concept

    The problem is that genetic groups aren’t just subdivisions of the traditional “black, white, yellow” races. For example, Cavalli-Sforza has shown that Hungarians and Japanese share genetic markers to the exclusion of Chinese. So, what? You want to raise the Hungarian/Japanese group to the status of a race?

  11. #11 Ken Camargo
    April 22, 2006

    I couldn’t agree more with Lewontin and consequently with PZ on this subject, but just to highlight one issue present in a few of the comments, that of sexual selection, I’d like to point out how widely (and wildly) definitions of desirable traits vary, in a very short time, thus precluding IMHO “explanations” based on some sort of genetic determinism. Just an anecdotal evidence: when I was graduating from medical school here (Brazil, twenty-something years ago) the vast majority of breast surgeries were aimed at *reduction*; currently, *enlargement* surgeries outnumber those by a wide margin (reported by several colleagues and also easily observable by looking around ). The favouring of larger breasts in the “sexual market” (if you’ll pardon me) is quite unlikely to be explained by some genetic change in the population in such a small time interval; the overwhelming influence of American aesthetics through several media products seems a lot more likely.

  12. #12 windy
    April 22, 2006

    Races have been roughly defined for a long time, and now what will happen is we will match up genes with races. From what I’ve read some biologists have already done this. There are already groups offering racial-makeup tests. They will never be perfect but I think they are better than relying on self-report for affirmative action programs, for example.

    They are completely subject to interpretation. See the Gene expression blog article ‘Ancestry testing on crack’.

    If a test tells you that you are ’2% East Asian’ and you put that on your college application as your race, I don’t know how that’s better than self-reporting…

    And these are all neutral markers, and as you said, it’s about the phenotype, not the genotype. Why not go straight for the biologically meaningful genes and skip the messy correlation with “race” (unless you are studying the social dimension, for which the exact genetic ancestry matters less anyway?)

  13. #13 Jake
    April 22, 2006

    wrt Ken’s comment: and the popularity of large breasts suggests we are currently in a period of cultural conservatism. Big boobs have been historically most popular in periods generally considered to be conservative, such as the 1950s; smaller breasts are considered sexy in more liberal periods, such as the 1890s, the 1920s, and the 1960s.

  14. #14 windy
    April 22, 2006

    Darwin was horribly racist, at least according to our modern standards. It’s impossible to read “Descent of Man” today without wincing on nearly every page.

    If you define “horribly racist” as “probably less racist than my parents”, then maybe…

    “The American aborigines, Negroes and Europeans are as different from each other in mind as any three races that can be named; yet I was incessantly struck, whilst living with the Feugians on board the Beagle, with the many little traits of character, shewing how similar their minds were to ours; and so it was with a full-blooded negro with whom I happened once to be intimate.”

  15. #15 Jason Malloy
    April 22, 2006

    One person asked me for my view on human races. My default view was that of Richard Lewontin’s – that race is a biological construct, and that biological differences between “races” is less than within-”race” difference. But I hadn’t given it much more thought than that

    Jon Wilkins admits that he hasn’t thought about the subject much and it shows. Of course every misconception Richard Lewontin has offered on race has been debunked (PDF). The genetic differences within Chihuahuas are also greater than the differences between Chihuahuas and wolves and coyotes – different species! Isn’t it interesting that no one ever uses this to argue these species are somehow illusionary? Or that all their differences must be just “skin deep”?

    They rightly observe that while there are continental differences in genetics, there is no hard division. . . There is a genetic substructure to the human population, but it isn’t racial.

    A trite argument: define ‘race’ as species (or something even more concrete than that!), and then say it doesn’t exist. None of the architects of neo-Darwinism defined race like this – Theodosius Dobzhansky, Ernst Mayr, Sewall Wright, and GG Simpson – and I’ll accept their definition over ideologues like Lewontin, thank you. That genetic substructure is race, as they defined it. Race is what we call differentiation below the species level. It’s fine if PZ wants to disagree and call it something else, but I’ll go with the opinion of a majority of biologists and call it ‘race’. Especially since the campaign against the word seems to be away to get people to think that cultural differences should have no theoretical connection to biological differences – which itself isn’t true.

