Mailbag: Modern High-Throughput Genomics Versus Race Realism

32 thoughts on “Mailbag: Modern High-Throughput Genomics Versus Race Realism”

• Emil Karlsson,

  I came across this one racist loser, who tried to convince me that black people never invented or discovered anything of value. He claimed that they couldn’t be equal to white people because Africa was so backwards. I than gave him a list of black scientists and inventors. He than moved the goal post and insisted that they weren’t pure black and that they wouldn’t have been able to do any of the things they did without their “white genes.” Never mind that he couldn’t even prove that such a thing people who are “racially pure” or “white genes” genes exist.

  I also encountered another ignorant racist who tried to convince me that every single invention attributed to a black person was really created by a white person, and guess what his source was? A “white genocide” website that talked about how it was all just an evil conspiracy to make white people think that blacks were their equals, so they wouldn’t try to protect their “race” from the “white genocide.” It listed a bunch of inventions attributed to black invented and claimed that they didn’t really invent them. Unsurprisingly, I don’t think the website listed its sources. Basically the guy expected me to believe that just about ever single modern day historian was involved in the “Anti white” conspiracy to make white people think they weren’t a “superior race.”

  It manages to be hilarious, terrifying and depressing at the same time, the degree to which some people will go to try to justify their irrational hatred and bigotry.

  • Unfortunately, what you describe is fairly typical tactics deployed by many race realists and white supremacists.

    It is certainly both depressing and entertaining when some of these individuals attempt to discuss modern genetics. Much of their talk about “white genes” and “racial purity” is of course telltale signs that their position is an imposter of science: trying to mimic the outward appearance of science, but having close to nothing in common with it when it comes to content.

    I like to point out to them that according to the largely mainstream out-of-Africa model, race realists are filled to the brim with gene variants that arose in Africa in individuals who were not white by any stretch of the imagination. The cognitive dissonance usually makes them deny the out-of-Africa model, settling instead for the multiregional origins model. Besides the problems with this idea, and their misinterpretations of it, their claims reduce to their standard claims about differential recent positive selection that is refuted or at the very least strongly contradicted by empirical data.

  • Emil Karlsson,

    I suspect that the typical “racial realist” probably has only a slightly better understanding of evolution and phylogeny than the typical creationist. I came across one that argued that the fact that blacks have apelike features in their skulls proved that they were a separate race. Never mind that all human beings have apelike features, including in their skulls, because humans are apes. I don’t remember his exact words, but I think he also said that blacks had a skeleton that had features that made them more suited for walking on all fours, which of course is utter nonsense. I have no idea why he thought that. I don’t even think that the most extreme racists in the antebellum south, or apartheid South Africa thought that black people were quadrupeds. I mean, I think it was pretty obvious even to them, that black people, like all human beings were bipeds. If there are really a significant number of white racists who actually believe that its normal for black people to walk around on all fours, they think that the most obviously baseless stereotype ever is somehow true.

  • I suspect that the typical “racial realist” probably has only a slightly better understanding of evolution and phylogeny than the typical creationist.

    Indeed. For many proponents of pseudoscience, evolution is merely a highly credibility scientific field that they try to appeal to in order to give superficial legitimacy to their irrational and unscientific opinions. They show very little interest in challenging their own beliefs and biases.

    Another commonality between the two is the repeated recycling of decades old claims that have been refuted thousands of times. But I guess that applies to pseudoscience more generally.

• Hello, this is the same RH that sent you the email. Thank you for writing such a detailed reply to my email. I think I understand the points you made about race and national development as well as your points about race and crime. There are some things I still don’t understand and I hope its okay if I ask for your help again.

  Firstly, I am a confused about the implications of Rosenberg’s and Li’s research about how 84.7%-95% of human genetic variation occurs within populations. Why does this fact refute race realism? I thought humans and chimpanzees share 98.8% of their genes, and yet humans are much smarter than chimpanzees. Even if whites and blacks share ~90% of their genes, why can’t the remaining ~10% of genes explain racial differences in intelligence and athletics? Is it because the between population genetic variation is only for genes that affect skin color, hair texture, etc.?

  Secondly, when companies like 23andMe do genetic testing and find someone to be 99% European, why doesn’t this support race realism? I don’t understand how people can be classified as having, for example, 10% sub-Saharan African ancestry and 90% European ancestry, if Europeans and Africans are not distinct groups. Are white Europeans more related to other white Europeans than they are to black Africans? I don’t understand this idea of “admixture”. 23&Me has this map where they divide the world by regions like West African, Northwest European, etc. Why do they divide people like this if Africa has most of the world’s genetic diversity? Europe has more divisions than Africa.

  Thirdly, is U.S. President Barrack Obama more related to white Americans or to black Americans? How will Obama’s children test when checked for their admixture? Will they show up as 25% European? If one of your grandparents is from Europe, does that make you more European than if your last European ancestor was 5 generations ago? Does that make you more European than someone who has one black African parent but many white parents? Does recent admixture count more than historical admixture?

  Thanks for taking the time to answer my questions.

  • Thank you for your continued interest in these issues. I will do my best to try to answer your remaining questions!

    Genes and alleles

    Firstly, I am a confused about the implications of Rosenberg’s and Li’s research about how 84.7%-95% of human genetic variation occurs within populations. Why does this fact refute race realism? I thought humans and chimpanzees share 98.8% of their genes, and yet humans are much smarter than chimpanzees. Even if whites and blacks share ~90% of their genes, why can’t the remaining ~10% of genes explain racial differences in intelligence and athletics? Is it because the between population genetic variation is only for genes that affect skin color, hair texture, etc.?

    This is a really good question. It turns out that it boils down to the difference between genes and alleles. A gene is, to put it simply, some bits of DNA that code for a functional product (such as a protein). An allele is a variation of a particular genes.

    For humans and chimpanzees, the genetic difference is about ~1% when it comes to single nucleotide polymorphisms (SNPs). A SNP is a place in the genome where e. g. the average human might have a T nucleotide, whereas the average chimp might have an A. There are also some insertions and deletions (i.e. segments that humans but not chimps have, and vice versa) that changes this figure slightly, but let us focus on the 1% SNP rate.

    If we look at the corresponding difference between any two randomly selected humans, this is ~0.1% (or 1 in 1000). A lot of this variation occurs in regions of the genomes that does not have any genes or regulatory elements and thus does not have any effect on any biological function or how we look. Again, if we look at things that are not SNPs, such as copy number variations (i.e. some humans can have two copies of a given gene, others can have only one. or three etc.), the figure is larger, but let us focus on SNPs.

    So your human-chimp figure you mentioned of 98.8% is for SNPs between humans and chimps. The cited figure of 84.7%-95%, on the other hand, is 84.7%-95% out of this 0.1% SNPs between any two humans.

    So roughly 95% of all SNPs between humans (0.1%) are between individuals, not groups. So the variation that are not between individuals is thus somewhere around 0.05*0.001 = 0.00005 or 0.005% or 1 in 20000 nucleotide bases. This is likely an overestimation, because humans can differ due to being different populations within the same continental group as well.

    So to sum up,

    (1) the vast majority of genetic differences occur between individuals,
    (2) the genetic difference between groups is so very small,
    (3) human genetic variation is mostly continuous across the globe rather than in discrete groups,
    (4) genes that have recently undergone differential positive selection have more to do with immune system function, metabolism and superficial morphological traits etc. than with functions suggested by race realists (such as intelligence or athletic ability)

    If race realism was valid, we would expect to see a large portion of the human genetic diversity to be between groups and not between individuals, human genetic variation would not be mostly continuous across the global and there should be dozens and dozens of genes that have undergone recent positive selection that relate to those features that race realists think differ among groups, such as intelligence and athletic ability.

    Taken together, these data are very different from what we would expect to see if race realism was valid. Thus, the available evidence is simply not consistent with race realism.

    23andMe, genetic testing and admixture

    Secondly, when companies like 23andMe do genetic testing and find someone to be 99% European, why doesn’t this support race realism? I don’t understand how people can be classified as having, for example, 10% sub-Saharan African ancestry and 90% European ancestry, if Europeans and Africans are not distinct groups. Are white Europeans more related to other white Europeans than they are to black Africans? I don’t understand this idea of “admixture”. 23&Me has this map where they divide the world by regions like West African, Northwest European, etc. Why do they divide people like this if Africa has most of the world’s genetic diversity? Europe has more divisions than Africa.

    Another excellent question. The reason 23andMe and other genetic testing companies do that is because they rely on genetic markers that correlate with recent geographical ancestry. They have a very informative page about it here.

    Their methods revolve around choosing their genetic markers to fit with their geographical divisions. This is easy, since you can find genetic markers for any arbitrary geographical division you pick and it shows that the clustering results of collected genetic data is strongly influenced by the initial hypothesis and what you sample. A fun example comes from Weiss and Fullerton (2005):

    Suppose, instead, that you chose Icelanders, New Zealand Maoris, and Mayans as your major ancestral populations. They would form three nice clusters and you’d be free to call them “major” or “races” if you wanted to. But ask yourself why you’re more likely to prefer the first choice of samples to the second. These are not inherent features of the data, but choices investigators make for a host of reasons, some so taken for granted that we may not recognize them as choices at all.

    Make no mistake. This is not postmodernism. Far from it. Instead, it is the admission that clustering will depend on what hypothesis the researchers have and how they design their sample selections. If you sample 20 people in those three regions, you will get genetic clusters. But if you had sampled homogeneously from people living between these groups, it would no longer be an obvious clustering of a “Icelandic race”, a “Maori race” and a “Mayan race”.

    Let us get back to 23andMe. They do not use a single genetic marker, because that does not give enough resolution or certainty. Instead, they use several. But this too is not enough, as a lot of people have multiple ancestries (humans obviously migrated during the past tens of thousands of years). This issue of multiple ancestries is called admixture. Since the genome of a person is a mixture from his or her parents, some regions of the genome could be more typical one some ancestral populations, whereas others might be more similar to others. It is these proportions that get translated into a percentage.

    However, this crucially depends on the reference populations available. The reason why 23andMe has more divisions in Europe than Africa is likely due to two reasons: (1) they do not have enough reference genomes from all around Africa to get a good resolution (they “only” have about 10k reference genomes worldwide) and (2) most of their customers are likely to be in North American and Europe, so they focus their efforts on regions where their customers live.

    So how come the 23andMe method is possible with most human genetic variation being between individuals? They simply look at the small fraction of human genetic variation that occurs between geographical groups (and probably the variation that occurs within populations within the same continent). If you use enough markers, this is enough to get a rough idea about geographical ancestry. Of course, the mainstream scientific position is not that all humans are clones, but rather that the vast majority of human genetic variation occurs between individuals, not groups.

    So are white Europeans more related to other white Europeans than to black Africans? It turns out that this is a hard question to answer. Partly because of admixture but also because traditional racial categories are typically not monophyletic; they do not include a common ancestor and all of its descendants, but typically either exclude some descendants, or include some other descents for which they do not share the most recent common ancestor with. For instance, there are people with dark skin color around the equator, but not all of them are more related to each other compared with other groups.

    The most honest answer, but perhaps the most boring answer, is that it depends on which white Europeans and which black Africans we talk about. We have a thirst for simplicity and lack of ambiguity, but sometimes the world just isn’t that simple or unambiguous. This is one such case.

