Adaptive evolution is an ongoing process, and can result in interesting changes  over historical time – and it has, because the human race, most of it,  has been experiencing drastically different environments in recent millennia, environments that are largely the products of our own crazed imaginations.

You’d think that anthropologists, historians, and sociologists would be fascinated by these new vistas opened up by modern genetic technology, but of course they’re not.  Either they pay no attention or they promulgate reassuring nonsense about how people haven’t really changed.  (“They’re still just a bunch of rats”).

The hell with them all.

Now that I’ve gotten that off my chest,  one of those subjects that people ought to be excitedly exploring is caste, the nutty social system in India.

As Wiki puts it, “Indian society has consisted of thousands of endogamous clans and groups called jatis since ancient times.” And we know that: the genetic evidence indicates that many of the major castes have been highly endogamous  for  about three thousand years.

This has implications.  An endogamous group that is exposed to unusual selective pressures – say because of their occupation –  is going to change in response to those particular selective pressures.  That is not the case in a society that mixes fairly freely,  one in which smiths marry farmers’ daughters and the daughters of clerks marry shoemakers.  In that case,  the only selective pressures that result in much change have to affect a major fraction of the population, often everybody – as with adaptation to new foods or new communicable diseases.

If the endogamous group isn’t too big (which often seems to be the case), you’ll see genetic drift. There should be lots of jati-specific genetic diseases in India, if this picture is correct. Which should result in a torrent of interesting medical-genetics papers out of Indian, but I don’t believe that is actually happening.

The third point, which I’ve mentioned before, is that a population made up of a zillion little endogamous groups should slow the spread of selective sweeps.

By contrast, places like Europe (that has had only two small endogamous occupational castes, Jews and Gypsies)  should have no specialized occupational adaptations (except maybe for farming), little drift, and a more rapid spread of selective sweeps.  Selection on standing variation should work at about the same speed in an endogamous or open group,  except that it would respond to general  selective pressures in the open group and caste-specific selective pressures in the Indian-like population.

This entry was posted in Genetics. Bookmark the permalink.

79 Responses to Caste

  1. redzengenoist says:

    If we’d been exploring population structure 100 years ago, I believe we might have discovered a (much less rigid) endogamous group structure in the nobility of Europe.

    The family of Darwin was clearly very distinct from the surrounding population, both in traits, pressures, and breeding pool.

  2. thrasymachus33308 says:

    And yet Darwin and his relatives weren’t nobility, just upper class. The whole English class structure is pretty endogamous I’ll bet.

    • redzengenoist says:


      I suspect that the European class system might have even have been a little bit more efficient at concentrating traits (though not ROH) precisely because the titling was not legally formalized, like the Indian caste system, but a matter of cultural enclavization.

      European Jewry probably worked somewhat in this manner – fuzzy edges, with a not infrequent boiling in and out along the periphery, yet nonetheless a clear understanding on all sides of an in-group and an out-group.

      • gcochran says:

        Not at all. Fuzzy is exactly what they couldn’t be. Even a little inward gene flow – say 3% – stops local adaptation cold.

      • redzengenoist says:

        By definition, the edge is fuzzy, even with something as defined as a legal caste – the question is only of quantifying the flow.

      • highly_adequate says:

        “Not at all. Fuzzy is exactly what they couldn’t be. Even a little inward gene flow – say 3% – stops local adaptation cold.”

        I’m not sure how you’re defining “local adaptation” here, or “inward gene flow”, and maybe I’m missing something here, but on a simple account of how selection occurs on the many metric traits we might care about, such as those for IQ — which certainly plays a major role in differentiating castes — this doesn’t seem to be close to true.

        Suppose 3% of group A with average IQ 100 intermarry with those of group B with IQ average of 115, and the SD of IQ is 15. Then, assuming IQ heritability of .6, the breeder’s equation predicts that the progeny of group B will be only about .6x.03X15=.27 IQ points lower — which, iterating (though with less of a drop each generation), would require probably over a hundred generations, or over 2500 years, to bring the average IQ down to, say, 101 in group B.

        Of course in most situations the reality of the circumstances is even less likely to bring this IQ down. The same selection pressures that put A at a much higher IQ to begin with likely continue to operate, counterbalancing, and in some cases perhaps even overpowering, the downward trend otherwise introduced by the intermarriage.

        And of course also, in the case of castes, the groups might have started out as the same group, and then begun to differentiate in place geographically. In such a case, the relative selection pressures on the group tending higher would not have to be very great in order to overcome at first a fair amount of intermarriage. The amount by which IQ would be reduced by intermarriage at first would be very small, since the average IQs of both groups would be nearly identical. At some point, the effect of intermarriage in dragging down IQ would reach equilibrium with the selection pressures.