    . . .and genetic variation doesn’t match up with cultural differences per se.

    O RLY? Interesting use of ‘per se’; This is nonsense. Neighboring regions are almost always more similar across any number of measurable variables – physical, mental, behavioral, temperamental. For instance (just as an example) I see a big cluster of Christianity in the Europe region, and a big cluster of Islam in the Middle East region, and a big cluster of Buddhism in the Asia region. Cultural? Genetic? Could be either, or both. That’s for us to find out, not to simply assert – that’s dishonest.

    There are prominent markers that we use to recognize “race”, but they tend to be superficial and not really relevant to the substantive interactions between people

    False.

    As you would expect with populations with different evolutionary histories, allele frequencies differ across the genome. The differences aren’t simply found in junk DNA.

    I recommend Jon Goodrum’s Race FAQ for a more honest and substantive view than anything ever offered by Lewontin on the subject.

  16. #16 Jason Malloy
    April 22, 2006

    Sorry, a working link to AWF Edward’s paper on Lewontin’s Fallacy:

    http://www.goodrumj.com/Edwards.pdf

  17. #17 Buffalo Gal
    April 22, 2006

    I’m not a geneticist, but I know one who pointed out that racial differences are small but real. Some heart medication has been shown to be more helpful to people of African rather than European ancestry. In the article we read about this, medical ethicists were arguing that this sort of thing was somehow inappropriate for study because it discriminated among people. It seems to me that being able to test patients for a particular genetic trait that would help determine the best treatment for them would be a good thing.

    My point – we shouldn’t go so far in our attempts to be non-racist that we go around the bend and come out on the other side.

  18. #18 Brent T.
    April 22, 2006

    The statement by thwaite seems prophetic. “I predict there will be incoherence in this discussion using the word ‘race’ given its several definitions in Wilkins’ article, ranging from Blumenbach (old bad value-laden essentialism) to contemporary multi-allelic genetic clusters.”

    As someone who has engaged in genetic epidemiology in a population with about a 50/50 split between people who self-reported their race/ethnicity as “black or African American” and those who reported “white or Caucasian”, I can say from exeperience that self-reported race/ethnicity is related to genetics. This isn’t to say that race is an important biological construct for humans (it is in my belief primarily a social construct), but it is correlated with ancestry which is an important biological construct. Therefore, until ancestry can be reliably and conveniently measured either with genetic or non-genetic methods, self-reported race/ethnicity will have a place in genetic analyses. The term race really needs to be replaced when we talk about it in a biological sense for human, since humans do not have anything near the level of within species genetic variation that would qualify for the term race in other species. I think that non-genetic differences are much more important in terms of health and other things. I also think that racial disparities in health are not going to be corrected by race-specific treatments. My only point is that self-reported race/ethnicity does correlate with allelic frequency for many polymorphisms and if you want to investigate the role of a polymorphism and a health outcome you had better be sure that your results are not confounded by self-reported race/ethnicity. I have seen huge differences (10% in one group and nearly 60% in another) in frequencies of polymorphisms when stratified by self-reported race. I have NOT seen evidence that these differences substantially explain health differences.

    Why then is it important to at least think about self-reported race when looking at genetics?

    Example:
    Group A: has an allele frequency of 80% T and 20% t and 50% have health outcome Z
    Group B: has an allele frequency of 20% T and 80% t and 20% have health outcome Z

    An analysis that does not take “Group” into account will find that the T allele is highly predictive of outcome Z. When there are allelic differences between Groups of this magnitude it is important to check to see if the associations hold up within Group (i.e. within Group A do people with the T allele have greater outcome Z than people with the t allele?).

