The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 Gene Expression BlogTeasers
Selection? Let Us Compute

170px-Charles_Darwin_by_Julia_Margaret_Cameron_2Over at The New York Review of books H. Allen Orr has put up a reaction to A troublesome inheritance. It’s very similar to Jerry Coyne’s take, the part about science (e.g., population structure being non-trivial) is deemed acceptable, but speculations in the second half of the book are not as appreciated. This is not surprising, and seems typical for working population geneticists (though do note that R. A. Fisher’s A Genetical Theory of Natural Selection has quite a bit of sociological speculation in the second half). But I have to say that I disagree with Orr when he says that “it seems hard to maintain that educated people deny that DNA sequences differ subtly among continents.” Jerry Coyne has a follow-up post where he praises Orr’s review, but adds:

This is what I also claimed, and of course got slammed by the race-denialists who are motivated largely by politics. To a biologist, races are simply genetically differentiated populations, and human populations are genetically differentiated. Although it’s a subjective exercise to say how many races there are, human genetic differentiation seems to cluster largely by continent, as you’d expect if that differentiation evolved in allopatry.


You can see what Coyne is talking about in the comments of his blog, for example:

“The idea that human populations are genetically identical, and “races” are purely social constructs, reflecting nothing about genetic differences, is simply wrong." [quoting Coyne -Razib]

This is completely false. I don’t see how any analysis of genetic differences will produce a ‘black’ race that combines Africans, Sinhalese and Australian Aborigines, nor that will justify the ‘one drop’ rule.

Can you believe that Jerry Coyne actually has to respond to this sort of thing? The whole point of formal means of clustering populations is to avoid these a priori social constructions. Gross phenotypes like skin color seem only moderately informative of population history. But a great number of educated people talk about human variation with the taxonomic sophistication of an 11 year old from the Jim Crow south.

In any case, I want to highlight a second area where Orr has a mild slap at Wade, and that’s on the science. Others have touched on this, but a key issue is that Nicholas Wade is covering a beat which is changing by the month, so obviously a lot of the science has now been superseded. Orr states:

Wade’s survey of human population genomics is lively and generally serviceable. It is not, however, without error. He exaggerates, for example, the percentage of the human genome that shows evidence of recent natural selection. The correct figure from the study he cites is 8 percent, not 14, and even this lower figure is soft and open to some alternative explanation…

What’s the truth here? This is a very live area of science. Last summer a preprint was posted on arXiv, Genome wide signals of pervasive positive selection in human evolution. The title makes the conclusion pretty clear. It’s now been published in Genome Research. The authors argue that background selection is confounded with regions of positive selection, in such a manner that the latter is obscured. I blogged it when it came out, if you want to dig deeper. At this point all the controversies about selectionism vs. neutralism really are irrelevant, as there’s enough data to go around that you can actually concretely test hypotheses. In regards to humans my own position is now leaning more toward greater, rather than less, selection, despite the small effective size of our species. That’s because I suspect that we’ve only scratched the surface when it comes to ‘soft selection’ on standing variation….

 
• Category: Race/Ethnicity, Science • Tags: Genomics, Selection 
Email This Page to Someone

 Remember My Information



=>
Commenters to Ignore...to FollowEndorsed Only
[]
  1. Any insight as to what the MO of people like Aylwyn Scally is? Why are they so reluctant to say anything “controversial”? He disliked the even the use of the word “race” when it came to describing different populations. Is he representative of the field? What’s his deal? Liberal? Boring? Afraid of his career being Watsoned?

    Read More
    ReplyAgree/Disagree/Etc.
    AgreeDisagreeLOLTroll
    These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are only available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also only be used once per hour.
    Sharing Comment via Twitter
    http://www.unz.com/gnxp/selection-let-us-compute/#comment-169000
    More... This Commenter Display All Comments
  2. Let us be less cherry-picking with our quotes here. The full quote from Coyne is clearer…

    “I have read this book, and I think it’s pretty awful. One part of the book, though—Wade’s discussion of genetically differentiated subgroups, whether or not you want to call them “races”—is not too bad. Although there aren’t a fixed number of “races”, we can identify individual humans’ ancestry very well by using an assemblage of genes, and in some cases even identify the particular European village from which an individual’s grandparents came. The idea that human populations are genetically identical, and “races” are purely social constructs, reflecting nothing about genetic differences, is simply wrong. But as we all know, those genetic differences are not profound—they’re seen by aggregating data from many genes across the genome, and doing a kind of “cluster analysis.” In other words, “races” (or “subgroups” or “populations”) differ statistically, not absolutely. And most of those differences are not in genes whose function we know well, although a few, like some genes involved in skin pigmentation, do show, as expected, more profound differences among populations.