    Genetic versus genealogical ancestry

    Thirdly, is U.S. President Barrack Obama more related to white Americans or to black Americans? How will Obama’s children test when checked for their admixture? Will they show up as 25% European? If one of your grandparents is from Europe, does that make you more European than if your last European ancestor was 5 generations ago? Does that make you more European than someone who has one black African parent but many white parents? Does recent admixture count more than historical admixture?

    Good question. The answer to this question involves distinguishing between genetic ancestry and genealogical ancestry.

    The number of genealogical ancestors double each generation we go back. You have two parents, four grandparents, eight great-grandparents, sixteen great-great-grandparents and so on. After just a few generations, you would (according to this model) have millions of ancestors! This is not really possible since that would soon be more people than lived on the planet in the past, so some of those ancestors are the same person.

    You get half of your genetic material from mom and half from dad. They got half of their genetic material from their parents and so on. If we go back long enough in time, you will soon get on average less than a single nucleotide from each of your ancestors! But this clearly do not make any sense.

    In reality, you get half of your genetic material from your mother who got half of her genetic material from her mother, but this does not mean that you got 1/4 of the genetic material of your grandmother. This is because of stochastic variation during the formation of sperm and egg. These genes do not know from which parent they came from, so in that half of all of your genetic material you got from your mom, it can very well be 75% from grandmother or 20% or 45% or essentially any figure.

    If we go back far enough in time, you are still genealogically related to your great-great-great-great…great grandparents, but you are almost completely genetically unrelated.

    I am not aware that we have access to the genome sequence of President Obama, so a robust genetic analysis is probably not possible. However, a lot of your questions in this third block does not distinguish between genealogical and genetic relatedness. If a man with one white and one black parent got a child with a black woman, his children would only be “25%” white in the genealogical sense of having 1/4 white grandparents. But this need not at all be the case when it comes to genetic ancestry. It depends on how much of the white parent’s alleles were passed on in the sperm and to the admixture of the child’s grandparents.

    If we look at African-Americans more generally, the average person has about 25% European admixture, but of course individual African-Americans can have much less or much more. From the same source, about 4% of European-Americans have (recent) African admixture. From the above arguments, recent admixtures will be more visible to researchers. This is because after a longer time period (and now we are speaking in geological time), gene variants tend to either reach fixation (everyone in a given population has them) or elimination (no one in a given population has them).

    I think I have answered most of your questions to the best of my ability. Feel free to ask follow-up questions on anything you want!

• Wow, that was very educational. What I cannot understand is why these facts are not taught in Canadian and American schools. I don’t know if they teach this sort of thing in Sweden and Europe, but across the pond they definitely don’t. In America they even screw up the history of race in their own country. For example, they act like Abraham Lincoln was anti-racist and wanted to end slavery for moral reasons. Far from it. He hated black people. He wanted to prevent slavery from spreading to the North because where slavery existed in the South, blacks and whites sometimes married. More slavery meant more black people and eventually more mixed race people.

  Americans don’t believe this today, because they think that white slave owners raped black women and that this is the source of the European ancestry of many African Americans. In reality many white slave owners freed and married their slaves, or married some of the few free blacks. Or sometimes white slaves would marry black slaves. Rape certainly happened, but the vast majority of mixed children were born to voluntary couples.

  There were many mixed race people who supported slavery because they thought it was a noble institution that would lead to miscegenation and more racial equality. However, Americans today cannot understand this because they are taught slavery = racism, and therefore don’t understand that support of slavery did not equate to hating black people.

  There were also a few “black” people who fought for the Confederates, and some Confederate flag types today say that this proves that the Confederacy wasn’t supportive of white supremacy. The Confederacy definitely was supportive of white supremacy, its just that the “blacks” who fought for them were actually mixed race people who were considered legally white, since they looked white enough. Nowadays, they might be considered black. Most Americans don’t understand this because they believe that the One Drop Rule was used to determine if someone was a slave. In reality the Rule was developed in the Jim Crow era.

  Frank W. Sweet has some great videos and articles on the “U.S. Color Line” where he busts a lot of misconceptions about race in America and talks about the things I mentioned above. If you ever have the time I encourage you to take a look at his videos on Youtube and his website.

  https://www.youtube.com/user/frankwsweet/videos
  http://essays.backintyme.com/item/category/the-color-line

  I don’t know why they don’t teach this kind of stuff in America, considering they have so many race issues. If they taught a combination of the history of race and the science of human genetic variation, I think it would help their country immensely. Then again, there is that thing in America with creationism and evolution, so maybe I shouldn’t set the bar so high (>.<)

  In Canadian schools, they just teach everyone to be nice, they never bother to provide substantial scientific evidence to show that racialism is nonsense. As a result, people think that being racist is mean or rude, rather than pseudoscientific madness. This becomes a problem when someone comes across white nationalism or books like "The Bell Curve", and he/she finds himself unable to refute the arguments of race realists. However, their isn't much racism in Canada, especially when compared to America. Canada has much more racism to the Native aboriginals than to immigrants. There is no mainstream political party that has a platform of anti-immigration. It's weird, it's like all the foreigners gang up on the indigenous population.

  I think the method you explained in your article about debunking bigotry is the best way to handle it.
  https://debunkingdenialism.com/2015/09/26/the-value-of-debunking-irrational-bigotry-over-emotional-outbursts/
  I think of race realism now as pseudoscientific bigotry, it's both inconsistent with biological reality and based on hatred and lazy stereotypes. Its bad in a scientific sense and bad in a moral sense.

  Also, sorry for the rant in my email. It was final exam week and I was feeling very anti-university. Re-reading my email I'm sure I probably came across as a right-wing loon, ranting without a care for citations. Thanks for putting up with my attitude and educating me.

  Sorry, but now I have another few questions to ask.

  I am curious, do you know if they teach human genetic variation in Swedish schools the way you explained it to me? Or do they teach it more like Canada? I don't know if there are many non-European people in Sweden so maybe they just don't bother.

  It's hard to tell because right-wing American sites like Breitbart act like Sweden is being invaded by millions of bloodthirsty-welfare-craving-muslim-demons, 100% of whom are male, who are viciously raping everything that moves and that Swedes are so left-wing, self-hating, and terrified of being labelled as racist that they are humbly submitting to the utter destruction of their glorious white Nordic nation by the Muslim hordes!!!!!!!!!!!!!! Either that or the EU along with the Kenyan-born Imam in Chief, Barrack Hussein Obama, are hellbent on Islamizing Europe at all costs and displacing Swedes with Somalian monsters!!!!!!!!!!!!

  Or at least that's how the narrative goes.

  Thanks again for your time. I really appreciate it.

  • Wow, that was very educational. What I cannot understand is why these facts are not taught in Canadian and American schools. I don’t know if they teach this sort of thing in Sweden and Europe, but across the pond they definitely don’t.

    It is taught in university biology classes, most likely in North America as well. It is hard to teach it properly on lower levels of education, since it involves concepts such as “nucleotide”, “single nucleotide polymorphism”, “admixture” etc. Perhaps it is possible to do it in a more simplified way, but a lot of these studies are quote recent in educational terms, typically being less than 10 years. In biology, 10 years feels like forever, but in education, things move more slowly.

    If they taught a combination of the history of race and the science of human genetic variation, I think it would help their country immensely. Then again, there is that thing in America with creationism and evolution, so maybe I shouldn’t set the bar so high (>.<)

    Yes, the contemporary and historical widespread nature of creationism in the public school system is a major stumbling block for good science education. The U. S. also has no national science standards, and it can differ between and within states. So there is often not enough cohesion.

    In Canadian schools, they just teach everyone to be nice, they never bother to provide substantial scientific evidence to show that racialism is nonsense.

    This is probably because it is easier to win the war of the values than the war of ideas. Basically, getting young people in schools to understand basic decency and fairness is easier than schooling them in fairy complicated biology.

    There is also the problem of the so-called familiarity backfire effect. Basically, the more you mention a myth, the more it spreads regardless of whether you have provided a refutation or not. So it is tricky to dispel myths in a way that is both pedagogic and effective.

    As a result, people think that being racist is mean or rude, rather than pseudoscientific madness. This becomes a problem when someone comes across white nationalism or books like “The Bell Curve”, and he/she finds himself unable to refute the arguments of race realists.

    Indeed, this is why we need to immunize young people against pseudoscience. For instance, explain that there is always a debate among scientists, especially about details. So next time they hear a sensationalist claim about how this or that detail of evolution is being debated, they do not falsely infer that evolution as a whole is in doubt among scientists.

    I think the method you explained in your article about debunking bigotry is the best way to handle it.
    https://debunkingdenialism.com/2015/09/26/the-value-of-debunking-irrational-bigotry-over-emotional-outbursts/
    I think of race realism now as pseudoscientific bigotry, it’s both inconsistent with biological reality and based on hatred and lazy stereotypes. Its bad in a scientific sense and bad in a moral sense.

    It is indeed important to combat pseudoscientific bigotry at both ends — point out why their claims are scientifically wrong and point out why their values are incompatible with legal and moral principles — because otherwise they will always find an escape hatch.

    Also, sorry for the rant in my email. It was final exam week and I was feeling very anti-university. Re-reading my email I’m sure I probably came across as a right-wing loon, ranting without a care for citations. Thanks for putting up with my attitude and educating me.

    No worries. I have had worse. 🙂

    I am curious, do you know if they teach human genetic variation in Swedish schools the way you explained it to me? Or do they teach it more like Canada? I don’t know if there are many non-European people in Sweden so maybe they just don’t bother.

    In first grade to junior high school, it is mostly about getting the basic principles right: what are atoms, what is DNA, how do we measure water content in plants, how does cell division work etc. At the university level, you get the fuller picture. There are many things I’d wish that they would cover in first grade to junior high school: critical thinking, evaluating sources, logical fallacies and cognitive biases, how not to be mislead by statistics etc. But in the end, it is a question about time. Adding something in means something else has to go.

    It’s hard to tell because right-wing American sites like Breitbart act like Sweden is being invaded by millions of bloodthirsty-welfare-craving-muslim-demons, 100% of whom are male, who are viciously raping everything that moves and that Swedes are so left-wing, self-hating, and terrified of being labelled as racist that they are humbly submitting to the utter destruction of their glorious white Nordic nation by the Muslim hordes!!!!!!!!!!!!!! Either that or the EU along with the Kenyan-born Imam in Chief, Barrack Hussein Obama, are hellbent on Islamizing Europe at all costs and displacing Swedes with Somalian monsters!!!!!!!!!!!! Or at least that’s how the narrative goes.

    This is a great example where critical analysis of facts, sources and narratives comes in. Breitbart is obviously an alternative far-right source, it is a narrative that provokes strong emotions and most of their core claims are false. This ‘triad of unreasonableness’ (low credible source, emotional provocation and lack of factual basis) should make large alarms go off in our heads.

    In Sweden, we have a political party in the parliament that was founded by Neo-Nazis in the late 1980s. Although it never explicitly promoted a Nazi political platform, a substantial part of their members had connections to organized Neo-Nazism before, during or after their stay. The party is called The Swedish Democrats, but this is an Orwellian name since they are nothing like the U. S. democrats. Currently, polls put them at 17%-20%. So 1 in 5 people in Sweden say they would vote for this party. Today, they have attempted to “clean up” their ranks and evicted the worst-of-the-worst Nazis in their lower ranks, but, if you don’t mind the Game of Thrones reference, the North remembers.