  19. #19 Jason Malloy
    April 22, 2006

    The term race really needs to be replaced when we talk about it in a biological sense for human, since humans do not have anything near the level of within species genetic variation that would qualify for the term race in other species.

    Brent, check my link, this isn’t true.

  20. #20 Jonathan Badger
    April 22, 2006

    If you define “horribly racist” as “probably less racist than my parents”, then maybe…
    “The American aborigines, Negroes and Europeans are as different from each other in mind as any three races that can be named; yet I was incessantly struck, whilst living with the Feugians on board the Beagle, with the many little traits of character, shewing how similar their minds were to ours; and so it was with a full-blooded negro with whom I happened once to be intimate.”

    If that’s all he had to say on the issue, then you’d be right, but I’d hope your parents wouldn’t agree that “the civilised races of man will almost certainly exterminate, and replace, the savage races throughout the world.”

  21. #21 anon
    April 22, 2006

    wrt Ken’s comment: and the popularity of large breasts suggests we are currently in a period of cultural conservatism. Big boobs have been historically most popular in periods generally considered to be conservative, such as the 1950s; smaller breasts are considered sexy in more liberal periods, such as the 1890s, the 1920s, and the 1960s.

    All the more reason for me to hope there will be a liberal resurgence in the good old US of A…..

  22. #22 Brent T.
    April 22, 2006

    Jason, I don’t see where in your link humans are compared to other species.

    I am not saying there is not variation. I intended to imply that sometimes there are important variations on the population level, but that race is probably not a good enough marker of genetics to work well on the individual level. My point is that on a population level there is important variation. On an individual level, it is somewhat doubtful that self-reported race/ethnicity is an important enough marker to dramatically alter treatment. I think that it might be possible to use population differences to help find individual (or sets of polymorphisms) that are actually causal. Then these actual polymorphisms could be used as targets for treatments. Targeting treatments at self-reported race is unlikely to be helpful, since it won’t be very specific/sensitive as a clinical tool.

    The finding mentioned earlier about heart medications has yet to be confirmed, and I am doubtful it will be very good as a “race-specific” treatment. This is a long story I don’t have time to get into.

  23. #23 Jason Malloy
    April 22, 2006

    Jason, I don’t see where in your link humans are compared to other species.

    Sorry, I didn’t specify which link. The FAQ explicitly compaes humans and animals. I’ll quote from a large post of mine, that deals with a lot of the misconceptions in this thread:

    Wise also completely invented the fact that human races are less genetically differentiated than animal races. Using Sewall Wright’s genetic distance statistic, FST, the value of differentiation for human races is typically 12-15%, Wright suggested this figure is on the high to moderate end of genetic differentiation among other animal races. This figure, by the way, clumps together a number of low differentiation populations and drags the number down; if we compare just the two largest sub-divisions of the human species – Africans and non-Africans – we get a number of 25-30%, which is huge. So at the very least we can say the human species has two enormously genetically differentiated races compared with the rest of the animal world.

  24. #24 blah
    April 22, 2006

    Nothing I have seen indicates that humans nicely group into distinct populations of less than the 54 found by Feldman’s group

    We have spent hundreds of millions of dollars on mapping the haplotype structure of Northwest Europeans, East Asians, and Yorubans from Nigeria. And the Hapmap was on the cover of Nature in 2005.

    Quote:

    http://www.hapmap.org/
    http://www.hapmap.org/hapmappopulations.html

    The goal of the International HapMap Project is to compare the genetic sequences of different individuals to identify chromosomal regions where genetic variants are shared. [See What is the HapMap?] By making this information freely available, the Project will help biomedical researchers find genes involved in disease and responses to therapeutic drugs. [See How Will the HapMap Benefit Human Health?] In the initial phase of the Project, genetic data are being gathered from four populations with African, Asian, and European ancestry. Ongoing interactions with members of these populations are addressing potential ethical issues and providing valuable experience in conducting research with identified populations…

    The International HapMap Project is analyzing DNA from populations with African, Asian, and European ancestry. Together, these DNA samples should enable HapMap researchers to identify most of the common haplotypes that exist in populations worldwide. [See What Is the HapMap?]