        • gcochran9 says:

          You’re missing something. The point is that you start out with IQ = 100 in group A and group B – they are products of fission. Suppose that you have unusually strong selection for some quantitative trait (IQ for example) in group A, while B does not experience this. Let us also suppose that group B is much larger than group A. Even a little bit of gene flow from group B to group A greatly limits the possible change, while gene flow from A to B has almost no effect.

          If S1 is the local selective differential, and h2 is the narrow-sense heritability, the change per generation is R1 = h2*S1 with no gene flow. If the local population experiences a fraction g of inward gene flow per generation with a large outside population not subject to this selection, the change per generation is

          R2 = h2*(1 – g)*S1 + g*(global average – local average)

          The maximum possible difference between the local and global average is then S1*h2*(1-g)/g.

          For example, suppose that a population experienced selective pressures similar to those posited for the Ashkenazim, such that the parents of the next generation averaged 0.5 IQ point higher than their current population average, while also experiencing significant gene flow from the general population. For 10% gene flow and a narrow-sense heritablity of 0.5, the maximum IQ increase over many generations would be 2.25 IQ points. For 20% gene flow, the maximum increase would be 1.0 point. Clearly, any signi cant amount of gene flow greatly inhibits local adaptation.

      • highly_adequate says:

        I meant above,

        The same selection pressures that put B at a much higher IQ to begin with…

        Also, one of the interesting possibilities with castes is that they might reach a kind of “escape velocity” in which, at some point in their differentiation, the average differences are so pronounced that intermarriage plummets, allowing further differentiation.

      • highly_adequate says:

        One further point.

        If, by “gene flow” one means new alleles, that is likely mostly irrelevant to the caste issue (and often irrelevant even to the issue of differentiating populations which are separated geographically), because those differences can easily be based, or mostly based, on standing variation.

      • highly_adequate says:

        Bear with me please, because I’m trying to understand how this works.

        I don’t really get why the formula you’re using is the correct one. In particular, why doesn’t the selection of the smaller group apply to those parents who intermarry, as well as to those who don’t? (I’m assuming that when you talk about gene flow here, you mean the rate of intermarriage.)

        I can see that for those who intermarry the average of their progeny will be lower than the average of those who don’t intermarry because of regression to the lower mean of the larger population due to one member of the couple. But the selection at least presumably applies to them, or might well do so. If, say, higher IQ in a couple means more children in this group, and this is how selection operates, and this is how selection operates in general in that group, then selection certainly seems to apply — though, again, I can see how the simple formula I used needs to be modified to handle the difference in population means in intermarrying couples.

        In general, it does seem to me that the exact manner in which selection operates is pretty important in coming up with a correct model here. Your model seems to assume that couples who intermarry do not have the same sort of selection operating as in the small population, but rather have the selection (or lack thereof) of the larger population (say the number of children does NOT vary with IQ in cases of intermarriage). But that model appears to be very different from one in which the selection operates as it does in the smaller population. Of course, anything in between is possible as well — a comprehensive model should incorporate all of the possibilities, I would think.

        • gcochran9 says:

          I suppose I should include the effects of selection in the admixing group. It’s insignificant, though, because plausible amounts of selection per generation are always small, and the effect you’re talking about multiplies that small amount of selection by the small amount of inward gene flow. A second-order effect.

          My conclusion stands. Dog-breeders and agricultural geneticists understand it. I would guess that the fraction of people in the human scientists who understand the importance of tight genetic isolation for adaptive change is very small, easily bounded in a nutshell.

  3. Pingback: Greg Cochran on Caste

  4. dave chamberlin says:

    Speaking of adaptive evolution being an ongoing process, anybody notice who always wins the 26 mile marathons in international races? Ethiopians. Gee wilikers, what a coincidence, they must train harder. It just so happens that this part of the world has rather loose rules on theft, and the more cows you steal the more wives you can have. But to successfully steal a cow you have to run longer and faster than the guy you are stealing it from. So what happened rather quickly in this scenario is the better runners had more wives and more children. So surprise, surprise, they evolved into a group of fantastic long distance runners. The Swedish national team, the best long distance runners from the whole country, went down to this neck of the woods and got beat by the equivalent of the local high school team. They went home feeling really depressed.

    • Gerard Mason says:

      Hmm… and am I carrying the cow?

      • dave chamberlin says:

        If you were then Ethiopians wouldn’t be winning just the marathon

      • dave chamberlin says:

        I guess the real question you are asking is don’t cows run long distance a whole lot slower than humans and yes indeed they do. But I reckon they do the same thing in Ethiopia to caught cow thieves that they did in the wild west to horse thieves. Kill ’em. So if you run a 6 minute mile your cow runs a 8 minute mile and the tribe you stole said cow from runs a 6:15 mile then you may not get away with said cow but you will another day to try again.