    But, except for politically motivated denialists, this has been known for a long time. Wade’s main thesis, and where the book goes wrong, is to insist that differences between human societies, including differences that arose in the last few centuries, are based on genetic differences—produced by natural selection— in the behavior of individuals within those societies. In other words, societal differences largely reflect their differential evolution.”

    Spot on. The book is “pretty awful” and “goes wrong”.

    Read More
  3. “For this Wade offers virtually no evidence, because there is none. We know virtually nothing about the genetic differences (if there are any) in cognition and behavior between human populations.”

    What a clown. We know something very important about them, they exist. ex. The iq gap between blacks-whites at higher ages – where iq is most heritable – is still incredibly high. The best explanation is that the gap is at least partially genetic. The evidence of recent polygenic selection for IQ also points to that fact. That fact that brains are different between the races (ex. brain size) obviously means different genetic evolution greatly affects the brain. It has been shown that blacks consistently commit more crime across many different situations. Genetic dispositions undoubtedly can exist to cause these behaviors. Selection for genetic differences causing behavioral differences undoubtedly took place in recent human evolution and the evidence does not point to them being small (them being nonexistent between populations on average would be a literal miracle).

    As shown by Freedman’s and Kagan’s works incredible behavioral differences are present before culture can make any significant difference. The race behavioral denialism is just as stupid as the sex behavioral denialism. Men committing more crimes is undoubtedly linked to genetics, it is the same with blacks. Entire neural networks are set up at birth to make people behave consistently differently, plasticity isn’t magical and the organisms will tend to certain behaviors. So, given the evidence, why is the purely cultural explanation taken as truth a priori? A partial genetic explanation for the behaviors should be the default. Demanding absurdly high standards, gaining popular consensus, and then throwing around words like “xenophobia” to silence others is a sad way of confronting what clearly aims to be a purely descriptive hypothesis. We shouldn’t throw around these hypotheses unless they’re supported by a lot of data claims Coyne. But how does Coyne expect a lot of data to come without people being able to voice the hypotheses publicly? Wade’s book may actually make some brave people start researching say the effects of MAOA-2R which can and should be used to reduce suffering.

    “In other words, societal differences largely reflect their differential evolution.
    For this Wade offers virtually no evidence, because there is none.”

    What? Every difference is, by itself, evidence that supports wade’s hypothesis. Why is there assumed to be no genetic component to the different behaviors? Assuming no genetics is garbage speculation.

    The sociological explanations are a complete joke a lot of time with nearly zero real predictive utility and are just as speculative. Sometimes they are just a fun exercise in interpreting history to fit current political prejudices (against men, whites etc.) Sometimes the field is filled with garbage meaningless postmodern language like “social contruct”, “gender” they fail to meaningfully define (along with bad philisophical diatribes against language that is actually meaningful like the classical use of the word race to describe consistent phenotypic differences between human groups. Nobody can deny the consistent phenotypic differences, we give these differences a word, that word is race, this isn’t very difficult.)

    A single mutation in one person may possibly make a big difference with a butterfly effect and must be taken into account in attempting to come up with the best explanation possible. A gene that is in 5% of the population must certainly make an incredible difference at the societal level. MAOA-2R has a solid theoretical foundation between brunner syndrome and the 3R allele. An incredibly dangerous gene that has survived in 5% of the population was probably not selected alone.

    “I seriously doubt that. My prediction is that nobody will remember this book a century from now, except, perhaps, as an example of how to justify your preconceptions by using weak data.”

    I’m afraid Jerry Coyne is going to be in for a big surprise as genetics continues to accelerate.

    Read More
  4. Scott P. said:
    I don’t see how any analysis of genetic differences will produce a ‘black’ race that combines Africans, Sinhalese and Australian Aborigines

    Did any of the scientific racialists of the past ever subscribe to a taxonomy that lumped Australian aborigines and Sub-Saharan Africans together?

