 In the 2008 Olympics, the women's 100 meter was won by three women from the same country. Shelly Ann Fraser, Sharon Simpson, and Karan Stewart are all three from Jamaica and placed gold in a two-way tie for silver, respectively. Jamaica has a population of 2.7 million people, one one-hundredth of the population of the US, and an economy about 1,000 times smaller than the US. Yet they outperformed the US team in this event. Jamaica, in general, does well in Olympic sprinting events for a country of their size and economy. It's a matter of national pride. The question I'd like to answer in this video is, why are Jamaicans so good at sprinting? Do they have some racial ability for fast sprints? Is there some genetic variation crafted by natural selection that has lent them a slight advantage? It's an interesting question and should be something we can answer from the science of human population genetics. We'll ask some related questions as well. For example, Ethiopians and Kenyans tend to be overrepresented in marathon events, but not Jamaicans, or for that matter, Ugandans or Tanzanians. It's also been pointed out that there's a gap in the IQ measurement of people who self-describe as white versus black Americans. Why does this gap also exist between and within countries of the world? Why are Jewish people supposedly overrepresented in Nobel Prize recipients and on the high end of the IQ scale? Let's dispose of a few hypotheses. 1. The myth of the pure population. Look, outside of a few isolated islands in the Indian Ocean or Northern Pacific, there are no groups of people who have not been actively marrying outside their own small tribal unit. The human gene pool is a shared gene pool across the entire globe. This doesn't mean that the ancient Romans were marrying Aboriginal Australians. The process is analogous to a bucket chain, with each tribe only marrying to the group proximal to itself. But this creates a flow of genes that span the globe, jumping from old to new world at the northern extremes of Alaska, Siberia and Iceland. Humans are one of those rare species that span the globe without geographic isolation. Thus, whenever we apply markers of ancestry to any population, there is always evidence of significant mixing of heritage. Each person, regardless of self-identified ethnicity, is a mix of multiple continental races. There are therefore no pure races anywhere. How many commenters on white supremacy websites are aware of the fact that they likely have 5% or more ancestry from Africa in the very recent past, probably back about 6-10 generations. We are all mutt blends. This was demonstrated very clearly in the following incident. In 2006, James Watson, who received the Nobel for his work in discovering the structure of DNA, in a newspaper interview made comments that linked the social problems in Africa to the lower genetic potential of Africans. It was an off-the-cuff common, the kind of thing that pops up in casual chat with someone who grew up in a different era. The controversy he sparked led to his resignation as head of Cold Spring Harbor Labs in 2007. James Watson was one of the first people to have his genome sequenced, so his genetic markers are publicly available on the internet. The Icelandic Genomics Company, Decode, was reported as claiming that the Watson genome showed greater than 15% African ancestry. The data is still somewhat questionable, since the early genomes were very poorly annotated and aligned, but this would not be out of the ordinary. What that means for Dr. Watson is that he has a one in six chance of having won his Nobel thanks to genes inherited from his recent African ancestor. It also shows that no matter how European someone appears, you can't imagine that they are by any measure of pure ancestry. Let's test this idea of a pure population against our Jamaican athletes. None of them are entirely African or entirely European. All of them contain significant markers of both ancestries, and also Asian and indigenous American markers. The term for this blending of different groups in our recent genetic past is admixture. This is also the name of a program that calculates and represents the genetic mixing of populations. Here we have one such graph, sometimes called a waterfall graph, that represents the global admixture by cultural ethnicity or tribe. Each bar on the graph within the brackets represents a single individual, and the colors represent groupings. The human analyst tells the program how many clusters or groupings to create. Recall that this arbitrary and inferred number of groupings is called K, and admixture represents membership in these groupings that best explain the total diversity. Notice for example in this global population that none of the bars is entirely of a single color when we allow for a larger number of inferred groups, that is high value of K. Every tribe or ethnicity contains within its gene pool plenty of mixtures of markers from adjoining groups. Let's take a look at admixture in Jamaica. The average Jamaican has 20% or more European genetic markers, slightly higher than the rest of the Caribbean. Does this European ancestry explain their apparently superior athletic performance at sprinting? One recent paper addressed this question by measuring ancestry in both elite athletes and non-athlete controls, and the answer was a definitive no. The percentage of ancestry was almost identical in both groups. However, in the same paper, elite African-American athletes possessed significantly more European ancestry than non-athlete controls, which is the opposite of what most biological race proponents would have expected. It also raises interesting questions about why this should be so. Feel free to leave your own hypothesis in the comments. 