    • bbartlog says:

      You are thinking of the Kenyans, though in any case someone who is interested in racial biology would probably sooner refer to the Kalenjin specifically or the Nilotic people more generally, rather than a nationality. The rest of your ‘just so story’ is completely retarded and presumably reflects your own prejudices regarding Africans and crime. Hint: cows aren’t endurance runners, they would die of overheating. And the old Irish had a long tradition of cattle raiding and somehow didn’t end up with this physiology.

      • dave chamberlin says:

        I’ve been called worse. But what bugs me worse than insults is inaccuracy. Now Mr bbartlog, if the truth is what you truly desire then kindly read “Taboo, Why Black Athletes Dominate Sports and Why We are Afraid to Talk About It.” It is written by not by a racist babbler like me but Jon Entine a highly decorated author. Now I read the book 10 years ago and I’ll bet I got some details wrong. And if you like you can pick on them too, but hey, we are out to prove our little worlds are AOK aren’t we. I’ve met the folks all my life waving their vanity flags like they were leading a parade, funny thing is……no one is following.

      • Toddy Cat says:

        I take it that by “prejudices regarding Africans and crime” you mean ststistics?

      • Hugh Mann says:

        “And the old Irish had a long tradition of cattle raiding and somehow didn’t end up with this physiology”

        Yes, but all those bogs slowed them down, so they couldn’t run – but had to stand and fight. Over the years they’ve become pretty good at it. Same goes for the Highlanders and Borderers, now I think of it.

      • reiner Tor says:

        It also matters a lot how they started out. For example if they had been persistence hunters before (like the Tarahumara of Burn to Run fame), then they were more likely to adopt cattle raiding methods involving endurance running, whereas if they had hunted in another way, they might have adopted other cattle raiding methods.

    • RT says:

      This is slightly off topic, but the post above provides an entry into this question: I just finished reading David Epstein’s “the Sports Gene” (I heard about the book via Steve Sailer’s site) and there are several chapters on African runners. The Swedish team referred to above found that Nilotics don’t have higher VO2 max or more fast twitch muscle fibers than Europeans. The main reason for their success in distance running doesn’t seem to be greater metabolic efficiency, but greater biomechanical efficiency (better running economy) resulting from their physiques.

      West Africans, who dominate the sprints, on the other hand, appear not only to have more efficient physiques for sprinting, but also metabolic advantages – more fast twitch muscle fibers and more efficient anaerobic metabolisms. Epstein writes about the theory proposed by Morrison and Cooper (both Jamaicans) that the West African anaerobic advantage is due to selection for resistance to Malaria – they note that among the adaptations of West Africans to malaria is low hemoglobin levels and that the efforts of aid organizations to supplement West African diets with iron to raise hemoglobin make malaria worse. They posit that West Africans have been selected for a metabolic shift away from aerobic to anaerobic processes resulting in a higher percentage of fast twitch muscle fibers and higher metabolic activity in pathways that rely less on oxygen to create energy. They posit that a side effect of this change is to confer an advantage on West Africans in sports requiring short anaerobic bursts and disadvantages on longer duration endurance activities.

      Are either Greg or Henry familiar with Morrison and Cooper’s work and does either of them have any opinions on its plausibility?

  5. In my experience, Swedes are often depressed. I thought Bergman was a genius until I visited Sweden. He just pointed the camera at them. Running lifts the mood, apparently, though not if you do it competitively against Ethiopians.

  6. says:

    Apaches also had a culture based on cattle and horse stealing. They relied on stealth to get the animals without being noticed and then drove them as fast as possible without stopping to rest. Sometimes a raiding party would go 4-5 days continually without sleep or rest to get back to their encampment. If pursuers caught up with them they would scatter each individual in a different direction to maximize each individuals chance of escape.

  7. dearieme says:

    But cattle and horses are incapable of running for 4-5 days continually without rest or sleep. Aren’t they?

  8. Hallie Scott Kline says:

    That’s my understanding, too. (Thanks to True Grit.)

  9. Greying Wanderer says:

    Just on the point of how little gene flow is needed to mess up this small endogamous group evolution thing – gene flow doesn’t can happen without effecting the endogamous group if certain conditions apply.

    1) Intermarriage is always marrying out from the point of view of the endogamous group i.e. the spouse from the endogomous group lose their in-group status.

    2) Men from the endogamous group have children with women from outside the group but those children aren’t seen as part of the endogamous group.

    Those two conditions seem pretty common. If you take euro aristocracy as a (much less strict) example of this kind of caste behavior then kids being cut off from their family for marrying out is pretty common (at least in fiction) and men having lots of illegitimate children who weren’t considered full members of the endogamous caste is also very common. I’d imagine these conditions are commonly applied in Indian castes also and possibly more strictly?

    So it seems to me you could have 3% gene flow and it not have any effect on the endogamous group if the caste rules about the product of that gene flow are strict enough.