    This is a stark contrast to the claim that Sweden is generally left wing. Furthermore, there is a recent creep towards anti-immigration in establish left and right parties as well, so the fringe political stance against immigrants is slowly seeping into the mainstream.

    I covered the claims about immigration and rape in a previous post called How Anti-Immigration Activists Misuse Rape Statistics. Basically, the increase in the number of reported rapes comes primarily from a radical widening of the definition of rape that took place 1998, 2005 and 2013 and a doubling of the proportion of committed rapes that are reported to the police. If we look at crime victim surveys, the actual rape rates are in a long-term decline, although it fluctuates from year-to-year.

    But of course, if all you say is that rape rates are roughly constant or declining, you cannot tap into fear by using political propaganda.

    Thanks again for your time. I really appreciate it.

    No problem. I want to illuminate issues with science and critical thinking as much as I can. I don’t expect to convince every person with unreasonable beliefs out there, but rather that people might hear a claim and do a Google search and find posts such as this one. I try to focus mostly on the long-term.

  • RH and Emil Karlsson,

    America has a pretty severe problem when comes public education with elementary though high school, in particular when it comes to be the people who are in charge.

    Texas schools board rewrites US history with lessons promoting God and guns
    http://www.theguardian.com/world/2010/may/16/texas-schools-rewrites-us-history

    Map: Publicly Funded Schools That Are Allowed to Teach Creationism.
    http://www.slate.com/articles/health_and_science/science/2014/01/creationism_in_public_schools_mapped_where_tax_money_supports_alternatives.html

    Its better or worse depending on which state you happen to live in. I don’t think its as bad if you happen to live in say California where I live, which is one of the most progressive states. However, Texas is such a large state that US textbook manufacturers usually write their textbooks around what Texas wants, which is a pretty big problem.

    By the way, contrary to the title of the second article I linked to, they’re technically not allowed to teach creationism in American public schools. The Supreme court has ruled repeatedly that they cannot, that it violates the establishment clause, but they continue to do so anyway, either by doing so openly and risking a lawsuit, or trying to find a way to circumvent the law.

    Ten Major Court Cases about Evolution and Creationism
    http://ncse.com/taking-action/ten-major-court-cases-evolution-creationism

  • Hi RH,
    I agree with everything else and relish this dialogue you and Emil are having, having knowledge is of utmost important when we must defend science from those and misunderstand and abuse it. However, having just finished my degree in American South History, I must write to inform you of the fact that literature and study regarding the antebellum South has consistently found it is a myth that Masters married their slaves. Rape was extraordinarily common, and was the cause of the vast majority of “mixed” babies in the South at that time. It caused broken families, terror, and the children were never accepted by the white master and were outcasts in the slave communities in some cases. There is no evidence for white/black marriages in the South prior to the Civil War, and for likely even a hundred years after that racially mixed marriages in the American South were extremely uncommon, and frowned upon. Excuse the length of the post, but I do feel it important to correct the record on that as the true horrors of slavery are still being uncovered.

  • CHW,

    To quote Frank W. Sweet’s “Myth’s across the color line”:

    Myth 3. European genetic admixture in African Americans comes from rape. — This myth comes in two versions. Some Whites say that admixture resulted from Black-on-White rape after the Civil War. Most African Americans and White liberals attribute it to White-on-Black rape during slavery. In fact, although rape has happened across the color line in both directions, the current levels of African genetic admixture in White Americans and of European genetic admixture in Black Americans match what one would expect from the number of actual documented intermarriages. Before 1691, intermarriage was routine. Before the Civil War it was nearly as common as today. The intermarriage rate fell drastically during the Jim Crow era but has now recovered beyond its former level. Also, for as long as records have been kept, Black male/White female intermarriages have been more common than White male/Black female intermarriages. In some places and times, such as 19th-century Boston, the difference was extreme. Finally, among Americans with “mismatched” mtDNA and Y markers (matrilineal from one continent and patrilineal from the other), most have patrilineal European markers, but a large minority show the reverse.

    And from Sweet again on “How Much of the European Contribution to the African-American Genome Comes From Females?”:

    Virtually all African Americans have some detectable European genetic admixture. Some carry few Euro DNA markers and some have many; but overall, about 17 percent of the collective African-American gene pool comes from Europe. What fraction of that 17 percent comes from European females? Can DNA tell? The short answer is “no”. Let’s see why not.

    The question arises because it is fraught with ethno-political implication. Many Americans see Blacks and Whites as mutually hostile groups. About half of those who see “races” as adversarial blame Whites for the challenges facing the African-American community, challenges such as gaps in net worth, test-scores, and criminality. They attribute these challenges to White-on-Black racism past and present. The other half blames Blacks for their own plight, attributing it to an opposition culture that idolizes victimization and disdains diligence. The former ideology sees the European genetic contribution as resulting from the rape of Black female slaves by White male slaveowners. If much of the European DNA in the African-American genetic enclave were to actually come from European females, it would weaken this ideology.

    After some discussion, Sweet concludes that we cannot put a percentage on how much of European admixture in black Americans comes from European females because of several problems:

    There are other problems as well. In addition to the unsuitability of haplotype lineages for estimating genetic contribution, three unexamined assumptions lurk in the original question about rape by slaveowners.

    It assumes that all mating between Euro males and Afro females was rape. In fact, although intermarriage was illegal in most states during the 50 Jim Crow years, it was not uncommon in the Deep South before 1850.[v] It was common before the color line was invented in 1691.[vi] And in Boston of the 1860s, it was very popular indeed.[vii]

    It assumes that all slaveowners were White. In fact, many slaveowners in South Carolina and the Gulf Coast were biracial. In 1830, one out of every 25 Louisiana slaves,[viii] and one out of every hundred South Carolina slaves was owned by a biracial slave master.[ix] In 1839, 41 percent of the Coloured households of New Orleans owned at least one slave.[x]

    It assumes that all slaves were Black. In fact, tens of thousands of biracial slaves were considered White, due to their appearance and descent from Europeans. The one-drop rule of invisible Blackness was unknown in the antebellum South.[xi]

    In conclusion, it is not safe to put any number on the female share of the Euro contribution to the African-American gene pool. The most that you can say is that there exists an odd contrast between, on the one hand, the remarkable stability of the 65-35 male imbalance in legal B/W intermarriages over the past two centuries and, on the other hand, the fact that more African Americans carry European patrilineal haplotypes than carry European matrilineal haplotypes.

    I should mention that I am in no way an expert and I don’t even know the basic history of the antebellum South, so I could be absolutely wrong. Your personal knowledge of this topic definitely exceeds my own. Sweet’s website has a lot of info on the U.S. color line so I encourage you to look around. He also has a YouTube channel that has lots of good information. I also don’t want to hijack Karlsson’s comment section by getting too far off topic.

    If I am in fact wrong, I thank you for correcting me.

    Myth’s across the color line
    http://essays.backintyme.com/item/27

    How Much of the European Contribution to the African-American Genome Comes From Females?
    http://essays.backintyme.com/item/577

    Frank W. Sweet YouTube channel – The U.S. Color Line
    https://www.youtube.com/user/frankwsweet/videos

• Emil Karlsson,

  Thanks for your help. I find it sad though that you have less than 400 followers on twitter. Meanwhile race realists like HBD Chick have over 3500 followers on twitter. I think your blog is truly amazing but I find it disheartening that people who are so wrong can have so much more support.

  CriticalDragon1177

  Wow, that is very disturbing! Thankfully, I don’t live there. Then again you guys elected a muslim born in Kenya, so America must be a crazy place 😉

  I am curious about how you feel about the rise of “The Donald”, especially from your perspective of anti-bigotry (as claimed on your twitter page).

  • Thank you for your kind words. I do not really use Twitter that much to promote this website, but Debunking Denialism has about 2100 likes on Facebook. I do not care that much about celebrity status and focus more on putting out long-term useful content. People find a questionable claim online, Google it and end up here.

    I find it disheartening that people who are so wrong can have so much more support.

    Welcome to the realities of scientific skepticism.

    Although your question about Trump was to CriticalDragon1177, I want to point out that Donald Trump has pseudoscientific stances on many things, such as vaccines, climate change and renewable energy.

    I get why people support Trump. Many people think he feels like a fresh breath of air in American politics, he is not part of the dominating political families, he does not care about what other people think about him, he does not need the money from lobbyists, he says he wants to look out for the “common citizen”, and he appeals to white Americans who are tired of the liberal left, feminism, immigration, affirmative action and talk about racism and sexism (and some Trump supporters are likely not in favor of gender and ethnic social, economic and political equality either, to say the least).

    …but we have to look beneath the surface. Trump does support a wide range of pseudoscientific nonsense, and he has incited and take advantage of very questionable political stances, such as building a wall against Mexico and halting all immigration based on religious and ethnic background. One wonders, and now I am being intentionally provocative, if he has ever looked into any historical examples of building large walls (e. g. Germany and Israel) and read about how that went.

  • RH,

    I really don’t like Donald Trump. I fear for the future of America and the world, if he wins the election. I didn’t even like him before he ran for president, but I feel even more uncomfortable with him now. I’m a Bernie supporter, but I’m definitely going to vote for whoever the Democratic nominee is. I do not want that guy anywhere near the White House, especially since he’s been getting so many glowing endorsements from white nationalists, and he clearly is a racist himself. He also wants to stop all Muslims from entering the US, even if they’re refugees fleeing terrorist groups like ISIS. Last but not least, Donald Trump has a bizarre fantasy about how he’s going to solve our illegal immigration problem.

    Trump 2016: White Nationalists Throw Their Support Behind The Donald
    https://www.splcenter.org/hatewatch/2015/08/05/trump-2016-white-nationalists-throw-their-support-behind-donald

    Donald Trump Retweets One of His Favorite White Supremacists AGAIN
    http://littlegreenfootballs.com/article/45862_Donald_Trump_Retweets_One_of_His_Favorite_White_Supremacists_AGAIN

    Donald Trump: Ban all Muslim travel to U.S.
    http://www.cnn.com/2015/12/07/politics/donald-trump-muslim-ban-immigration/

    Trump reveals how he would force Mexico to pay for border wall
    https://www.washingtonpost.com/politics/trump-would-seek-to-block-money-transfers-to-force-mexico-to-fund-border-wall/2016/04/05/c0196314-fa7c-11e5-80e4-c381214de1a3_story.html

    In addition to all that, as Emil has pointed out, the guy also believes in a bunch of unrelated pseudoscientific nonsense, that would make him far from my first choice, even if it weren’t for everything else about him.

• Welcome to the realities of scientific skepticism.

  Lol, sounds painful. Keep fighting the good fight!

  The funny thing about American politics is that possible solutions to many of their most important issues require a level of humility and political realism that is not very popular. If they want to stop illegal immigration from Mexico they could focus more on how to “Make Mexico Great Again”, so that people don’t want to leave Mexico. There is a good reason why millions of Canadians don’t sneak into America. Canada is a prosperous country and there is no reason to go to America en masse. Perhaps trying to find a bilateral solution to the drug war would be a good start.

  As for terrorism, America keeps trying to blow people up in Afghanistan, Somalia, Iraq, Libya, and now Syria. All of those interventions are disasters of varying degrees and have large terrorist groups operating in those countries today. I can’t understand why some Americans think that they just need to blow up a few more people and everything will be okay. Trump thinks they need to kill the families of terrorists, because terrorism would somehow go away if some more non-combatants died. Seems legit (/sarcasm). Some of his foreign policy ideas make more sense though. Perhaps trying to chip away at the American military-industrial complex and decreasing military interventions would be a good idea instead.