2. Natural selection has guided different groups to be good at something. Evolution is constantly occurring in the human population, that much as certain. Natural selection is also constantly occurring in the human population. However, if we scan our genome for evidence of certain genetic variants differing in frequency between two populations, the results are surprising. The scientific phrasing for this is high FST alleles. Recall from the last video that FST is a measure of variation between groups. Taking the genome as a whole, about 11% of human variation can be best explained by variation between the classical races. The remaining 89% is the result of variation within those groups. This is an important fact, and I hope you'll keep it in mind when we come to some calculations later. If we zoom in on just a few genetic variations that happen to be very different between, in this example, the Yoruba people of Africa, the French people of Europe, and the Han people of Asia, we can see that there are some that show a strong difference in frequency depending on whom we compare to each other. These are the race specific or high FST alleles. So how many of them would we predict? There are about 3 million of a particular class of genetic variations called SNPs, single nucleotide polymorphisms. But a lot of them are only measured in a few individuals in the whole world, so we should consider the total variation as about 900,000, which all round to a million for this narrative. Of the million shared variations, about 70, total, not 70%, are found to have undergone what we would call race specific or region specific natural selection. 70 out of a million, 0.007% of human variation is the result of strong natural selection acting differently on people as far apart as France, Nigeria, and China. What genes do we find that show evidence of positive selection in different areas of the world? Well, you won't be too surprised, skin color, hair texture, resistance to malaria and pox virus, lactose intolerance, salivary amylase, and a few blood antigens relating to maintaining intergroup pregnancy. That is the only genes that we can detect that have been under racial selection or the ones you might have learned about in freshman biology. There were no genes associated with athletic performance or brain development that were identified as being under positive selection in a single group. This makes sense in our understanding of evolution. Can you imagine an environment that wouldn't select for smart athletic people over dumb, slow people other than the modern United States? No, just kidding. Even here the person most likely to pass their genes on to many offspring are the ones who are successful for reasons of intellect, personality, or physique. For most of our history, hunting skills, long distance endurance, disease resistance, fertility, and cunning have been positively selected to about the same degree in all environments. And again, the degree of selection of these types of traits tends to be pretty low. Even dumb people are quite capable of locating a mate, producing an offspring, and surviving long enough to raise those children to independence. Thus, the degree of selection is expected to be quite low. Scientists measure the presence of positive selection by what are called selective sweeps. This means that random junk DNA just around a gene that is being selected for tends to be highly conserved across the entire group, where it would normally just randomly mutate in the background if it weren't right next to a gene variation that is being preserved by selection. When we measure those genes that show selective sweeps, then back calculate the degree of selection needed to explain the conserved variation. It comes out to less than 1% fitness increase. That is, in those genes, which are acted on by natural selection differently between different population groups, the effect is really, really weak, less than 1 in 100 difference in relative survival. So, what this allows us to do is to toss out strong selection on human populations, on anything other than disease resistance and skin pigmentation. There could still be even weaker selection than 1 in 100 difference, but it would be hard to distinguish from background variation. What many proponents of biological race differences are really saying is that running from lions in Africa produces superior runners, but inferior brains, where the European or Asian was less reliant on physical prowess and more likely to need social or intellectual skills at the expense of brawn. This is a ridiculous, unscientific idea, and smacks of Lamarckism, the idea that the struggles and desires of the parents produce advantageous changes in the children. The genomics data suggests very little adaptive change in human populations, and very little positive selection for desirable traits that differs between groups. So if strong positive natural selection accounts for about 1% of differences between groups, and the total variation between continental groups is about 11%, why are those two numbers so different? The remaining variation is the result of founder effects, genetic drift, as well as what is called background selection. That is when ancient humans left Africa to colonize Asia, Europe, Australia, and the Americas, they represented only a small group of individuals, sampled from a much larger, more diverse population. In most cases, what happened is a natural bottleneck in the population. The remaining variations that had been present in the African population never made it to Asia or Australia. Thus if we were to generate a proper