    • gcochran9 says:

      I said “inward gene flow” for a reason. I’m sure that the European aristocracy could have gone in some interesting direction, if had they had just managed to avert their gaze from a tanner’s sexy daughter.

  10. Jim says:

    Dearieme – They wouldn’t be running but walking.

  11. Jim says:

    The Apaches themselves would be walking. It was a long time after the Spanish arrived before they learned to ride horses. They took the cattle and horses to eat. It was their almost only source of meat.

  12. Jim says:

    As fast as possible meant as fast as the Apaches could walk at a sustained pace. But they moved continuously day and night until they arrived in safe territory.

  13. Jim says:

    The Apaches became highly dependent on living off the Spanish. Had the Spanish left the Apaches would have been in deep trouble.

  14. Jim says:

    The Apaches recognized their dependence on the Spanish and had no desire to drive them out. On their cattle and horse stealing raids they made every effort to avoid any fighting with the Spanish. If however an Apache was killed in a raid then it was the duty of his kinsmen to organize a revenge raid not for stealing animals but to kill some member of the group (it didn’t matter who) who had killed their relative.

  15. RS says:

    > The rest of your ‘just so story’ is completely retarded and presumably reflects your own prejudices regarding Africans and crime. Hint: cows aren’t endurance runners, they would die of overheating.

    Not necessarily so dumb, since as he said the long runs may be more about escaping with your hide in case of a raid that has attracted pursuit. At that point the cows are neither here nor there. The running ability may also be equally about pursuing — the other side of the game.

    Now, you might think all of the above applies to every possible form of theft, so what’s it got to do with E-Africans. Well, cows are probably harder to guard than jewels and money. And valuables other than cows, jewels, and money can be pretty hard to move if you’re afoot. Not to mention, some people probably don’t even have anything valuable besides livestock, jewels, and money: except maybe houses and bales of hay, good luck with those.

    > And the old Irish had a long tradition of cattle raiding and somehow didn’t end up with this physiology

    Were they mounted or afoot? If mounted you would have to look at their horses’ endurance performance not their own. They themselves would more aptly be adapted for unarmored strife in the case of their horse taking an arrow in the glute. (You will need to be unarmored if you want to be fast.) Perhaps it’s not a coincidence that they are(?) the largest people in the world. We know that’s not because Ireland is freezing cold. As for gangsterish and pugnacious tendencies they do have a degree of it, maybe less than some people.

    • RS says:

      > Well, cows are probably harder to guard than jewels and money.

      Also hard to put in your pocket, making pursuit more likely.

    • albatross says:

      The just-so story involving cattle raids suffers from the fact that cattle raids were common in lots of places, most of which didn’t end up with huge numbers of natural marathon runners.

      • Richard Sharpe says:

        Sure, but isn’t selection path dependent and contingent?

        And, there might well be different factors in other places, like the ability to appeal to outside players (the law) to deal with cattle thieves.

  16. RS says:

    > My conclusion stands

    Sure, other than the fact that you were talking 3% inflow and are now talking 10%….. also the h^2 is likely above 0.5 for IQ, as you might well agree, though it does seem closer to 0.5-ish for most traits.

    Not that I’m ungrateful for the mathematization you’ve given, which I’d have struggled with and ultimately botched.

  17. kai says:

    indeed, significant divergence is compatible with (very) low gene flow. Not sure about the formula, it seems it assume perfect heritability for the genes coming from outside from the gene flow, while heritability is h2 for the genes submitted to the in-group selection…but I may be wrong. Maybe the equilibrium would be S1*(1-g)/g. But it does not change Greg’s number, as it is obtained relative to generational speed of change (S1*h2).

  18. Greying Wanderer says:

    For anyone who isn’t sure another example of the general idea might be black-white in South Africa or in the US South before desgregation. You could have a lot of gene flow in total but if all inward flow on one side of the line becomes outcast then it wouldn’t effect the sealed-in nature of their evolutionary bubble i.e. if White people could marry out but Black people couldn’t marry in then all the gene flow becomes one way.

    • albatross says:

      I think the HIV rates among black and white South Africans suggest very little causal sex across racial lines. From Wikipedia I see the claim that in 2008, about 13.6% of black South Africans and 0.3% of white South Africans were HIV+.

      Along with that, it seems like any time you have obvious physical markers of interbreeding that will cause you to be cast out (blacks/whites), that can work to block gene flow. But it has to continue to exist for a really long time, right? And even so, you get people with some black ancestry who can pass for white, and who can introduce some genes back into your isolated population.

      Are the castes in India rigid enough to exclude all gene flow? An occasional adoption, cousin from far away where caste rules are looser, or other guy’s baby slipped in while you weren’t watching will give you some genes flowing in.

      • Greying Wanderer says:

        Dunno. I was just trying to use a well-known example to illustrate the idea that you could have gene flow without *inward* gene flow for one of the two groups concerned. I’d guess it would have to be very strict to work long-term.