  Helping Mexico develop and decreasing foreign military interventions may be easier said than done, but it definitely makes more sense than a giant border wall or continuing Operation Blow-Stuff-Up. Unfortunately, I can’t imagine someone being elected on a platform of “Make Mexico Great Again because it will indirectly help Make America Great Again”.

  Now Cruz and Kasich have left the race and so it will probably be Trump vs. Hillary Clinton. They both seem pretty terrible to me, they are just terrible in different ways. I’m not sure who is worse though.

• CriticalDragon1177,

  I am just speculating but I think Trump just wants to win the election and write a book and call it “The Art of the Presidency”. If he said he wanted to deport every muslim in America I am certain that his poll numbers would go up. I think that even if Trump loses, next election you will see another nationalist, anti-muslim, anti-illegal immigrant, protectionist republican candidate. Maybe the next Trump-like candidate will want to stop immigrants from third world countries.

  Everyone said Trump would lose early on, but he has made it this far. It seems like nothing can stop the Trump Train. Frankly, I think he will win the election. His polling is bad now but that can change. He is anti-muslim and against illegal immigration so he has the vote of many on the right. Meanwhile, he said he won’t cut social security and he wants protectionism to save blue-collar jobs so he has the support of many on the left. Everybody loves to hate Black Lives Matter, muslim terrorists keep blowing people up in Europe, and your muslim president gave gay people some rights. I think most people don’t say how they feel out loud, but I think most people want America to stay white and conservative, minorities be damned. This is all just speculation though.

  The funny thing is Sanders seems like a poor choice to me, since I am the kind of guy that would privatize the police if they would let me. Then again I don’t really now what he wants exactly. I live in Canada, and here we have so much coverage of Trump, but almost no coverage of anyone else. The news will talk about Sanders and Clinton but it isn’t said what their policies are. Same for Cruz and the other republican candidates.

  I think Trump will be the loudest, he will stand out, he will win, and then you will move to Canada.
  See you soon, lol.

  • RH,

    Things will have to get pretty bad for me to move to Canada. Trust me, the Democrats won’t just let him have free reign, even if he wins, but it will still be a bad day, if he does.

    As for my top pick for the job, I don’t really have time to go into too much detail on Sanders, and we’re kind of getting off topic, but lets just say that he’s a social democrat. You can checkout his campaign website if you would like to know more about what he stands for. Off course even if he wins, it would be naive to think that he would be able to do everything he says he is going to do.

    https://berniesanders.com/?nosplash=true

• A race realist blogpost by a person called SL has decided to respond to this article. I will not link to it because this website is not a platform for race realists, and I do not want to improve their search engine rankings.

  SL promises to write three parts, and the first part contains the same general fallacies and errors that have been refuted thousands of times before on this website and others. Unless there are any actually new arguments that are interesting to refute, I will settle with focusing on the countless errors in the discussion of genetic clusters and human genetic variation by SL:

  1. SL claims that the Serre and Pääbo (2004) study only had a sample size of 90. This is not true. The first two initial datasets used in the study was at 89 and 90 individuals, but they also analyzed the CEPH diversity panel (1066 individuals). This is the same dataset that the initial Rosenberg paper looked at. This shows that SL has probably not even read the paper he objects to.

  2. Serre and Pääbo (2004) objected that the sampling density was low: “we recommend an approach in which individuals from as many localities as possible are sampled”. Yet the Rosenberg et al. (2005) response paper cited by SL does not increase sampling density across the world (they used a dataset with the same individuals, but expanded with more markers), so they did not actually address this question.

  3. The Rosenberg et al. (2005) paper explicitly states that its data and conclusions do not support race realism:

  Our evidence for clustering should not be taken as evidence of our support of any particular concept of “biological race.” In general, representations of human genetic diversity are evaluated based on their ability to facilitate further research into such topics as human evolutionary history and the identification of medically important genotypes that vary in frequency across populations. Both clines and clusters are among the constructs that meet this standard of usefulness: for example, clines of allele frequency variation have proven important for inference about the genetic history of Europe [15], and clusters have been shown to be valuable for avoidance of the false positive associations that result from population structure in genetic association studies [16]. The arguments about the existence or nonexistence of “biological races” in the absence of a specific context are largely orthogonal to the question of scientific utility, and they should not obscure the fact that, ultimately, the primary goals for studies of genetic variation in humans are to make inferences about human evolutionary history, human biology, and the genetic causes of disease.

  It is hilarious that a race realist such as SL would cite a paper he thinks support race realism, when the paper itself explicitly says that it is not evidence for race realism. This shows, yet again, that SL probably does not read the papers that he cites or if he did, he has not made a serious effort at understanding them.

  4. Rosenberg et al. (2005) reiterates that the genetic variation that are the basis for their clustering is a very, very small part of total human genetic variation:

  The effect of a barrier is to add 0.0153 to Fst beyond the value predicted by geographic distance alone. As 0.0153 is not a large value of genetic distance, and because the addition of the B term produces only a modest increase in R², the discontinuities that give rise to genetic clusters—as we have stated previously [3]—constitute a relatively small fraction of human genetic variation.

  So there is no genetic basis for claiming that modern genomics validates traditional racial categories.

  5. Rosenberg et al. (2005) also points out that clustering depends on the K parameter that has to be set by the user:

  This method requires that the number of clusters be prespecified, and assumes either a particular model of allele frequency correlations across clusters [12,13] or that allele frequencies are uncorrelated

  So the researcher’s initial hypothesis affects the results (see e. g. figure 2, and think about what would happen if you had continued to increase K gradually, to, say, 30), not just the number of clusters, but also a model for allele frequency correlations.

  This is consistent with arguments that I have made before, namely that of Weiss and Fullerton (2005). It is easy to find any genetic markers for whatever geographic hypothesis you use to decide sampling. Another lethal blow for the race realism of SL.

  6. SL rehashes the claim that human F_st values are the same for other species that we have no problem labeling different populations as “subspecies”. But F_st cutoffs are ultimately arbitrary, since there is no objective definition of a subspecies and the concept is just based on them being “fairly” geographically isolated and human categorization.

  This is unlike species, which is based on the ability to produce fertile offspring. The biological species concept also has limitations, but it is not as arbitrary as the subspecies concept. Basically, researchers define subspecies in a very subjective way, then try to harmonize this with F_st values. But there is no F_st value that objectively define a species where there are subspecies compared with a species for which there are no subspecies. It is just an arbitrary limit set on continuous changes. Race realists do not understand that the subspecies concept is used because it is useful for describing species where there is substantial geographical isolation. It isn’t an objective classification scheme and humans are not geographically isolated to a large degree.

  I could go on, but I have made my point. Race realism stacks falsehood upon error upon misunderstanding when it comes to modern genetic research. Race realism is an imposter of science that tries to abuse research results to prop up its ideological belief system. This is especially and painfully obvious when you catch proponents not even having read or understood the papers they reference.

• Emil Karlsson,

  … we must not be misled by how the world seems. Instead, we must boldly explore beyond the limited scope of our own personal beliefs and biases by testing them against broad scientific data without being selective and seeing what we want to see.

  Very true. The funny thing is even if we just look at how the world “seems”, certain claims made by race realists still don’t make sense. I looked at some other facts and it seems that even superficially, national economic development is inconsistent with race realism.

  If race is a major factor in national success than why the massive development gap between North Korea and South Korea? Both countries are populated by Koreans. The CIA puts North Korea’s GDP per capita at $1,800 US. That’s lower than Swaziland, Lesotho, South Sudan, Sudan (North), Kenya, Rep. Congo, Nigeria, Ghana, and many other “black” African countries. Race realists may blame North Korea’s current poverty on dictatorial communism. Yet, Ethiopia is a communist, one-party state but few race realists will accept this as one of the reasons for Ethiopia’s underdevelopment. If North Korea is poor, they pin it on communism. If Ethiopia is poor they pin it on race. Race realists will look for confounding variables and alternative explanations only when it supports their theories. That’s how many of those people think.

  Furthermore, the Bahamas and Seychelles are “black” countries and yet they both have a higher GDP per capita as determined by the CIA than “white” countries like Serbia, Poland, Hungary, and Latvia, and Asian countries like China, Mongolia, and Vietnam. But of course no race realist will say that this is explained by the racial IQ superiority of blacks over whites and Asians.

  So even a lazy glance around the world doesn’t match with the claim that supposed higher racial IQ causes higher national economic success. Admittedly, scientific evidence that undermines the central tenets of race realism is better than trying to prove that some black countries are richer than some white and Asian countries.

  I think pseudoscience is probably more dangerous than political correctness (PC). Most people on the street can tell when PC culture is going too far but how many laymen have the skills to dissect and refute pseudo scientific claims?

  I found that race realist blog run by SL that was trying to refute your claims. I find it funny how you are referred to as “the race denier”. Just using that terminology portrays you as being ignorant. I am curious though, do most anthropologists, especially physical anthropologists, believe in human races as a valid biological concept? SL tries to paint it as Western European countries are PC and so they claim races don’t exist but non-PC researches in China believe in race. When I search “does race exist” online the first thing that comes up is this NOVA PBS page that gives both support and opposition to race realism. In the proponent section, the person supporting human racial categorization says that physical anthropology shows evidence for racialism. Do you know if that’s true? I can’t understand why PBS treats racialism as a 50/50 debate if mainstream science rejects racialism.

  Links below. Thanks.

  List of countries by GDP per capita:
  https://en.wikipedia.org/wiki/List_of_countries_by_GDP_%28PPP%29_per_capita

  Communist ideology of the ruling party of Ethiopia:
  https://en.wikipedia.org/wiki/Ethiopian_People's_Revolutionary_Democratic_Front#Ideology

  Does Race Exist – NOVA:
  http://www.pbs.org/wgbh/nova/evolution/does-race-exist.html

  • I think pseudoscience is probably more dangerous than political correctness (PC). Most people on the street can tell when PC culture is going too far but how many laymen have the skills to dissect and refute pseudo scientific claims?

    The answer is very few. That is why pseudoscience is so successful: alternative “cures” for cancer, creationism, conspiracy theories, HIV/AIDS denialism, hatred of modern agriculture that uses biotechnology etc.

    I think both pseudoscience and (anti-science) postmodernism are symptoms of a larger culture of anti-intellectualism and a preference for superficial and easy solutions.

    I am curious though, do most anthropologists, especially physical anthropologists, believe in human races as a valid biological concept?

    Before I answer this question, we need to clarify a couple of things.

    The first is that a lot of surveys do not properly distinguish between the genetic question and the social / geographical / categorization question. It is perfectly possible to categorize people into different groups based on superficial morphology or geography if we want to and we do this all the time (“Swedes”, “green-eyed people”, “thin people”). But we should never fool ourselves into thinking such categorization has more underlying genetic validity compared with any other arbitrary grouping. If you start with almost any hypothesis of geographical grouping, you can find genetic markers that distinguish your groups simply because the large size of the human genome (over 3 billion nucleotides) and the effects of geographical distance.