family tree of humanity, and if it was to represent all of human diversity, it would have to be rooted in Africa. Out of every 10 branches, a good 7 or 8 would still reside in Africa. Europeans, Asians, Australians, and the first Americans might all be twigs off a single branch. We might call the branches, for example, Pygmy, San, Yoruba, Bantu, and then all the non-Africans would be lumped into a Mero Eurasian with Australian Aborigines possibly being their own branch. There is one other contribution that is still the subject of active research, and that is ancient hominid admixture. You may be familiar with the fact that Neanderthal markers are suspected to be present in all living humans, though far more in non-African groups. While this has not been definitively established, and may be the subject of a future video, we have reason to believe that there was some intermixing between anatomically modern humans and the Neanderthals, possibly other ancient hominids. We can set an upper bound to this contribution and about 3% of markers. There are some pretty silly ideas that Europeans somehow contain genes from Neanderthals that Africans do not possess. This is simply not the case. At most there are some variations in certain groups of modern humans that are best explained by admixture with these ancient hominids. These same variations are also found in Africans, but at a much lower statistical frequency. If you are relying on Google or the popular press for the facts on this topic, you most likely have them completely wrong. Let's return to the topic of why Jamaicans do so well at Olympic sprinting events. How much of that is cultural and how much is heritable? We already have the numbers we need to make an approximate calculation. We established early on that about 11% of genetic diversity is best explained by differences between continental groups, the FST measurement. To what degree is athleticism heritable? If we propose a very large number, like 50% for example, then the racial sprinting ability of a Jamaican can at most be one half of the FST value of 11% or about 5-6%. If the heritability is more like 20%, a more realistic number, then we are talking about 2% of someone's talent being the result of their membership in a particular continental group. At the level of Olympic performance, where the difference between a silver and gold medal is measured in microseconds, this slight genetic advantage could translate to something actually measurable. A large country though should still have an advantage as there are no performance genes denied to any population anywhere. We see this in the sudden dominance of Chinese athletes in all categories of Olympic sport. At best, we could say that there are 2-6% more people per capita in Jamaica versus Ecuador, with exceptional genetics for sprinting. But this small difference certainly doesn't account for the large imbalance in Olympic models. Instead, we have to look to the interplay of culture, environment and the genetic landscape. I started with 50% as an example because that is one proposed measure of heritability in IQ. In identical twins, there can be a pretty strong correlation between IQ values. We can't distinguish though, what effect sharing the womb has on both developing brains in maintaining this correlation, and how much is strictly the result of shared genetic variation. 50% heritability is a pretty high degree of genetic correlation, but it's important to note that 50% is how much your parent's genes determine your phenotype. The unit that this gene travels in is a direct family line. Let's take the best case scenario. If Albert Einstein and Marilyn Monroe were among your grandparents, you would have only a 1 in 4 chance of getting Einstein's genes for intellect or Marilyn Monroe's genes for looks if such genes existed. Then you would have only a 50% chance of those genes putting you on the same intellectual or physical potential level as either one. So your chances of either being exceptionally smart or attractive are only 1 in 8 if the degree of heritability were 50% for both traits. I would not expect Einstein's grandson to be exceptionally smart, even if such ability were inherited according to the Mendelian rules of genetics. The statistical models suggest that everyone alive today is, at most, 50th cousins. That is, if we go back only 50 generations, everyone alive has an ancestor in common. That's about our hereditary relatedness, not our genetic relatedness. Our genetic ancestry is diluted very rapidly. It's therefore unlikely that anything more than third or fourth cousins are likely to be genetically very similar. If we take the idea of a few hundred races, they are constructed of groups of people with some arbitrary degree of relatedness that is higher than that of more distant people. What we really are talking about with race are very, very extended families that blend together at the edges. You may notice similar characteristics in your own extended family, certain type of nose or eyebrows, or in my own case, abnormally large craniums run in my family, at least on my father's side. If we want to view race as the intermediate between proximal family relationships and our relatedness by common ancestry, then we need to understand why one extended family would have different attributes from another. We've discarded the idea of natural selection and instead acknowledge that founder effect, drift, and negative selection dominate those small differences that exist between groups. Three, statistical differences are inherently meaningful. I think it's important that we discard the idea that a statistical inequality has any special meaning. If I divide any group of