        Cape Coloreds are another odd example where the result of gene flow is outcast from one of the parent populations and outcast themselves from the other.

        I wonder if there are Indian castes like that?

  19. Jaim Jota says:

    Regarding Greg’s “the tanner’s sexy daughter” (above), he must be talking about Herleva (Arlette), daughter of Fulbert the Furrier, mother of William of Normandy. The rest is (English) history. I’m not sure that the European aristocracy could have gone to a more interesting direction, even if those Norman leaders had they had just managed to avert their gaze from sexy commoner girls. Maybe by “more interesting”, Greg is pointing to the Habsburg King Carlos II of Spain, a degenerate with an enormous misshapen head. His Habsburg jaw stood so much out that his two rows of teeth could not meet; he was unable to chew. His tongue was so large that he was barely able to speak. His intellect was similarly disabled. Muy interesante. Other aristocratic rulers fixed their eyes on close female relatives and avoided those sexy devils, such as the Greeks ruling Egypt, but all ended producing imbeciles.

    • RS says:

      I was reading about Cleo Ptolemy. Hard to know if her physis was stunning or just attractive, but they all say her geist was ravishing. While it could be that politics was one motive of his, she passed muster with J. Gaius Caesar. I was blown away, then, to see how mega inbred she was.

      Maybe her father of record is spurious, or one or both grandfathers — or even a female ancestor.

      • reiner Tor says:

        The only bust of her that is accepted as contemporary shows her to be pretty in a kind of queen-next-door-ish manner. At any rate i wouldn’t call her ugly, even if obviously she wouldn’t qualify for the front page of a magazine.

        On the other hand, she seems to have been remarkably competent both as a ruler and as a seductress, even if eventually she was to fail in both roles.

        I guess her family must have had very good genes to begin with.

      • RS says:

        > I guess her family must have had very good genes to begin with.

        Ah, that didn’t occur to me. Having a high-homozygosity genome won’t be as bad if your line is fairly low in deleterious variants in the first place.

      • RS says:

        Or I should say, (relatively speaking) very low

      • reiner Tor says:

        @RS: I think that being a very successful soldier and warlord in antiquity is a good indicator of low genetic load, since even the job of the warlord is highly physical (he probably needs to do some fighting himself, if for no other reason then to inspire his men, as did Alexander), he needs a lot of intelligence and charisma, as well as physical prowess, probably needs to be taller than average, maybe even being handsome could help. (We tend to vote for better looking candidates in elections, I guess people were more likely to follow the better looking warlord.)

        Possibly not always true, but could often be true. If you have low genetic load, than mega inbreeding might not be worse than any other strategy. (See the recent post on Ottoman fratricide and the line on the useless poet…) At least it worked in the case of the Ptolemaic dynasty, since their last ruler (queen Cleopatra VII) was still competent. (I’m not sure about the rulers in between, unfortunately I know very little about the Hellenistic empires.)

  20. dearieme says:

    With the Habsburgs you’ve moved from discussing aristocracy to royalty.

  21. Jaim Jota says:

    We are discussing if European aristocracy/royalty could have turned into a caste. Apparently, no. I know of no ruling group turning into a durable caste. You know, those tanner’s sexy daughters.

  22. Matt says:

    Re: the gene flow discussion upthread, does pre-selecting the inflow for matching traits (i.e. assortative mating between the selected and unselected group) change the impact of long term flow, so its more compatible with selection within the group?

    I would have thought that it would (although still blunt it, because the selected group would have more of the trait from genes), BUT, that this wouldn’t make any practical difference to the kind of really intensive breeding practiced by artificial selection, because you’d quickly get such a high divergence that there would be no possible trait matching. And not that European ruling classes actually did this.

    In terms of Europeans being subject to general selection, I take general as meaning what works (boosts fitness) on net across all the occupations in society, scaled for their contribution to the population. So if you had a situation where patience was a fitness boosting trait for 20% of the population and neutral for the other 80%, the whole population would get more patient over time. They’d get more patient more slowly than if that 20% were separated off as a population, but on the other hand, selective sweeps for new mutations are faster in panmictic populations.

    • Greying Wanderer says:

      That’s a thought. If the only inward gene flow accepted by a caste was from individuals with high levels of the traits that the caste was itself selecting for? Like a warrior caste that would take people from anywhere but only if they had numerous medals, scars and bullet-holes.

      • reiner Tor says:

        Yes, or a priestly cast would marry intelligent people.

        Jewish rabbis and merchants did marry, although I understand Jews as a whole were a ‘caste’ of sorts in that respect. But in societies where priests are indeed a separate caste, such inward gene flows might not slow down divergent evolution at all. In fact, they might hasten things up, since then there would be a selection against the specific trait in other groups (b/c people with the given trait would simply leave the gene pool of that other group), leading to more not less divergence.