    Read More
  5. […] – Selection? Let Us Compute from razib khan – “Over at The New York Review of books H. Allen Orr has put up a […]

    Read More
  6. I got my copy yesterday and read it. There are quite a lot of errors regarding basic nomenclature in the relevant sciences. For instance, a few times he talks about correlations in percents (nonsense, but interpretable). Another time he talked about percents of IQs (nonsense, IQs are not ratio scale).

    Then there were the times he wrote that we haven’t found that genes for IQ yet. Clearly wrong. We have found 10-20 by now. Then he wrote that no one would count these by racial groups. Well, that affair was begun last year already by Davide Piffer. See most recent publication: http://openpsych.net/OBG/2014/05/opposite-selection-pressures-on-stature-and-intelligence-across-human-populations/

    Read More
  7. […] responds. Coyne replies (big debate in the comments). Peter Frost on recent evolution. Razib Khan responds to Orr & […]

    Read More
  8. My take on the book is similar to that of Orr/Coyne: does a decent job of explaining population structure; too much speculation in the second part.

    However, I think Orr/Coyne/Wade all miss the most interesting piece of science regarding strength of recent selection: evidence that the N-S height gradient (about 1 SD of difference between the two regions) in Europeans is due to selection pressure. That would constitute an example of fairly strong (in the context of the debate over group differences in humans) selection pressure acting over relatively short periods of time (~ 10 kya or less). I would think this result, if it holds up, might require significant updating of priors for certain people. It also provides a good example of how science in this area should be done: observed phenotype group difference, large data sets (GIANT) teasing out the genetic architecture, tests for selection on associated genetic variants.

    http://infoproc.blogspot.com/2014/05/whats-new-since-montagu.html

    http://infoproc.blogspot.com/2012/08/recent-human-evolution-european-height.html

    Another point, for the cognoscenti: Wade does a good job explaining the difference between soft and hard sweeps. Orr notes that small adjustments of allele frequencies is one of the primary mechanisms for evolutionary change (so nothing new in Wade’s discussion; goes all the way back to Fisher), but many many readers, even biologists who aren’t in population genetics, don’t understand this point very well. So reading that section in the book would increase understanding for a large number of people.

    Read More
  9. Coyne didn’t actually respond to the comment, except basically to abuse the commenter.

    The basic model is three races, Caucasian, Negroid and Mongoloid. The most common alternative model, White, Black, Yellow and Red is even more superficial. And the notion of racial contamination (carried in real life to the absurdity of the one drop rule,) is still implicit in the everyday usage of race. It is a natural corollary to the notion of a hierarchy of races. And race is conceived by many as a biological cause of different historical and social outcomes. Races are phenotypes.

    As I understand it, the consensus between Wade, Pinker, Coyne and yourself, is that it has been demonstrated that you can scientifically demonstrate the objective existence of human races, descent populations, relying fundamentally on shared SNPs demonstrating common descent. I believe alleles in operational genes (protein coding and regulatory) are mostly disregarded. The few exceptions I know of, genes for adult lactose tolerance and high altitude adaptations and melanin/Vitamin D production, do not seem to produce results that can be mapped to global populations. Thus, scientific race is a statistical variation in genotype.

    Scientific race and race in ordinary usage are not even in the same category. One refers solely to phenotype, the other predominantly to genotype and they do not intersect in any practical way. We could think of the word “race” as a plastic tumbler. The ordinary word has practically no objective meaning, the “glass” is practically empty. And it seems to me that when pushed scientists will admit that their fine, hand-blown glass goblet, the one which does hold significant meaning, is closeted away with the expensive silverware and fine porcelain dishes. That is, scientific race is part of the laboratory or field expedition, no part of everyday life.

    When a scientific race advocate tells us that the cheap plastic tumbler version isn’t empty, it doesn’t seem very helpful. Wouldn’t it be better to just respond that study of the scientific “race” has disproven practically everything meant by the every day “race?”

    Read More
  10. And the notion of racial contamination (carried in real life to the absurdity of the one drop rule,) is still implicit in the everyday usage of race.

    this is cognitive science. not genetics.

    Scientific race and race in ordinary usage are not even in the same category. One refers solely to phenotype, the other predominantly to genotype and they do not intersect in any practical way.

    both these assertions are false. you can’t assert by fiat.

    also, why are you bringing pinker into this? he’s barely talked about this in any technical depth, because it’s not his field.