    The second is that we must take care to ensure we understand who are the relevant experts on the question of human genetic variation. The experts on climate are climate scientists, the experts on evolution are evolutionary biologists, the experts on engineering are engineers and so on. Others may have some relevant things to contribute, but they are not the substantial experts on these topics. Mathematicians can make valid contributions to evolution (there is more mathematics and statistics in evolution than a lot of people realize!), but they are not the most appropriate experts. This is also obvious from polling data, where mathematicians reject evolution to a much larger degree than biologists. In general, the appropriate experts on the topic of human genetic variation are molecular biologists and/or geneticists who work with human genetic variation data. There are other researchers, such as anthropologists, forensic medical experts that occasionally make claims about human genetic variation, but they are like the mathematicians who make claims about evolution. Sometimes they are right, but sometimes they are wrong because it is not their field of expertise.

    Third, most large mainstream anthropological associations reject race realism. Here is the position statement by American Anthropological Association:

    Physical variations in any given trait tend to occur gradually rather than abruptly over geographic areas. And because physical traits are inherited independently of one another, knowing the range of one trait does not predict the presence of others. For example, skin color varies largely from light in the temperate areas in the north to dark in the tropical areas in the south; its intensity is not related to nose shape or hair texture. Dark skin may be associated with frizzy or kinky hair or curly or wavy or straight hair, all of which are found among different indigenous peoples in tropical regions. These facts render any attempt to establish lines of division among biological populations both arbitrary and subjective.

    […]

    “Race” thus evolved as a worldview, a body of prejudgments that distorts our ideas about human differences and group behavior. Racial beliefs constitute myths about the diversity in the human species and about the abilities and behavior of people homogenized into “racial” categories. The myths fused behavior and physical features together in the public mind, impeding our comprehension of both biological variations and cultural behavior, implying that both are genetically determined. Racial myths bear no relationship to the reality of human capabilities or behavior. Scientists today find that reliance on such folk beliefs about human differences in research has led to countless errors.

    The American Association of Physical Anthropologists have a similar position statement:

    10. There is no necessary concordance between biological characteristics and culturally defined groups. On every continent, there are diverse populations that differ in language, economy, and culture. There is no national, religious, linguistic or cultural group or economic class that constitutes a race. However, human beings who speak the same language and share the same culture frequently select each other as mates, with the result that there is often some degree of correspondence between the distribution of physical traits on the one hand and that of linguistic and cultural traits on the other. But there is no causal linkage between these physical and behavioral traits, and therefore it is not justifiable to attribute cultural characteristics to genetic inheritance.

    11. Physical, cultural and social environments influence the behavioral differences among individuals in society. Although heredity influences the behavioral variability of individuals within a given population, it does not affect the ability of any such population to function in a given social setting. The genetic capacity for intellectual development is one of the biological traits of our species essential for its survival. This genetic capacity is known to differ among individuals. The peoples of the world today appear to possess equal biological potential for assimilating any human culture. Racist political doctrines find no foundation in scientific knowledge concerning modern or past human populations.

    So to finally answer your question

    – anthropologists are not the proper scientific expertise for questions about human genetic variation. That would be molecular biologists / geneticists who work with human genetic variation data. So those anthropologists who believe in the genetic validity of traditional racial categories have the same status as creationist mathematicians. We know the exist, but we should not overestimate their scientific relevance.

    – anthropologists generally reject race realism (see quotes above from anthropology associations), but may subscribe to the usefulness of social / geographical categorization. The surveys that race realists often trot out rarely distinguish between these two meanings of “race”. Even in the cases where surveys use “biological race”, this does not distinguish between genetics and e. g. superficial morphology.

    – race realists often appeal to the fact that many forensic anthropologists work with trying to “determine race” of a victim (some race realists, including SL, mistakenly use “physical anthropologist” here). If race does not exist, they say, how can forensic anthropologists routinely identify the race of a victim? But forensic anthropologists merely identify the social race category this person would be designated by society, not a supposed genetic race. A classic paper is Sauer (1992). The abstract to the Sauer paper (which is entitled “Forensic anthropology and the concept of race: if races don’t exist, why are forensic anthropologists so good at identifying them?”) is revealing:

    Most anthropologists have abandoned the concept of race as a research tool and as a valid representation of human biological diversity. Yet, race identification continues to be one of the central foci of forensic anthropological casework and research. It is maintained in this paper that the successful assignment of race to a skeletal specimen is not a vindication of the race concept, but rather a prediction that an individual, while alive was assigned to a particular socially constructed ‘racial’ category. A specimen may display features that point to African ancestry. In this country that person is likely to have been labeled Black regardless of whether or not such a race actually exists in nature.

    A good popular science coverage of this issue is How to Paint your Panda: Forensic Anthropology and Race.

    SL tries to paint it as Western European countries are PC and so they claim races don’t exist but non-PC researches in China believe in race.

    This is probably a correlational fallacy. It is telling that SL only used two data points, because you can always connect two data points with a line. In order to fully address the question if western European countries systematically differ in their conclusions about the supposed existence of genetic human races, we need a lot more data points. Even if there was a difference, this difference cannot immediately be attributed to political correctness, since correlation does not entail causation. There are probably other differences between scientific research in western Europe and China that could play a factor: relative amount of funding, quality, size of collaborations, size of data sets, government intrusion into research etc.

    When I search “does race exist” online the first thing that comes up is this NOVA PBS page that gives both support and opposition to race realism. In the proponent section, the person supporting human racial categorization says that physical anthropology shows evidence for racialism. Do you know if that’s true? I can’t understand why PBS treats racialism as a 50/50 debate if mainstream science rejects racialism.

    The reason why PBS does this is because they do it for all of their investigations, such as vaccines or supplements. They make use of a common journalistic practice known as false balance, which means they give the scientific position and the nonsense position equal time. It comes partly from a notion of “fairness” and “let all voices he heard” and partly from journalistic laziness, as it is often easier to do a “he-said, she-said” piece than to really investigate the issue and find out what is legitimate and what is not. Some news media has now rejected false balance, such as the BBC with respect to e. g. climate change. They will no long give 50/50 to people who say there is no climate change or that humans are not an influential factor.

• Some commenters have asked me to address the claims made by race realists about national IQ, brain size and why this does not vindicate race realism. This is because (1) we know from high-throughput genetic data that race realism is false, and that (2) the arguments and datasets used by race realists when they make claims about IQ and brain size are extremely poor.

  I have declined publication of these comments because Debunking Denialism is not a platform for proponents of pseudoscience to spread their nonsense, but I am more than happy to address the claims they make. Their views are heard and discussed, not just in a venue and manner of their own choosing (this is one reason why I started the mailbag series in the first place, even though of course many people email me or submit comments that do not promote pseudoscience).

  In any case, this article and several comments already address (1), so the rest of this comment will address (2).

  First, race realists think that the national IQ datasets they use are uncontroversial. In reality, there are extremely flawed:

  (i.) over half of their dataset (> 100 countries) do not have any IQ data, so the “interpolated” their data by taking averages of surrounding countries. This might be a reasonable approach if the number of countries without data was very few, but when you are basically inventing half of your dataset, it undermines the entire argument.

  (ii.) for those countries were IQ data does exist, a lot of the research is based on small sample sizes, few studies and unrepresentative / non-random samples. This cannot be arbitrarily extrapolated to national IQ. That would be like interviewing fifty 15-year-olds outside a local Starbucks and making claims about the coffee consumption of the U. S. population.

  Race realists, when talking about economics and national IQ, also fail to adequately eliminate reverse causation or third variables. The national economy influences the size of the middle class, which influences how many and what proportion of people get better access to nutrition and education (and other factors), which affects IQ. In addition, we know that there are many environmental factors that affect IQ, so comparing IQ scores without adjusting for these factors is a flawed approach.

  Brain size is affected by a host of environmental factors, and one cannot simply correlate brain size with IQ without taking these factors into account. Race realists often try to point to brain size at birth, but that obviously depends on how much energy could be invested by the mother, which in turn depends on her environment. So you do not wash away environmental difference because you compare things at birth.

• Yet another commenter appealed to the common race realist myth that Asians have higher IQs than whites. In reality, they are pretty much comparable, but Asians outperform Whites on standardized knowledge tests such as the SATs, which researchers such as Nisbett et al (2012) attribute to social environment:

  The academic achievements and high occupational profile of Chinese and Japanese Americans have inspired speculation about genetic superiority (Lynn, 1987; Rushton, 1995; Weyl, 1969). Flynn (1991) analyzed data from the Coleman Report for the high school graduating class of 1966. That large representative sample included a substantial number of Asian Americans. The Asian Americans had about the same mean IQ as White Americans (actually slightly lower) but scored one third of a standard deviation higher on the SAT than did White Americans. SAT scores may reflect motivational differences—for example, taking more and higher level math courses—to a greater degree than do IQ tests. Remarkably, Chinese Americans in the class of 1966 attained occupations of a professional, managerial, or technical nature at a rate 62% higher than White Americans. The picture that results is that Asian Americans capitalize on a given level of intellectual ability much more than do European Americans. The differences in achievement between Asian Americans and White Americans are not hard to explain on cultural grounds. East Asians are members of cultures having a Confucian background. An endemic belief in those cultures is that intelligence is primarily a matter of hard work (Chen & Stevenson, 1995; Choi & Markus, 1998; Choi, Nisbett, & Norenzayan, 1999; Heine et al., 2001; Stevenson et al., 1990). For thousands of years in China it was possible for the poorest villager to become the highest magistrate in the land through study. Families with a Confucian background exert far more influence on their children than do most families of European culture (Fiske, Kitayama, Markus, & Nisbett, 1998; Markus & Kitayama, 1991). They can demand of their children excellence in education and preparation for high-status careers and expect their children to try to comply.

  I suspect this myth persists among race realists for two reasons: (1) it is an easy way for race realists to deflect the claim that they are often white supremacists (“but we think Asians are genetically superior!”) and (2) it attempts to hide the impact of social environment as an explanation of the superior average performance of Asians on SATs, which is an obvious defeater for the race realist position.

• Another commenter asks if the rejection of race realism implies that all groups have the same intelligence.

  The implicit assumption is that since race realists believe that different groups have different intelligence (by citing the flawed and severely interpolated IQ datasets discussed above), this must validate race realism and anyone who questions race realism must therefore allegedly subscribe to the notion that all groups have the same intelligence. This is then mocked as political correctness or a belief in a blank slate.

  In reality, intelligence does differ among humans, but this difference is primarily between individuals, not between groups and any supposedly observed difference among groups cannot simplistically be attributed to any alleged genetic basis for group status, since many other things differ between groups as well. Correlation does not imply causation.

• The same commenter who wrote the comment criticized here responded to my objections.

  He acknowledges that correlation does not imply causation, but claims correlation is evidence for causation. However, the entire point with designating “correlation implies causation” as a fallacy is that correlation is extremely weak evidence that, in a world with thousands of factors and data mining, is hardly worth anything.

  He defends his correlational fallacy by historical revisionism of the research on the relationship between smoking and lung cancer. His implication is that smoking-causes-lung-cancer research was merely correlational research, yet researchers concluded that smoking caused lung cancer. However, this is demonstrably untrue.

  This is because research in the 1950s that showed that smoking causes lung cancer relied on prospective cohort designs and not correlational studies. They used a sample size of 188 000 people who they prospectively followed to 1955. A follow-up study with more than 1 million people showed similar results. More information about these studies can be found at the American Cancer Society in the article The Study That Helped Spur the U.S. Stop-Smoking Movement.

  Also, there was no possibility that lung cancer caused you to start smoking, so the problem with reverse causation did not exist.