    Because of the history of the human species, most of the common haplotypes in human chromosomes occur in all human populations. [See The Origin of Haplotypes.] However, any given haplotype may be more common in one population and less common in another, and newer haplotypes may be found in just a single population. Efficiently choosing the tag SNPs needed to identify haplotypes therefore requires looking at haplotype frequencies in multiple populations. Also, genetic data from more than one population will enhance the ability of researchers to study the genetic contributions to diseases that are more or less prevalent in different groups.

    The DNA samples for the HapMap have come from a total of 270 people. The Yoruba people of Ibadan, Nigeria, provided 30 sets of samples from two parents and an adult child (each such set is called a trio). In Japan, 45 unrelated individuals from the Tokyo area provided samples. In China, 45 unrelated individuals from Beijing provided samples. Thirty U.S. trios provided samples, which were collected in 1980 from U.S. residents with northern and western European ancestry by the Centre d’Etude du Polymorphisme Humain (CEPH).

    Myers says:

    It is not real biologically. There are prominent markers that we use to recognize “race”, but they tend to be superficial and not really relevant to the substantive interactions between people.

    Frankly, I can’t believe that Myers is lying to his audience about this. This is not a word game. There are real implications for the health of ethnic minorities, and lying about the supposed “superficiality” of these markers is just not legitimate in 2006. Bidil is just the start. Read the Hapmap papers to see just how important and far reaching the variation, especially the variation related to immunity and pharmacological response is. Quote:

    http://www.pnas.org/cgi/content/full/103/1/135

    Overall, the observed genes in overrepresented GO categories are not random. For example, six functional categories constitute 82% of the HapMap CEU -log(EASE) scores of >0.65, represented by color flags in Fig. 5. We have defined these more general functional categories to include a number of individual GO categories associated with pathogen-host interaction, reproduction, DNA metabolism/cell cycle, protein metabolism, and neuronal function. We emphasize that many genes appear in multiple GO categories, and hence exact classification is not possible. Nevertheless, the clustering of most high-scoring GO categories into one of these generally defined functional categories is striking (Fig. 5). In the 112 genes with evidence for selection in all populations (Table 1) the proportion of genes in each of these categories is as follows: reproduction, 7%; host-pathogen interaction, 10%; cell cycle, 13%; protein metabolism, 15%; neuronal function, 17%; and DNA metabolism (including putative transcription factors), 21%.

    Selection for alleles in some of these categories might be anticipated, such as host-pathogen interaction and reproduction, given prior selection studies in humans and other organisms (15, 38, 39). Pathogen defense has long been suspected to be under constant evolutionary pressure. The beginning of agriculture and animal domestication 10,000 years ago not only brought domesticated animals close to humans but also established permanent human settlements (24). Such shifts from a hunter-gatherer nomadic lifestyle to agrarian societies likely facilitated the wide spread of infectious agents (38, 40). Our results suggest that human populations may have encountered many selective events associated with pathogen-host interaction. Examples of genes identified under host-pathogen interaction include CSF2, CCNT2, DEFB118, STAB1, SP1, and Zap70, and under reproduction, BIRC6, CUGBP1, DLG3, HMGCR, STS, and XRN2.

    The other overrepresented GO categories contain a number of unexpected genes. For example, it has been suggested that changes in organic compound metabolism may have been influenced by increases in meat consumption by early humans (41). Overrepresented genes in protein metabolism could be the result of this shift in dietary composition and/or the profound changes associated with a restricted agrarian diet (40). The large number of selected genes under DNA metabolism is also unexpected. We suggest that many of these selected alleles may be involved in the recent inferred increase in longevity of humans (42). Modifications to our immune system, increases in tumor suppression, and enhanced DNA repair (Fig. 5) are likely molecular components of our unique primate longevity. Some examples of selected genes in protein metabolism include ADAMTS19-20, APEH, PLAU, HDAC8, UBR1, and USP26, and under DNA metabolism CKN1, FANCC, RAD51C, HDAC8, PDCD8, and SMC1L1.