    , no part of everyday life.

    which is why population origin is an important in biomedical research?

    your comment doesn’t know you know much about the scientific field

    Read More
  11. “also, why are you bringing pinker into this?”

    I didn’t, he volunteered his support for scientific race and Coyne quoted his remarks.

    “which is why population origin is important in biomedical research?”

    The populations of interest in biomedical research don’t correlate with race as it’s used in daily life. This is grossly obvious for most genetic disorders, despite the obvious intense interest in descent in those studies. Studies of something like adult lactose tolerance pay much attention to population origin but those studies have nothing to do with popular notions of the Caucasian or Mongolian/Mongoloid or Negroid races. Studies of the effects of racial discrimination are not studies of scientific race.

    I don’t know how “aren’t a fixed number of ‘races’…an assemblage of genes…aggregating data from many genes across the genome, and doing a kind of ‘cluster analysis’… differ statistically, not absolutely,” is so badly misrepresented when I wrote “refers…predominantly to genotype…” In any case, statements like this one by Coyne were my source, not my personal fiat.

    “this is cognitive science, not genetics.”

    There is no possible referent for your “this” to anything in what you cited, which can make this sentence mean anything.

    Read More
  12. I think the point Steve Hsu makes is important regarding the ground-breaking result on the genetic variation in height between northern and southern Europe.

    This was established by showing that *standing variation* could account for the differences across these broad groups.

    A reasonable inference is that one might be able to demonstrate something very similar for IQ, using this model, combined with sorts of genetic variation uncovered in the Visscher et al. paper regarding IQ.

    Perhaps the most promising group would be Ashkenazi Jews, who have probably evolved higher IQs in the same frame of time as the differences in height from north to south Europe. Show that those differences mainly derive from standing variation, and it’s game over for the idea that differences between groups in IQ are purely cultural in origin.

    Read More
  13. stevenjohnson, no offense, but your comments are kind of stupid IMO. in fact, there are so many stupidities in each comment i’m overwhelmed attempting to reply. though i guess you think you’re really smart, so perhaps i’m the stupid one. please don’t comment again.

    Read More
  14. There are two issues here. The first is a sort of statistical one: does ‘race’ describe something meaningful or useful about human genetic variation? It would help a lot if users of the term were able to be clear about what they meant, but generally they seem to talk about it as a way to explain apparent clustering in summaries of human genetic variation, such as PCA and STRUCTURE plots. Now we know there is plenty of structure in human genetic diversity, and indeed we know exactly the kind of processes which generated it: demographic processes of population change, migration and interbreeding. This gives rise to a roughly tree-like relationship between groups of individuals, and in genetics we often use the term ‘population’ to indicate a collection of individuals who share some degree of ancestry. This is a loosely defined term, in that it can refer to very small and local groupings, e.g. the inhabitants of a single mountain valley, or very large continental-scale groupings, depending on the context of the discussion. Sometimes, the populations considered are not well defined in terms of ancestry, e.g. one is really strictly a subset of another – as is the case for example if you talk about Africans and non-Africans. In a technical context we accept such ambiguities on the assumption that everyone understands the real situation.

    Importantly, there is no ‘natural’ division of humans into populations, meaning a level at which we could cut the tree-like structure relatively cleanly and divide the branches into well-separated groups. In particular there is no level where you could cut it and get groups corresponding to continental populations, or indeed corresponding to most of the apparent clustering in PCA plots. That’s because in making these plots we select samples from predefined cohorts, rather than sampling evenly across human diversity. These plots, useful though they are, say as much about our choice of samples as about the underlying structure of genetic ancestry.

    Note also that in other species, where there is often far more genetic diversity and substructure than in humans, such natural genetic divisions do exist and are usually referred to as subspecies. And I have even seen different subspecies of butterflies referred to as races in this context.

    But where are human ‘races’ in all of this? Well that’s for someone who uses the term to explain. Is it a general non-specific term to refer to the fact that there exists human population structure? Or is it a more specific set of categories – in which case what are these, and since it isn’t a natural categorization, why is it useful?