  In order to move from a mere correlation, race realists need to greatly improve their case. A quick look at the Bradford Hill criteria shows that they are far, far off.

  Sooner or later, a lot of race realists brings up Jews and this commenter is no exception. He thinks that Jews have higher IQs than others throughout time and place, and that this difference is likely genetic because he does not personally buy a particular straw man of one particular non-genetic explanation.

  However, such conclusions are based on improper sampling procedures for such inferences. Here is Nisbett et al (2012) again:

  There is little good evidence about just what IQ levels are typical for Ashkenazi Jews. (There is even less evidence available for Sephardic Jews, and in the rest of this section “Jews” should be taken to mean “Ashkenazi Jews of European descent.”) Jewish IQs have been variously estimated to be 7–15 points higher than those of White non-Jews in Britain and America (Flynn, 1991; Lynn, 2004, 2006). All available studies, however, are based on samples of convenience. It is not clear to what we should attribute the greater overall intellectual ability popularly attributed to Jews and supported by (weak) data.

  Thus, is it not scientifically defensible to ignore the problems of weak data and even if the data was good, we cannot arbitrary attribute it to genetics. This is also the case because far from being genetically homogeneous, Jews (like many ethnic groups) have a varied genetic background that does not allow such simplistic inferences to “this must be genetic”.

  Another curious piece of evidence is that Jews achieve higher outcomes than predicted on IQ alone (ibid.):

  It is important to note that even at the highest estimates we have of Jewish IQ, Jewish accomplishment exceeds what would be predicted on the basis of IQ alone. Nisbett (2009) has argued that the numbers of Jewish Ivy Leaguers, professors at elite colleges, Supreme Court clerks, and Nobel Prize winners are greater than one would expect even if average Jewish IQ were 115. He has also noted that remarkable as the superior achievement of Jews is, the achievement difference between Jews and non-Jews is far less extreme than differences between groups in many other comparisons that cannot be explained on purely genetic grounds, such as the achievements of the Italians versus the English in the 15th century, of the English versus the Italians after the 18th century, of Arabs versus Europeans in the 8th century, of Europeans versus Arabs after the 14th century, and of New Englanders versus Southerners throughout American history.

  Another awkward series of fact for race realists.

• Emil Karlsson,

  If human beings are created by their genes, and human beings create their culture, can it be said that human genes create culture? For example, when Japan has less crime and corruption compared to Russia, some might say that Japan has a culture that discourages criminal activity and corruption to a greater degree than Russian culture. But that culture was created by Japanese people, who were in turn created by their genes.

  Even though both Japan and Russia have modernized (Japan especially) Japan and Russia are distinguishable from say Norway and Britain. Even when ideas of Western science and liberal democracy were introduced to Japan, Japanese culture still remains distinct from general Western culture and there are big gaps in certain areas of achievement. For example, Japan may be rich today, but it still doesn’t have as many billionaires as America. According to Wikipedia (https://en.wikipedia.org/wiki/List_of_countries_by_the_number_of_US_dollar_billionaires#Africa) Japan has 24 billionaires and the US has 536. The population of US is only ~3 times the size of Japan and Japan has been substantially Americanized, but America still has over 22 times as many billionaires as Japan. Japan and America have so much in common today, but some things about America just cannot be replicated by the Japanese, and vice versa. Could this mean that Japanese people as a society have certain biological variations that have lead them to create Japanese culture? There may be variation among individuals but if the net effect of every individual’s actions creates Japanese culture, and those individuals are created by their genes, does that imply that the culture of Japan is dependent on the genetic stock of the country?

  Some might say that the environment is the source of the cultural differences, but to take another example, the whole world has had the example of Western civilization and yet most countries today are still far less successful than Western countries, even though Western countries encompass a wide variety of environments. Native American peoples never turned the USA into a superpower despite being in America for thousands of years. Once the Europeans showed up it took only 200 years to turn America into the most powerful country in the world. Same environment, but a different group of people gave a incredibly different result. People will sometimes say that Americans have so many Noble prizes because their country is rich and has many opportunities. But who created that wealth and those opportunities? Americans of European descent. Same with rich Western European countries. People will say that they are “fortunate” to have so many opportunities and to live in a culture/society that encourages science, etc. But again, that culture did not fall from the sky as a special privilege to Western Europeans. They made that themselves. Even things that were invented by other peoples were taken by Europeans and used in better ways, such as Chinese gunpowder. Why didn’t the Chinese make guns themselves? Why didn’t China colonize the world? Same tool, different peoples, radically different implementations. Some areas of Europe were unsuccessful for centuries like Britain, but eventually they developed and became at one time the strongest country in the world.

  Sweden and Australia have completely different environments, but have similar developments and culture. Both countries are mostly inhabited by European peoples. What about European Australians versus Australian Aborigines? Same environment, but completely different results. Yamato Japanese versus the Ainu indigenous people of Japan? Same environment, but completely different results. Some might say that Australia has skyscrapers and modern technology, which is not the same as the natural lifestyle of Aborigines. But again, who created that modern technology? Who changed the Australian environment? Who brought a new culture/society into Australia? The European Australians did all of those things. And what creates the European Australians? Their genes created them. So their genes are actually creating the current environment of Australia too. That’s why I don’t understand this dichotomy between genetic influence versus environmental influence on human development. Human genes allow us to alter our environment and create greater influences on the development of others than nature can. It seems to be less genes vs. environment and more your genes vs. other people’s genes, especially in the modern world.

  People will say that Africans are poor because of European colonization, but who told Africans to sit around and be demolished by advanced European empires? Who told Africans not to have the Renaissance and the Enlightenment? Today, nobody is forcing Africans to be unsuccessful. Maybe it difficult for them to adopt Western cultures/ideas because that cultures/ideas were created by different peoples with different biological traits than Africans, and so Africans can’t adopt them. Africans may immigrate to other countries and be able to function in Western societies but maybe that is because they are surrounded by Europeans and Western culture, a culture created by non-Africans. If those same African immigrants had stayed in Africa they would have been surrounded by African peoples and African culture, and so their life would probably be worse. The difference is that in the West they will be surrounded by Europeans, in Africa they will be surrounded by Africans. The Europeans are long gone, but African countries are now being taken over by China, as if there is some inherent biological feature of African populations that allows them to be used by others.

  I don’t understand how differences in culture can be so large and seemingly distinct and not be attributed to biological variation (genetic or something else), when those same cultures are created by people and those people are created by their genes. Even if their is no large genetic variation between groups, if a country like Japan or a region like Africa has certain individuals who when summed up create Japanese and African cultures, then can it be said that Japan has biological reasons for being “Japanese” and Africa has biological reasons for being “African”? Human beings are just proteins and other chemicals made by DNA so it’s not as if the individual choices that shape a culture are purely random. Maybe the biological differences are not purely genetic. Could epigenetics be the cause of vast cultural differences between groups? Or perhaps mysteries of brain development that are not yet understood? Can’t there be other biological differences that are not dependent purely on genes, or perhaps are not understood well enough that can better explain the difference in culture and advancement of different groups of people than simply saying that the environment causes all the differences? For if the differences in culture and long-term success of groups is not biological, then what is it? What is the alternative explanation, and how does it not fall back to human actions and ultimately back to human genes?

  • If human beings are created by their genes, and human beings create their culture, can it be said that human genes create culture?

    Unfortunately, it is much more complex. Humans are “made” from biology, psychology and environment (both social and non-social). In addition, remember that human genetic variation between groups is just a tiny part of the total human variation. For these reason, these very crude arguments about genetics and national prosperity are not reasonable.

    To make this point explicit, consider the large differences between subcultures within a group. Let us look at, say, the Goth subculture and the suits and ties subculture (e. g. lawyers, business people etc.). These differences are at least as large as many of the differences you bring up, but we would not spontaneously think of a genetic explanation. In fact, such a proposal would seem weird.

    Native American peoples never turned the USA into a superpower despite being in America for thousands of years. Once the Europeans showed up it took only 200 years to turn America into the most powerful country in the world. Same environment, but a different group of people gave a incredibly different result.

    Note quite. Native Americans never had an industrial revolution and were hard hit by infectious disease brought by the Europeans, such as smallpox. Europeans might have moved to the same environment, but they clearly had a very beneficial cultural and environmental background. This realization is also fatal to many of your comparisons. It is not just about the environment on a given place people travel to.

    I discussed some key reasons for why Europeans became successful. Here is what I wrote in the article above: “The second involves geographical and cultural factors as detailed in Diamond (1997). For instance had better crops, better climate, more domesticated animals, more immunity to devastating diseases because of this, better geography with long west-east distances that enable crops and animals domesticated in one place to work equally well in other areas, small countries that competed with each other in trade and war.”

    I highly recommend the book because it broadens the perspective on what factors affect national prosperity.

    Even things that were invented by other peoples were taken by Europeans and used in better ways, such as Chinese gunpowder. Why didn’t the Chinese make guns themselves? Why didn’t China colonize the world?

    The Chinese did make guns themselves! In fact, they were the first to ever make firearms.

    The reason Europeans got their hands on Chinese inventions were due to a blooming trade with the Silk Road. Why did China not colonize the world? Because their rulers were not terribly interested in that model of expansion (using mainly a tributary system instead) and the traders came to China (so there was no need to go to them). For instance, one of the main reasons that Europeans sailed was to reach China via the sea to avoid trade fees from the Silk Road.

    It seems to be less genes vs. environment and more your genes vs. other people’s genes, especially in the modern world.

    We should be skeptical of these kinds of simple soundbites, because we know they do not accurately represent the complexities of biology, non-social environment, culture and history. Perhaps the best analogy for you is to consider the term “patriarchy”. Perhaps you consider this to be an exaggerated, oversimplified and abused term used by radical feminists to explain almost everything and all differences between men and women. In the same way, most biologists view attempts to explain most or all differences between countries and groups by referencing genes to be wrongheaded.

    But if you force my hand, I would say that we should think about individual human phenotypes as a mixture between genes and environment, differences between countries as almost exclusively environmental and cultural (with some rare exceptions, such as lactase persistence in adulthood and milk consumption).

    People will say that Africans are poor because of European colonization, but who told Africans to sit around and be demolished by advanced European empires?

    Go back and think about the factors brought up by the Diamond book discussed earlier. How do Africa and Europe differ in climate? Geographical distribution? Species of plants and animals that can be domesticated? etc.

    I don’t understand how differences in culture can be so large and seemingly distinct and not be attributed to biological variation

    Consider the case of goths and suits again. They differences are very large in a wide range of areas, such as employment rates, economy, social influence and many other cultural aspects. Is your intuitive reaction to attribute these changes to “goth genes” and “suit genes”? If you can understand how cultural differences can cause those differences, there should be very little problem in understanding how environment and culture can create those differences you observe, particularly when you realize that culture builds upon culture for hundreds of years. It is rather this latter fact (exponential growth) that explains the size of cultural differences.

    Even if their is no large genetic variation between groups, if a country like Japan or a region like Africa has certain individuals who when summed up create Japanese and African cultures, then can it be said that Japan has biological reasons for being “Japanese” and Africa has biological reasons for being “African”?

    No, because you just said that there was no large genetic variation between groups. The big genetic difference is within populations. It is this “summing up” that results in small between-group genetic differences.

    Could epigenetics be the cause of vast cultural differences between groups?

    Unfortunately, epigenetic modifications occur in individuals based on their specific environment and are only inherited a few generations before they are reset, so they are inadequate to explain the differences between cultures.