    One of the more intriguing categories overrepresented in inferred selective events is neuronal function. We define this category to include a diverse assortment of genes, including the serotonin transporter (SLC6A4), glutamate and glycine receptors (GRM3, GRM1, and GLRA2), olfactory receptors (OR4C13 and OR2B6), synapse-associated proteins (RAPSN), and a number of brain-expressed genes with largely unknown function (ASPM, RNT1; see Fig. 4).

    Seed should be ashamed of itself for publishing scientific misrepresentations worthy of the Bush administration.

  25. #25 windy
    April 22, 2006

    Jason Malloy wrote:
    The genetic differences within Chihuahuas are also greater than the differences between Chihuahuas and wolves and coyotes – different species!

    This sounds exceedingly strange. What is the source for that?

    In your post you refer to mtDNA of dog breeds. I found the value 4 % for wolf-coyote sequence divergence. Are you saying chihuahua mtDNA is somehow more variable than this? If so, how can this variation not affect the dog-wolf distance?

    I assume the meaning is to contrast “variation within human races is bigger than between” with “well, variation within dog breeds is also bigger than between them, and yet there are clear biological differences” I don’t know if the dog breed thing is true, it may be: but extending this comparison to between species is a bit strange.

  26. #26 windy
    April 22, 2006

    Jonathan B wrote:

    I’d hope your parents wouldn’t agree that “the civilised races of man will almost certainly exterminate, and replace, the savage races throughout the world.”

    It seems that at Darwin’s time ‘civilised races’ were in the process of exterminating and replacing plenty of people. It was something that he thought was happening. Although we don’t call people “savages” anymore, I don’t see how this statement is completely wrong. He was very confused on the what’s race and what’s culture part, of course (but this seems hardly restricted to Victorian times!)

    The quote is followed by: At the same time the anthropomorphous apes, as Professor Schaaffhausen has remarked, will no doubt be exterminated.

    It seems he was not advocating genocide of native peoples and apes, but lamenting that it might happen.

  27. #27 Jason Malloy
    April 22, 2006

    I assume the meaning is to contrast “variation within human races is bigger than between” with “well, variation within dog breeds is also bigger than between them, and yet there are clear biological differences”

    The meaning is that Lewontin contrived his 85/15 statistic to “prove” a number of lies about human variation, which it actually tells us nothing about, and a large number of people (many of them scientists) have been repeating this statistic for 30 years to “prove” things it actually doesn’t speak to at all. For instance Lewontin implied his statistic proved that human racial classification is not justified by the same criteria as other animal racial classification (or perhaps that all racial classification was debunked by his statistic). But racial classification is not and was never determined by this statistic in any animal, and 85/15 can be true even for animals belonging to different species.

    It has also been implied, not only countless times in the popular press and on the Internet, but in a number of peer reviewed journals, that Lewontin’s statistic “proves” that any observed behavioral/physiological differences between human populations can’t be genetic in origin – not enough genetic differentiation, you see. Again, this is a lie. 85/15 can easily contain the differences we observe between breeds of dogs and even between three closely related species.

    I don’t know if the dog breed thing is true, it may be: but extending this comparison to between species is a bit strange.

    It’s only strange if you think the Lewontin statistic implies something that it doesn’t. I’m surprised you believe that there is big, fundamental difference between closely related, interfertile species and closely related, interfertile races within those species. Species are defined by a closed-off breeding behavior, not by some dramatic, essential genetic dissimilarity.

    This sounds exceedingly strange. What is the source for that?