      • Matt says:

        Greg’s equation basically combines what is lost and gained when combining the effects of selection within the group and changes from inflow (which may each may have a positive or negative effect on the trait).

        (There’s also the effect of selection on the incoming group, which gives a slight change, but not very much. Although selection in an incoming group might plausibly be much fiercer than in the local group, who may be more near the local adaptive optimum – e.g. of Chinese and Tibetans who both moved to a slightly lower oxygen environment that Tibet, selection would seem to be likely to be fiercer and the selection differential higher in the Chinese).

        For plausible levels of selection and heritability, the fraction lost or gained due to inflow can quickly balances with negative or positive selection, preventing directional change.
        Obviously, if there is no mean genetic difference in the admixing subset of the admixing population, then nothing is lost due to admixture.

        But finding an admixing subset that would come to a similar genetic mean becomes increasingly difficult as the groups diverge. If there is a 1 SD difference between the populations, then at heritability 0.5, the admixing group needs to be 2 SD different from its general population, a small percentage of the parent population.

        (If heritability is 0.7 and the divergence is only 0.5 SD, then of course, the overlapping subset of the lower population is a bit larger, and the admixing subset only needs to come to 0.7 SD above their population mean. I don’t know if that might fit well with Greg Clark style models of the European upper classes and recruitment from the “lower orders” or whether class mixture really was driven mostly by sexy tanner’s daughters).

        So in your example, the warrior caste’s out-of-caste recruits would probably have to have, at heritability 0.5, around twice the difference in badassness from the general population as the warrior caste, in order to not have at least a slightly negative effect on the warrior caste’s genetic mean. Lots and lots of bulletholes?

        And that preselection problem is larger if there are multiple independent (uncorrelated) large divergent traits between the populations (for instance, not just smarts, but also neuroticism and extraversion).

  23. RS says:

    > The just-so story involving cattle raids suffers from the fact that cattle raids were common in lots of places, most of which didn’t end up with huge numbers of natural marathon runners.

    That’s entirely fair as far is it goes, but you have to ask

    (A) did these pops raid on foot or on horseback. If the latter, you no longer have any question about fleet humans, but a question of fleet vs strong/durable horses. Suppose your horse just got shot. You may feel like running like hell, but that’s a poor idea since your antagonists are still mounted. It’s time for a confrontation — probably one which you are going to lose, unless your bros stop to make a stand with you.

    (B) how much did these various groups raid, at what pop size and for how long? This depends on various things. How exclusively pastoral were these groups, at what pop size, for how long? (Perhaps the Irish have had quite considerable plant culture in addition to their herds — as has generally been the rule for Nordish. If so, they probably stole less cows than pure herders did.) Also, what was these groups’ mating system (and for how long, at what pop size)? Suppose you and I are herders in a rather monogamous society. I might feel a no-joke itch to raid your herd once every ten years or so (and you, mine). Then a new extrinsic factor shows up that causes a considerable increase in polygyny. Let’s say falciparum malaria has spread to our district — or maybe mere vivax malaria — and a lot of people are febrile 30 days a year, while some are not. Now average males aren’t as much desired as before: they get cheated on more, dumped more, ignored in the first place more than used to be the case. After 2,000 years of this, you and I are reincarnated and discover that our raiding & mayhem drives have been cranked up 2x, 5x, 8x. If our society still has few horses and never has had, we may also discover that we are more formidable endurance runners than we used to be.

    • RS says:

      Not that degree of pastorality, degree of theft/violence, and degree of polygyny are necessarily separate. They could all be related. Suppose an economy goes from 33% pastoral to 67%. Cows being stealable, this increase in pastorality drives up theft/violence. The theft steepens the distro of wealth, so polygyny goes up. Then men do more stealing in response to the steepened distro of women. It’s a complex feed-fwd, but eventually it must come to equilibrium because of decreasing marginal utilities, ascending marginal costs, etc.

      • dave chamberlin says:

        I’m going to reread that book I recommended, John Entine’s book “Taboo, Why Black Athletes Dominate Sports and Why We are Afraid to Talk About It.” Maybe I’ll agree with Bart that his ideas regarding long distance running evolving in response to theft on foot of cows are silly. I read it a long time ago and I don’t remember it very well. I try to stay open minded and admit it when I’m wrong.

      • bruce says:

        You’d think lone shepherds guarding their sheep would have some incentive to go boot-scooting when giant bands of sheep-stealing sumbitches attack. A marathon run home to where their warrior band of avengers waits- well, more probable than some guy running away with a large herbivore under his arm, isn’t it?

        Just so. I don’t believe it, exactly- some mix of genes rewarding being a tall skinny guy with large lung capacity in East Africa -Africa is a high continent, lots of chances to evolve lung capacity- just seems more likely.