    Or perhaps mysteries of brain development that are not yet understood? Can’t there be other biological differences that are not dependent purely on genes, or perhaps are not understood well enough that can better explain the difference in culture and advancement of different groups of people than simply saying that the environment causes all the differences?

    Appealing to unknown mysteries when we do have robust social and non-social environmental explanations for differences, such as climate, geography, species available for domestication, trade, industrialization etc. seems to be grasping at straws.

    There are genetic changes that affect culture, such as how Northern Europeans and one area in Africa can digest milk in adulthood. But these example are few and far between and do not relate to classic race realist beliefs and tropes anyway.

    For if the differences in culture and long-term success of groups is not biological, then what is it?

    The simplified explanation is that cultural differences are based on past cultural differences and occur in a self-reinforcing manner. That is, if you discovered cultural element A you have a much better chance to discover cultural element B etc. The discovery of lenses enabled glasses, microscopes, telescopes and so on; the photoelectric effect enabled all technical inventions based on the photoelectric effect etc. Ultimately, this most likely originates in e. g. differences in environment discussed by Diamond.

    I told you this is a simplified explanation, and I truly mean it. There are rare exceptions, such as lactase persistence allowing milk consumption in adulthood that only some populations have and we know this has impacted culture of e. g. domesticating milk-producing cows. We should not forget these. But we should not also exaggerate their importance. To be objective and scientific, we must strike a good balance between theory and data.

    A journalist wrote a book (reviewed e. g. here) a few years ago that made similar arguments that you make. 144 population geneticists and evolutionary biologists signed a letter to New York Times Book Review rejecting his claims.

• It looks like SL posted two additional posts. Although virtually all of his claims have been debunked many times before throughout the Internet, let us look at the most embarrassing errors SL makes in his second part:

  1. SL spends a lot of time looking at unadjusted crime data stratified by group. However, he is not comparing apples to apples, but apples to watermelons, since those groups are not otherwise identical, but differ on many social parameters such as income, education, unemployment, amount of broken families etc. As we will see, these differences will account for a large chunk of the differences. In other words, correlation does not imply causation.

  For comparison, it is the same problem that occur when people look at wage data between men and women, without controlling for education, field, proportion of full-time workers etc. and then believing this is the pay gap that only depends on gender and on other factor. This is highly ironic, because many elements of the alternative right disagree with many forms of feminism.

  2. SL focuses a lot on whether there is a statistically significant difference between black and white violent crime rates. However, besides the problem with confounders discussed above, it is dubious to focus on statistical significance over effect sizes and scientific context (more on this later). This is because statistical significance does not imply that differences are large or that they are relevant or important in the corresponding scientific context. Conversely, statistical non-significance does not mean that two groups are equivalent. It is just the probability of how likely it is to obtain at least as extreme results given that the null hypothesis is true. I have written about the problems with an obsessive focus on statistical significance many times before on this website.

  3. SL suggests that blacks are overrepresented in crime statistics, but never bothers to put the observed unadjusted effect sizes in the scientific context. The most pedagogical example comes from Swedish crime data discussed in a previous article on Debunking Denialism (statistics from this report):

  – A person born outside Sweden compared with someone born in Sweden with both parents born in Sweden: 2.5x
  – A person being aged 15-17 versus those above 40: 2.5x
  – Men versus women: 3.5x
  – People who lack high school education versus people with a 3 year post high school education: 5.7x
  – People with an income below 4.8K USD per year compared with people with ~43k USD that amount: 5.3x
  – People on welfare versus people who have never been on welfare: 6.1x

  Thus, when we put the unadjusted overrepresentation of people born outside Sweden into the relevant scientific context, it is about the same as for young people and even less than men versus women. It is thus not a high overrepresentation and does not justify the intense obsession given by race realists and anti-immigration activists.

  A second problem is that SL only looks at overrepresentation and not absolute incidence. This is misleading because even a 2.5x overrepresentation translates into a very low absolute incidence. For instance, the same Swedish crime report states that over 99.7% of all immigrants have never been suspected of rape or attempted rape. Thus, focusing only on overrepresentation instead of also taking into absolute incidence paints a very misleading picture.

  It is also a bit ironic that SL uses the same general argument on overrepresentation and crime that some radical feminists apply to men. The reason this is ironic because many members of the alternative right rejects radical feminist power analyses like this with precisely the same arguments that I have used to reject the race realist stance on crimes and ethnic minorities: too narrow focus on overrepresentation, the ignoring of absolute incidence and so on.

  4. SL does attempt to discuss possible confounders, but do not think they are important. Unfortunately, his own data betray him. This is because SL only talks about how he thinks that ethnic group is a better predictor than these other confounders. But the argument is not that confounders are better predictors than group for crime, but that a large chunk of the supposed association between ethnic group and crime can be explained by these confounders. This fatal misunderstanding, together with the data presented by SL, severely undermines the race realist position.

  Looking at most of the graphs discussed by SL, it is clear that income, poverty rate, and unemployment explains a large proportion of variation in crime.

  5. Some of the studies cited by SL even contradicts his own position. For instance, Kposowa, Breault, and Harrison (1995) explicitly states that:

  “Even though blacks, Hispanics and Native Americans cannot be said to have similar cultures, it is at least conceivable, if not likely, that these groups may share subcultural values of violence. However, based on the above analyses we suggest that there are more plausible factors that these groups share, notably poverty, divorce and population density in the case of homicide, urbanity and population density for violent crime, and at least urbanity for property crime”.

  So what happens when they looked more closely?

  To explore this issue more fully a subsample of counties with high per cent black was investigated. Table Vll presents OLS regression estimates for homicide with a sample of 405 counties in which per cent black is greater than 25 per cent. This cutoff point is suggested first because it represents a substantial proportion of the population and second because it closely approximates one standard deviation from the mean for per cent black nationally (8.8 per cent). The mean for per cent black in this sample is 40.4 per cent. As expected, results indicate that poverty is the most important determinant. The only other significant factor is urbanity. The divorce measure and population density are not factors. Note that per cent black is also unrelated. […] Employing the same sample (405 countries with per cent black greater than 25), OLS regression estimates were computed. Results (not shown) indicate that the major determinant of violent crime is population density, followed by urbanity. Poverty is not significant, but per cent black is marginally significant at the .045
  level. Next, property crime was substituted as the dependent variable. Results (not shown) suggested that in this case density and urbanity are the strongest determinants of property crime. Virtually identical results for all three dependent variables were obtained when the sample cutoff point for per cent black was changed to 20 to 30 per cent.

  Unsurprisingly, other factors beside group was considered more important: poverty, urbanity, population density. Thus, SL rests his claim that ethnic group was still a relevant factor on a “marginally significant” p value of 0.045 without looking at effect sizes. In fact, the paper does not even report effect sizes, because data was not shown.

  So what did this paper conclude?

  In sum, the weight of the evidence suggests that the major causes of crime in areas in which blacks, Hispanics and Native Americans are strongly represented are the same factors that explain crime elsewhere. Thus these data suggest that race- and ethnic-specific explanations of crime are not necessary. We fail to support the subculture of violence thesis.

  This suggests that SL probably did not read even the paper closely, or did not completely think all of the results through. It is common among proponents of pseudoscience to just re-post descriptions of scientific studies they found on some other blog without going back and reading the entire paper in full. Similar issues occur with many of the other papers cited by SL.

  6. What happens if we look at other papers that adjust for confounders? One such paper, Sampson, Morenoff, Raudenbush (2005), found that 70% of the observed difference could be explained by confounders:

  The gap between Whites and Blacks in levels of violence has animated a prolonged and controversial debate in public health and the social sciences. Our study reveals that over 60% of this gap is explained by immigration status, marriage, length of residence, verbal/reading ability, impulsivity, and neighborhood context. If we focus on odds ratios rather than raw coefficients, 70% of the gap is explained. Of all factors, neighborhood context was the most important source of the gap reduction and constitutional differences the least important.

  Remember that this is just the confounders we are currently are aware of. There may be additional confounders that explain even more of the observed gap. Also note that the corresponding gap between “Whites” and “Latinos” completely vanished when confounders were adjusted for.

  So when SL claims that confounders are not important, he is mistaken about the facts.

  7. SL claims that the 18 factors that influence crime that I described in the article above is “guesses” and not based on evidence. However, these come from a mainstream criminology textbook, so they are certainly supported by evidence. The point with listing those factors is to present what factors mainstream criminology has concluded to be related to crime and group is not one among them. Another awkward fact for race realists.

  8. Another great irony is that SL posts a picture suggesting that the difference in incarceration rates might be explained by IQ and not “race”. But this undermines race realism itself.

  9. SL wants to make a big deal out of testosterone, but unfortunately we cannot naively infer that more testosterone means more aggression. This is because it can be reverse causation in that more aggression increases testosterone or that some other factor contribute to both. SL also does not look at effect sizes or the testosterone concentrations used in experiments and compare that with humans. So this line of argument falls flat.

  The study by Richard et al. (2014) only found a free testosterone difference of 2.5-4.9% and no difference in total testosterone. This is a tiny difference (about 4 picograms per milliliter) and neither this study nor SL has shown that this difference is practically relevant.

  In reality, this study dismisses practical implications, such as affect on disease risk, of this minor difference: “Based on previous studies on effects of sex steroid hormones on risk of chronic diseases or mortality, this modest difference is unlikely to explain racial differences in disease risk.”

  SL has not shown that this tiny, tiny difference is of any practical relevance. We should not forget that effects depends on dose and such a dose-free argument that SL makes is not reasonable.

  10. SL also trots out MAO-A and even briefly mentions that this is only relevant when it interacts with childhood abuse. However, he never bothers to discuss what percentage of variation in outcomes is explained by this or any other effect size. But one of his sources does!

  Ficks and Waldman (2014) produces meta-analytic odds ratios for MAOA – uVNTR of 1.08 or 1.14 depending on model selection. They chose the random effects model, which had an OR of 1.14 (95% CI: 0.98–1.32). Had this figure been 1.00, there would have been no increased odds. They also conclude that:

  Although the low activity alleles of these markers both appear to confer some risk, the magnitude of their effects suggest that each plays only a modest role in ASB [antisocial behavior] phenotypes. Findings from recent genome-wide association studies of complex traits indicate that genetic effects of this size or smaller are likely to be the norm for individual loci, and that hundreds or thousands of genes may contribute to variation in complex traits such as aggression and ASB

  This likely means that the percent variation explained by MAO A is very low. Looks like SL did not fully read or appreciate this paper, which has been a common theme. I wish that race realists would read and understand papers that they cite.

• Time to take a critical look at the third part written by SL. Like with the first two parts, almost all of the claims have been refuted repeatedly on this website and on other places on the Internet, so I will focus on the most vulgar problems.

  1. SL cites data from patents and Nobel Prizes, but fails to correct for confounders and known geographical biases. The most extreme example is Sweden, which has more Nobel Prizes than predicted based on population size. This is discussed in the tongue-in-cheek paper about the correlation between chocolate consumption per capita and number of Nobel prize winners. It also illustrates the general problem with naively inferring causation from correlation.

  2. SL wants us to look at GDP per capita and claims that since it appears to rise throughout colonialism, colonialism cannot have been much of a problem. However, this is not a good measurement to find the negative effects of colonialism. This is because GDP merely measures the total value of goods and services produced, but not e. g. the non-monetary cost of producing them, such as forced labor and resource depletion. GDP has many other limitations and flaws that makes in inappropriate in this context.