people into any arbitrary groups, there will always be one group with more of some exceptional or unusual trait. The error is assuming that this difference is somehow the result of the nature of the grouping. For example, if I take the US population and defy them into left and right handed, one of those two groups will have a higher income or more criminality. In the former case, it might be right handed people, and we find that left handed people are more likely to end up in jail. We can create from that an entire mythos or narrative of how handedness creates biological tendencies towards social success or failure. But the statistical difference is a false lead. It lacks biological plausibility. I might even find a correlation like the relationship between African Americans and heart disease. Is it then a fair assumption that this correlation between two factors is causative? To do so is to confuse correlation and causation. It could be that African Americans are more likely to eat a high risk diet or to receive poor medical care or the amount of stress they experience in daily life. These are all confounding factors we cannot discard. The issue of race is so bound up in culture, geography, and environment. This doesn't mean that it's impossible for genetic differences to exist between groups. After all, we know of some very well-documented cases where they do. But it means we approach each of these cases with a high degree of skepticism. This isn't about political correctness. It's about our ability to fool ourselves, our tendency to fall victim to our own confirmation bias. Biological race is an idea that dominated human population genetics and anthropology for the last several hundred years. We found all sorts of evidence to prop it up. And it wasn't until the modern era when we were first able to peer into our own genetic history that we began to realize how little these apparent differences were manifested in our DNA. On to myth four, the benefit of racialized medicine. A landmark medicine called BiDil was released in 2004 specifically marketed to be used in African Americans to reduce the risk of congestive heart failure. It was deemed to act on a genetic variant that affected nitrous oxide signaling in the blood vessels. African Americans, as a self-identified group, are more than twice as likely to develop CHF. And they are also 20% more likely than European Americans to possess a high risk marker. The numbers are 35% of African Americans and 15% of European Americans possessing a marker. 20% seems like a pretty small difference to me, on which to base a racialized marketing program for a drug that could still benefit 15% of European patients as well. It also begs the question, if we know the variant responsible for the improved performance of the drug, and if that test exists, why are we guessing using a self-identified race? To say that this drug performs best in black people is to miss the real point. It performs best in people with a particular genetic variation. Race is just a very poor approximation of this probability. The drug will fail to yield additional benefits in 65% of African Americans, a clear majority. This kind of thinking is actually harmful to medical treatment of both African and European Americans, and I'm not the only one who thinks so. Don't confuse a clever marketing ploy with good science or good medicine. So let me return to the question at the beginning of this video. Why are Jamaicans such good sprinters? I have no doubt that in every population, every race or tribe, the potential for great sprinters exists. It is in the interaction between the raw genetic potential and the cultural and environmental factors that champions are forged. Exceptional Jamaican athletes are encouraged to take up sprinting from a young age. If they were encouraged to take up tennis or beach volleyball, if those were a matter of national pride, I have no doubt that their exceptional athletes would perform as well at those sports as they do at sprinting. What about Kenyan and Ethiopian runners? Well, when genetically tested, they turn out to be a little different from non-elite athletes from the same general region. What is explanatory is what altitude they train at, their personal level of commitment, and the number of hours spent training. Their exceptional performance seems to be partly a product of living in the mountains of Kenya, rather than the genetic gifts, or else their non-mountainous neighbors would be just as competitive. Likewise, if Jewish or European or Asian Americans have dominated particular academic pursuits, it could very well be that exceptional intellects from those areas are encouraged to use their intellects in certain ways, are given opportunities denied to others. These are cultural differences acting on biological diversity, not the destiny of a people. If there is any lesson to be taken from these examples, it's that much of our potential as a species is wasted for lack of opportunity. The person with the greatest intellectual potential, the next Stephen Hawking or Einstein, may be alive at this very moment as a young girl living in the Sudan. Will she be able to use her tremendous intellect to solve social problems, to discover a unified theory of physics? Given the environment she's been born into, it's unlikely she'll even be allowed to attend school. How many of these exceptional people never developed their special skills for lack of opportunity, access to resources and equal rights? Racial discrimination is harmful in one simple way. By preventing everyone equal access to the opportunity to be truly exceptional, it limits the total potential of the human species. Thanks for watching.