    James Serpell’s The Domestic Dog (link in post):

    “Recently using genetic and biochemical methods researchers have shown domestic dogs to be virtually identical . . . to other members of the genus . . . Results using mtDNA (mitochondrial DNA) data . . . reveal startling similarities among canids . . . Greater mtDNA differences appeared within the single breeds of Doberman Pinscher or poodle than between dogs and wolves . . . to keep things in perspective, it should be pointed out that there is less mtDNA difference between dogs, wolves, and coyotes, than there is between ethnic groups of human beings. (pp. 32-33)

  28. #28 Jason Malloy
    April 22, 2006

    Also, Windy, read the Edwards paper, 85/15 is for single genes. It’s when we start to add them all up that the recognizable racial and species differences emerge. You aren’t going to confuse one (genetic) Nigerian for one (genetic) Swede or one poodle for one coyote, unless you are looking at single genes. It’s really not that strange at all.

  29. #29 barbarpapa
    April 22, 2006

    Jason Malloy: I see a big cluster of Christianity in the Europe region, and a big cluster of Islam in the Middle East region, and a big cluster of Buddhism in the Asia region. Cultural? Genetic? Could be either, or both. That’s for us to find out, not to simply assert – that’s dishonest.

    Are you saying that religion is genetically determined? The distribution of religion can be explained by where the religions were first founded, migration, wars, and conquests. That is by the recorded history of the world. Are you saying that studying history is dishonest?

  30. #30 windy
    April 22, 2006

    JM wrote: Greater mtDNA differences appeared within the single breeds of Doberman Pinscher or poodle than between dogs and wolves…

    Isn’t this the same kind of error as saying there are greater differences between Europeans than between Europeans and Africans?

    The only way would be, if the network is something like

    (Dobermans + poodles)-(wolves)-(Dobermans + poodles)

    but this is not the case, dogs nest within wolf sequences.

  31. #31 Ken Camargo
    April 22, 2006

    Buffalo Gal:
    Yes, people of African descent probably share a number of alleles including some that relate to certain diseases (sickle cell anemia is one that springs to my mind) and probably to the response to certain drugs.
    The problem is that simply by looking at someone it is hard, if at all possible, to ascertain that they have “African descent” and/or have those alleles. IOW, we might look at someone who would be classified (self or otherwise) as “white” and yet that person could be of African descent and/or have those alleles, whereas another person classified as “African-American” might not.
    In any event, we are *all* of African descent.

  32. #32 Damien
    April 22, 2006

    “The distribution of religion can be explained by where the religions were first founded,”

    It’s not inconceivable to me that the nature of the religions invented in various areas, or which became popular, might have been subject to genetic influence. Not precise details of doctrine, but one area might have had more personality types which tended toward a particular style of religion than another. Conversely, given a memeset which dominated for a long time, the genepool might have been selected for types which matched the memeset better, so that we’d find correlations now which weren’t originally causal (the genes didn’t cause the religion, but serve to reinforce it now.)

    E.g. when Christianity was stamped out of Japan, the Japanese at the time most genetically prone to Christian martyrdom (not in service of the dominant culture but in defiance of it) might have been removed from the genepool, leaving modern Japan as less prone to convert to Christianity. The genes might influence not the religion itself but one’s willingness to convert, or to hold exclusive memesets.

    America might be more religious because lots of Europe’s more religious types came over here.

    I wouldn’t bet on these genetic stories. But then I wouldn’t have bet on a single gene toggling pair-bonding in voles.

  33. #33 Ken Camargo
    April 22, 2006

    On the supposed “Lewontin fallacy”: one of the authors that Edwards quotes is Cavalli-Sforza (his work is defined as “magisterial”, no less). Here’s what Cavalli-Sforza has to say about “race”:
    “There isn’t, putting another way, a constancy that would respond to the current definition of the word ‘race’. Distinguishing the races is a complex enterprise: we have always to be based on frequence statistics of a large number of characteristics from a large number of individuals, never on a single character. But the worst is that we don’t even know how to answer the question: ‘how many races are there on Earth?’” (Cavalli-Sforza, Qui sommes-nous? Paris:Flammarion, 1997, pg. 315, my translation)
    Another interesting resource is a web forum organized by the Social Science Research Council at http://raceandgenomics.ssrc.org/

  34. #34 Jason Malloy
    April 22, 2006

    Isn’t this the same kind of error as saying there are greater differences between Europeans than between Europeans and Africans?