  Perhaps the greatest analogy for alternative right proponents to understand this issue is a growing Swedish economy, despite recently taking in 180k immigrants. Alternative right proponents often dismiss this growth as merely “artificial”, since the immigration causes more goods and services “without any real growth”. So alternative right proponents cannot have their cake and eat it too.

  3. SL invites us to look at population growth, thinking that if colonialism was as bad and deadly as historians say it was, how come population growth large and positive? To understand why such a question is fallacious, consider the HIV/AIDS epidemic in Africa.

  Some HIV/AIDS denialists such as Peter Duesberg does not believe that there is an HIV/AIDS epidemic in Africa and that there never has been one. How can he think that?! He bases his argument on the fact that population growth is large and positive in e. g. South Africa that was hit the hardest by HIV/AIDS. So how can we reconcile large population growth and a country hit hard by HIV/AIDS resulting in many fatalities?

  The answer turns out to be very easy: net population change is defined as births minus deaths. So a population can grow even though a lot of people are dying due to HIV/AIDS, as long as the number of babies born are larger. In other words, high birth rates mask the effects. How do we test this? We look at average life expectancy. When we do that, we see that during life expectancy dropped by about 15-30 years. This shows the massive impact of HIV/AIDS in South Africa and other African countries.

  The corresponding argument can be made when it comes to colonialism. We also have to keep in mind that it is possible to survive, yet have a really shitty existence, so number of fatalities is not everything.

  4. SL references a paper, Eaverly and Levine (2012), that argues for the benefit of European colonialism on income. However, this paper was not published in a peer-review journal, but on the website of a private think-tank in the U. S. There are a number of problems here: (1) it only looks at unadjusted correlations, which does not tell us anything about causation, (2) most data is missing between 16th century and 20th century, (3) they cherry-picked the time point at which to measure share of Europeans, and, perhaps most destructive to his argument, (5) there is no evidence at all that sampling was random or representative and (6) the paper carried out around 90 statistical tests (not including F-tests) without adjusting p values for multiple comparisons and (7) does not exclude reverse causation that Europeans stayed longer in richer African countries or third-variable problems. SL thinks that proponents of mainstream science has the burden of proof, but that is wrong. It is SL that has the burden of proof to establish that this preferred interpretation of correlational data is true. His claim that there did not appear to be a correlation between amount of precious metals and size of the European settlement, but this is not at all the same as saying that precious metals were not exploited, merely that there was no correlation between amount of precious metals and European settlement population size. Indeed, why would it be if metals was routinely taken back to Europe to fund religious wars?

  5. SL references a second paper from the same source, Feyrer and Sacerdote (2006), and this carried out > 200 statistical tests without adjusting for multiple testing. This seriously calls into question their results and interpretation.

  6. SL attempts to dismiss the negative effects of slavery by stating (1) that those slaves would probably have been slaves in Africa anyways and (2) being a slave was probably not that bad since it was tough being a white factory worker too. However, these claims do not hold up. The first one is a classic example of two wrongs make a right fallacy and the second one is a false equivalence, since it fails to take into account the negative psychological aspects of being a slave over being a paid worker, and SL does not present any evidence for his proposed equivalence.

  7. SL does not think that the fact that slavery broke families apart is such a big deal since it is less family rupture than modern times. But family rupture in the times of slavery was forced, whereas divorce or breaking up today is either something that both parties are alright with or one party wants. It seems unlikely that many slaves wanted their families to rupture while being a slave in a foreign land. Again, SL grossly underestimates the difference between coercion and freedom.

  8. SL thinks that discrimination has no relevance for income, but this is demonstrably false. We know lots of evidence for e. g. health and healthcare inequality between groups, and this obviously translates into income. If you are sick, you usually cannot work as efficiently or as much as you can when you are healthy. This is just health and healthcare, but it is easy to see that this is not an isolated phenomena, and similar forms of inequalities occur elsewhere as well. Together, this shows that there is a very real relationship between discrimination and income.

  SL is right in that overt, violent racism has declined, but this is not at all the same as saying that unconscious biases have declined. This has been investigated in the Implicit racism projects at Harvard, whereby they test people’s immediate unconscious biases. So when SL claims that income differences have widened despite less racism, this is not entirely accurate. If discrimination is driven by unconscious biases that a large segment of the population suffers from and has been roughly constant over time while income gap has increased, this refutes SL as ongoing unconscious biases and discrimination does not contradict a growing income gap. Instead, the evidence is consistent with discrimination influencing income gap. This is also self-reinforcing, since families with more money can give their children better opportunities which makes those children get more money and so on. This is not to say that there is not a lot of movement up and down for people in income brackets, because that does happen.

  9. SL does not deny the results of the callback studies whereby it is clearly shown that employers treat the same exact CV differently depending on if the name sounds stereotypically white or stereotypically black. Mainstream science considers this to be evidence of bias against African-Americans. Since all factors are controlled for since the CV is identical, this evidence is almost unassailable for the existence of discrimination based on unconscious biases.

  So how does SL try to get out of this? By citing a single survey on literacy from 1992 and claiming that whites are better than blacks even at the same education level, so a preference for people with “white-sounding” names over those with “black-sounding” names is rational. However, this line of thinking has many, severe problems:

  (1) it assumes that employers are aware of this difference, and that this is the reason they have a preference for the person with a stereotypical white name. However, it is unlikely that employers would know of this 1992 survey on literacy or that they would remember it when participating in these studies.

  (2) the survey is on literacy, but it probably does not matter if you write slightly slower for a career as a truck driver or other similar jobs. Even in academic jobs, the hard part is not necessarily the time spent on writing text, but on other abilities like e. g. pipetting, organizing courses and meetings or analytic thinking (two factors that are likely to strongly relate to obtained qualifications).

  (3) SL has not provided any evidence that the difference when controlling for level of education is big enough to be relevant for most jobs. It isn’t enough to say simply say that there is a difference. You have to evaluate the size in the scientific context.

  (4) the survey itself does explain the difference and they do not attribute it to group-status (p. 35, p. 60 in the file): “The remaining disparities in performance between White and Black adults may be the result of numerous factors. One plausible explanation is the variation in the quality of education available to these two populations. Differences in socioeconomic status are also likely to be a factor.”

  10. SL cites the supposed correlation between national IQ and wealth that I previously debunked and the same criticisms apply: (1) they invented over half of their IQ dataset, (2) the rest is based on few studies with small sample sizes and non-random / non-representative convenience samples. Also, neither SL or the paper excludes reverse causation whereby freer countries with better economy increases the average IQ by social and environmental mechanisms, such as the expansion of the middle class. Correlation does not mean causation and it is clear that al of the claims made by SL is based on wishful thinking. Most of his graphs on this issue are vulnerable to these objections. His claims about skin color predicting national wealth is no more evidence for his preferred causation theory than the correlation between the number of films that Nicolas Cage appeared in and the number of pool drownings (r = 0.66) is evidence that movies with Cage makes people kill themselves.

  11. The absolutely clearest example of wishful thinking on the part of SL is his claim that IQ predicts per capita rate of researchers engaged in R&D and his implicit belief that this means that IQ is the causal factor. Did it not occur to SL that countries that spend a lot on R&D would put more money into activities that raises average IQ, such as education (thereby expanding the middle class)? Later on, he does confess to this ability, but retorts that causation surely goes both ways. His argument for why the IQ -> R&D is dominant is that IQ predicts wealth within the same country. But this is unadjusted comparison for a lot of other factors. His own figure shows a 0.41 correlation with parental socioeconomic status in comparison with the 0.49 for IQ, hardly an impressive difference (indeed the paper itself dismisses the difference as insignificant). Had the study in question looked at the effects of IQ controlled for parental socioeconomic status, it would likely have shown a very different picture. Since this was also a correlational study, it could not answer questions of causality. Does IQ predict success on its own, or is it caused by parental socioeconomic status? In the discussion section, the paper freely admits that it could not address this question. But SL seems to think that it can. All of these studies also have the problem of not controlling for multiple testing.

  12. Perhaps stunningly, SL admits that the factors that I listed, such as climate, geography and availability of animals and plants for domestication is supported by evidence. One wonders then what the point was with his tortuous ruminations about race and IQ.

  13. SL describes a model for evolution that suggests that evolution could produce very large changes after 2000 generations. But this model has several flaws: (1) it assumes constant directional selection for 2000 generations, (2) it only takes into account a single trait, (3) it assumes a complete independence between traits (i.e. that it is not possible for selection to act on another trait that alters selection on the first), (4) it ignores the effects of drift, mutation, migration and other relevant factors we know exists in human history, (5) only takes into account additive genetic variation, despite the fact that we know that a lot of traits that we discuss is not primarily based on additive genetic variation, (6) assumes that heritability within a population is informative about the causes of between-group differences, (7) completely botches the definition of the selection differential, which is the the difference between the selected parents and the complete population, (8) other factors besides the trait influence survival, (9) ignores the data we have today indicating that the between-group genetic variation is only a small fraction of total variation and that (10) the genes that have undergone recent positive selection is largely unrelated to race realist fantasies about human evolution. I can go on, but you get the point.

  14. SL makes extremely simplistic arguments with respect to gene-culture co-evolution by proposing that violent crime selected against “gene variants associated with violent crime”. Unfortunately, reality is more complex than this. This is because such genetic variants may be beneficial in other contexts and factors (and we know that most genes do more than one thing), a violent lifestyle may better attract a partner (many famous criminals in prisons gets loads of attention from many women), and the violence might occur after the age where most reproduction happens.

  15. SL thinks that since culture varies between populations, selection pressures must also differ for many different kinds of behaviors. But this falls prey to sizeless science again. It is not enough to say that selection pressures differed, because those differences might be too small to be relevant compared with other factors (basically the same problem as the claims about testosterone we debunked earlier). Also, we know that genes that have undergone recent positive selections are not associated with the cultural factors that SL lingers on.

  16. SL says that race realists are not arguing that genetic diversity within a population is related to national success, only that genetic diversity between populations are related to national success. This is wrong on two counts. First, race realists do argue that ethnically homogeneous countries are more successful. Second, we know that the genetic variation between populations is only a small fraction of the genetic variation within a population. So if the later does not seem to matter, the former surely does not. Also, can SL show which alleles we thinks are relevant for national success what what percentage of variation in national success is explained by variation in those allele frequencies? Probably not.

• So there you have it! Due to popular demands, I have identified at least 6+10+16 (= 32) distinct flaws in the response written by SL, often with references to the primary scientific literature or using well-known methodological or statistical criticisms.

  I think this shows that race realism is based on a not-so-careful reading of the literature, reliance on grossly “interpolated” (invented) datasets, considerable usage of non-representative convenience samples, the stringing together of correlations without any particular focus on effect sizes or interpretations of research results in the appropriate scientific context (remember the 0.004 parts per billion difference in free testosterone without a difference in total testosterone that was suppose to cause African-Americans to commit crime!) and a willingness to almost always (with a few rare exceptions) ignore reverse causation, confounders and percent variation explained. I do not know any other word for this besides pseudoscience.

• Pingback: Mailbag: Water Fluoridation and Human Genetic Variation | Debunking Denialism

• Pingback: The Sixth Anniversary of Debunking Denialism | Debunking Denialism

• Pingback: Mailbag: What’s The Harm? | Debunking Denialism

Comments are closed.