        As to polygyny- I write from Peoria. We call that ‘getting a lot’. We all try for it.

  24. nickWH says:

    Probably Ethiopeans were already adapted for high endurance during their pre-pastoral evolution as hunters — so that they could over the long run outrun their prey. Many or most other human hunters quite likely had this high endurance adaptation too. But it was presumably costly, e.g. in terms of requiring greater stores of ready ATP, or a molecule that can be quickly converted to ATP, so that most populations that went through substantial adaptation to agriculture lost it.

    Ethiopian pastoralism, however, may not have exerted sufficient selection pressure in the other direction, or that selection operated over a short period of time, or (most likely) the population of the herders was too low for the mutation eliminating the (presumably costly) adaptation for high endurance to appear, or some combination of these causes.

  25. Greying Wanderer says:

    Off-topic, reading through old posts i was reminded of this link

    “stroke turned me gay”

    which is similar to this story

    “Rare condition makes James Bond theme music ‘orgasmic’ for man who suffered stroke”

    and it strikes me that the idea of not a gay bug exactly but a bug that damages the central wiring leading to repairs which can go wrong in many different ways could explain not just homosexuality but paraphilia generally and possibly pedofilia as well – which causes the most harm.

  26. Curt says:

    I have a quick question concerning Jewish intellectual abilities. Should they have higher standard deviations?

    • The fourth doorman of the apocalypse says:

      I don’t know, but if you take a model that says that there are, say, 10,000 genes involved in IQ (I just pulled that number out of my ass, but Npq and the measured variance suggests to me that there are at least 900 genes involved in IQ), then look at it this way.

      Some populations are fixed at some locations, either on the lower IQ contributing variant/variants or the higher IQ contributing variant/variants.

      Thus, the effective N in Npq is smaller than in the general population.

      This suggests to me that the variance is lower in both low IQ subgroups (and there appears to be evidence for that among AfAms) and high IQ subgroups. However, I am unaware of any evidence for the variance among various high IQ subgroups.

      Of course, I could be full of shit or yanking your chain.

  27. Pingback: linkfest: 08/18/13 | hbd* chick

  28. Steve Sailer says:

    “places like Europe (that has had only two small endogamous occupational castes, Jews and Gypsies)”

    Does anybody know what’s the story behind the black blacksmith castes of North Africa and the Middle East? Why in more than a few white Muslim cultures was it traditional for sub-Saharans to be the blacksmiths? In Europe, it seems kind of the modal non-farming occupation (thus all the people named “Smith”), while in nearby cultures it’s often reserved for blacks. But it doesn’t seem like the worst job, so I don’t get why Arabs often leave it for blacks.

    • Greying Wanderer says:

      Judging by this

      it looks like maybe a mixture of happenstance and superstition i.e. if ironworking took off in West Africa first and these blacksmith castes formed (possibly similar to the specifically stonemason caste aspect of freemasonry i.e. a top-end craft guild turned endogamous) and then spread eastwards as an endogamous caste then the locals wouldn’t ever get to learn how to replace them – which is the idea i guess. Judging by some of the other points in that wiki page i wouldn’t be surprised if the blacksmith castes spread the idea that there was witchcraft involved to reinforce their cartel.

      Just a guess though.

  29. Steve Sailer says:

    “places like Europe (that has had only two small endogamous occupational castes, Jews and Gypsies)” — Were there any other castes in Europe?

    I can only think of Irish Tinkers / Travelers, who fill a Gypsy like role where Gypsies are rare. But I have to imagine there were other caste-like phenomenon coming into and out of existence. In 19th and 20th Century England, for example, Advanced Intellectuals (Darwins, Keyneses, Arnolds, Huxleys, etc.) tended to marry within their own ranks. Galtonism was an intellectualization of this impulse.

    • reiner Tor says:

      I would think the English intellectuals lasted a few generations at most. That’s way too short.

    • Greying Wanderer says:

      I think there may have been a narrow class niche which was similar to a caste for a few hundred years made up of the top of the “middling folk” tree i.e. preachers, stonemasons, yeoman farmers etc – similar to the way skilled trades at the top of the bluecollar tree today tend to marry within that segment. If the euro late marriage model worked as truncation selection then i wonder how long it would have taken for a sub-population who practised it most religiously to shed a great deal of genetic load?

      It’s just a pure guess though that the people you mention (and the Washingtons, Franklins, Jeffersons etc) came out of that semi-caste of top-end middling occupations (just below the hereditary aristocracy) that had been heavily competed over for centuries. It’s just that idea combined with the potential load shedding aspect of the north euro marriage model (if true) seems to me like it could create an explosion of brains. Proving it would probably take looking back through all their family trees and counting the occupations but i’m too lazy for that so it’ll stay a pet theory for now.