    Yes! Rather it’s open to the same misinterpretation. Note that that is a quote from the book.

    Are you saying that religion is genetically determined? . . . Are you saying that studying history is dishonest?

    ‘No’ and ‘no’. Individual (genetic) differences affect religious preferences (PDF), and different aggregates of individuals (races), no doubt differ in the frequencies of these preferences as well. How or if this relates to current or historical events, I don’t know, but it is dishonest (and creationist) to write genetic differences off the story of cross-cultural differences a priori, yes.

    Incidentally, the leading scientist of human temperament, Jerome Kagan of Harvard, has in fact repeated the hypothesis in all of his books, that the (highly likely genetic) temperamental differences between low-reactive Europeans and high-reactive Asians has played a historical role in the relative success of Buddhism and Christianity among those populations. From Galen’s Prophecy:

    “This evidence invites speculation on the differences in the classic philosophies of Asians and Europeans. . . The commentaries on human nature by Martin Luther and John Calvin emphasize the anxiety, fear, and guilt that is endemic to the human condition and the extraordinary effort necessary to control these gnawing, unpleasant emotions. . . Buddhist philosophy, more attractive to Asians, emphasizes attainment of serenity as the ideal life goal. A person approaches this state by ridding the self of all desire because frustrated wishes are the primary cause of a state of unhappiness that most scholars translate as “suffering.” . . . The desirability of eliminating states of affective arousal is linked to a passive rather than an active attitude toward the world. It is hard to imagine many European philosophers or statesmen celebrating the virtue of a detached quiescence, as Lao-tzu did over two thousand years ago. . . ” (256-258)

  35. #35 windy
    April 22, 2006

    JM quoted “The Domestic Dog”: …there is less mtDNA difference between dogs, wolves, and coyotes, than there is between ethnic groups of human beings.

    What’s his source on that?

    VilĂ  et al’s Science dog paper (2000) says dogs averaged 5.3 bp difference over a 261 bp seq., and “dogs differed from coyotes and jackals by at least 20 substitutions and two insertions.”

    The famous Neanderthal paper had humans averaging 8 bp diff over 378 bp. Every Neanderthal differed from modern humans by at least 22 bp, and so were considered outside the range of modern humans.

    I don’t know if these are overlapping bits of the control region, and the human control region is a bit iffy anyway, but the values seem in the same ballpark to me. Humans don’t have more mtDNA variation than canids, unless Neanderthals are considered an ‘ethnic group’ :)

  36. #36 John Wilkins
    April 23, 2006

    Why couldn’t you all respond on my blog? This has been an excellent discussion. Don’t give Myarz the hits, he doesn’t need them.

    I have responded and clarified in a subsequent post.

  37. #37 Buffalo Gal
    April 23, 2006

    Brent T, Ken Camargo – Thanks for your comments. I admit to knowing just enough to get myself in trouble by speaking up. The article my acquaintance and I were discussing was from the Boston Globe science pages, not a scientific paper. Some of the ethicists quoted wanted to avert their eyes and pretend no differences existed between descendants of certain groups and shouldn’t even be studied. That’s the point that seemed counterproductive. I didn’t mean to imply that you could tell who has allele T vs. allele t by looking at them – I said being able to test for such a marker might be helpful.

    A non-race related example is the advances being made in identifying different types of breast cancer (estrogen-sensitive, not estrogen-sensitive) which may help guide treatment decisions. No one would say stop studying this and give everyone the same treatment because that’s the only fair thing to do.