    Which brings us to the second issue, namely that there already is a well-known and long-standing meaning to the term ‘race’. Before people understood human ancestry they often spoke in terms of supposedly well-defined races based primarily on appearance, like ‘Negroid’, ‘Caucasian’, ‘Mongoloid’ and so on. The logic of such categorization was such that individuals could be ‘pure’ or ‘half-bred’, and it was used to justify bigotry, persecution and worse. But geneticists abandoned this concept decades ago not because it is an unpleasant idea with ugly consequences, but because it has no foundation in genetics. Just like the aether, or phlogiston, it is an obsolete idea. Reviving it, given what we now know about human genetic ancestry, would be worse than idiotic.

    Yet equally, using ‘race’ now to mean something different (but not quite completely unrelated) is also stupid. The question arises: why would you want to use this word to describe whatever structure you are talking about, even if it was well defined, rather than any of the more neutral terms geneticists have consistently used over the last few decades? Using such terms enables us to talk about the real differences in environment and ancestry which have shaped recent human evolution, without invoking confusion with erroneous past ideas.

    Why is ‘race’ popular amongst certain individuals, almost exclusively non-geneticists? I think the first commenter above rather gives the game away by wondering if my and other scientists’ reluctance to use the term in this context was because we are ‘liberals’. I suspect his assumption of a political motive reveals his own unscientific and unpleasant agenda.

    Read More
  15. Well that’s for someone who uses the term to explain.

    http://whyevolutionistrue.wordpress.com/2012/02/28/are-there-human-races/

    “In my own field of evolutionary biology, races of animals (also called “subspecies” or “ecotypes”) are morphologically distinguishable populations that live in allopatry (i.e. are geographically separated). There is no firm criterion on how much morphological difference it takes to delimit a race. Races of mice, for example, are described solely on the basis of difference in coat color, which could involve only one or two genes.”

    i think this is part age. jerry coyne is older, and my personal experience is that older biologists are OK with the term. e.g., ernst mayr http://www.jstor.org/discover/10.2307/20027740?uid=3739256&uid=2&uid=4&sid=21104047087887

    Before people understood human ancestry they often spoke in terms of supposedly well-defined races based primarily on appearance, like ‘Negroid’, ‘Caucasian’, ‘Mongoloid’ and so on. The logic of such categorization was such that individuals could be ‘pure’ or ‘half-bred’, and it was used to justify bigotry, persecution and worse.

    this is somewhat ahistorical. this sort of scientific racism/racialism is really a function of 1800 and later. race concepts predate that, and though people would be prejudiced in a sense that we would understand it, they wouldn’t really be racist in such a crude manner. and they extend beyond europeans. e.g., indians and chinese had their own taxonomies (chinese during the ming dynasty made a distinction between northern and southern europeans).

    I think the first commenter above rather gives the game away by wondering if my and other scientists’ reluctance to use the term in this context was because we are ‘liberals

    if you read coyne’s opinions on this that’s his judgement.

    Read More
  16. “That’s because in making these plots we select samples from predefined cohorts, rather than sampling evenly across human diversity. These plots, useful though they are, say as much about our choice of samples as about the underlying structure of genetic ancestry.”

    I keep reading claims like this.

    But is there any reason to believe that the same clusters wouldn’t pop out if we did indeed sample “evenly across human diversity”? Is this a “problem” that may very well go away once we get more complete samples?

    Sure, there will be cases which will not fit particularly well into any cluster if we restrict the number of clusters to 5 or 6, and will more or less span the existing clusters. But that’s just the way clustering and genetics works: of course there are intermediary populations.

    I’ve read no evidence that it will break out in any other way. If this geneticist has such evidence, he should tell us what it is.

    Read More
  17. The other point regarding the quote is that it does not understand the perfectly legitimate reasons the cohorts might have been chosen, given that a full sample base is not available across all human populations.

    Namely, those cohorts would naturally be chosen so that they represent a fairly broad set of populations within areas that we know to be separated geographically. Why is this natural? Because we would reasonably expect that such barriers would limit gene flow, just as they do with animal populations. It would be irrational to start out with populations we would most expect to have experienced relatively significant gene flow even across those barriers, because they would confuse, rather than enlighten, the overall picture of gene flow. It makes vastly more sense to include those populations only after we have a better sense of the genetic variation and distances across the “core” populations of each geographically separated group, or, of course, better yet, when we have samples from all populations.

    Read More

Comments are closed.