  38. #38 Ken Camargo
    April 23, 2006

    Buffalo Gal: I don’t have the reference at hand right now, but there’s a researcher at McGill (Margaret Lock) who has been studying the impact of such studies (specifically, APOe4 markers wrt Alzheimer’s) on people’s lives, and to make a long story short, there are a lot of bad repercussions as well, so it’s not innocuous research (I think none ever is, anyway). The problem is that the connection between possible genetic markers and the actual development of disease is of a probabilistic nature, and relative risks vary a lot; what happens quite often, even in biomedical literature, is that people suddenly slide from probability into determinism, and thus “an allele that is associated with a relative risk of 0.n for disease x” becomes “the gene for x”.

    John Wilkins: Sorry! :)

  39. #39 TTT
    April 23, 2006

    It would make sense for human races / subspecies to exist. We see them in other species throughout the animal kingdom, after all. If there is (or was) a distinct, geographically isolated population, that population will come to display characteristics that others do not have.

    However…. the current understanding of “race” (black, white, brown, red, yellow) very likely has no correspondence to what the actual biological human races are.

  40. #40 Buffalo Gal
    April 23, 2006

    Ken – yes, things are complicated. I understand that knowing about risks makes people think the risk is deterministic. (Sigh) I don’t want to over-defend what was meant to be a simple proposition. I just wanted to make the point that knowing about a certain risk may help determine treatment. And that ignoring the differences between groups may put the ideal ahead of the real.

  41. #41 great_ape
    April 23, 2006

    Agreed Buffalo Gal. Until we have–and I suspect the day isn’t all too far away–a rapid and accurate system for teasing apart genetic ancestry to assess risk factors– “self-reported race” provides information about ancestry that can be useful.
    It’s not perfect by any stretch, but it’s *something*.
    It’s amazing how people will cling so desperately to notions that support their political ideologies that they will ignore scientific data and end up potentially harming the very groups they ostensibly aim to protect.

  42. #42 windy
    April 23, 2006

    It would make sense for human races / subspecies to exist. We see them in other species throughout the animal kingdom, after all. If there is (or was) a distinct, geographically isolated population, that population will come to display characteristics that others do not have.

    The latter part is true. But humans arrived rather late on the scene compared with many familiar animals, and have a long generation time, so probably subspeciation is too much to expect (whatever a subspecies is). If Australians had been isolated for the 50000 years they existed, that might be a start, but it seems there were repeated migrations to all continents.

  43. #43 Jason Malloy
    April 23, 2006

    I’ve posted a couple of comments since yesterday that haven’t shown up? :(

  44. #44 PZ Myers
    April 23, 2006

    Everything that I’ve seen has been posted here…I don’t see anything queued up requiring approval (and if there were, I’d just approve it.)

  45. #45 Jason Malloy
    April 23, 2006

    It would make sense for human races/subspecies to exist. We see them in other species throughout the animal kingdom, after all.

    The currently accepted subspecies definition/criteria in biology was formulated by Ernst Mayr, who (along with all the other leading figures of Neo-Darwinism) also held that humans qualified. As I pointed out above it’s exceedingly hard to find a definition of race that fits for other animals but not humans. Efforts like FST and 85/15 do not work, for the simple reason that our differentiation is comparable.

    “However…. the current understanding of “race” (black, white, brown, red, yellow) very likely has no correspondence to what the actual biological human races are.”

    Sure it does. This is an empirical question with an answer in Rosenberg et al. (2005):

    “In one of the most extensive of these studies to date, considering 1,056 individuals from 52 human populations, with each individual genotyped for 377 autosomal microsatellite markers, we found that individuals could be partitioned into six main genetic clusters, five of which corresponded to Africa [black], Europe [white] and the part of Asia south and west of the Himalayas [brown], East Asia [yellow], Oceania, and the Americas [red]“

    [parentheses added]

  46. #46 Jason Malloy
    April 23, 2006

    I think the spam filter digests trouble-makers with overly-formatted posts. Is there anything in there now?

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.