      I don’t really know how big a deal this genetic load thing is but assuming it’s very big then the north euro marriage model may have come unstuck as the average age of puberty dropped from the 1840s onwards (according to wiki) which would give a possible explanation why that explosive flowering from c. 1640s to 1840s didn’t carry on to produce steam-powered moon rockets.

    • Matt says:

      The Irish Travellers are one, although rare compared to Romani. The genetic studies indicate a thousand year isolation, which I thik would qualifies as long enough for adaptive evolution (or lack of adaptive evolution) to have a role explain them. They have a similar mean level of cognitive ability and performance for criminality and exploitative attitudes toward outsiders.

      “Advanced intellectuals” seem more like recent strong assortative mating of lots of outliers on the trait from a generally selected population more than long term occupational selection acting on a population of typical individuals.

      Not that recent strong assortative mating from a long term generalised mating would necessarily we “worse” for generating an intellectual caste (or extremely high ability intellectual outliers) than long term structure would (depending on the various parametrical possibilities of population size, drift, selective sweeps and selective forces in the two situtations)?

  30. RS says:

    > Just so. I don’t believe it, exactly- some mix of genes rewarding being a tall skinny guy with large lung capacity in East Africa -Africa is a high continent, lots of chances to evolve lung capacity- just seems more likely.

    I was assuming the person above was right: about Swedes finding that E-Africans high school champs who brushed off the Swedish national team were superior in build — not in cardiovasculo-pulmonary power, nor in (percent?) mass of endurance-type muscle. That could perhaps be wrong, or just somewhere short of categorically true.

    Anyhow, I don’t bet the farm on the Entine idea. I haven’t studied it. I only assert that prima facie it looks weighty, not idiotic.

  31. Greying Wanderer says:

    More on the black blacksmith castes in Africa.


    According to this (the section on copper)

    The Copper Age in Egypt was as advanced as a layman might imagine it to be with the copper sourced from Sinai and Cyprus. However one interesting point from


    “At the site, he found slag from ancient copper production. The slag had been heated so hot, it vitrified. To make fires hot enough to vitrify copper slag would have required vast amounts of wood. In fact, Lehner found that the floors of the site were covered with very deep, thick velvet carbon ash from the burnt wood. Lehner said that the cutting of wood to produce such huge amounts of ash contributed to the deforestation of Egypt at the time when the pyramids were being built.”

    Going back to the first link (the section on bronze this time)

    the Sinai copper didn’t contain tin so that retarded the Egyptian Bronze Age somewhat leading to importing Bronze from Syria.

    Then lastly (the section on iron)

    “Iron is a very common element and iron ores occur in the mountainous areas of the eastern desert and Sinai [6] though high grade ores are rare. That and the lack of hardwood or coal needed to achieve high temperatures prevented any large scale iron production in Egypt. Much of the iron used for the manufacture of tools and weapons was therefore imported.”

    I think all together that gives a plausible explanation. Although generally you might expect the more advanced technologies to mostly flow outwards from the middle and near east (simply because they had a head-start) this general pattern was likely modified by the availabilty of ores.

    So Egypt had a strong start with copper (possibly cutting all their trees down in the process) slowed in the Bronze Age through lack of tim and stalled in the Iron age through lack of high quality ore or lack of wood or lack of bronze expertise.

    Whereas West Africa may not have had a copper or bronze age (or maybe it did but on a smaller scale) but either way by the time it was iron’s turn they had plenty of iron and plenty of wood leading to these endogamous iron-working castes drifting west to east through supply and demand instead of the expected east to west – and retaining their monopoly to this day through being endogamous.


    On a separate note these castes make me wonder about the Bell Beakers who i’ve read are seen by some people as an intrusive, possibly endogamous population but not a conquering population which might tie in to the same idea of a metal-working caste.

    As there are two groups of Bell Beaker, maritime and continental, and two main sources of copper in Copper Age Europe, Iberia and the Balkans, that might fit. Although if the two forms of Bell Beaker had two separate sources then why would the beakers have the same form? One possibility would be the Iberian source ultimately derived from the Balkan source (via med islands with copper?). Another possibility would be the shape of their beakers had something to do with copper-working?

    This idea of an endogamous caste of copper-workers spreading out along the trade routes from the two main copper source regions leads to a possible mythological twist vis a vis the “dwarves” in Norse mythology.

  32. Jim says:

    The source of the tin used in the Middle East bronze age is obscure as there are few sources of this metal there. Egypt is one of the few places in the Middle East with deposits of tin ore but these deposits may not have been known back then. It is likely that much of the tin used in the Ancient Near East came from Europe, from Bohemia and perhaps even from Britain. Anatolia seems to have played an important role as a middleman in this trade.

    • Greying Wanderer says:

      You could make Bronze with arsenic too apparently but it wasn’t as good (apparently). It’s an interesting thought that the rise in relative Greek power relative to other regions in the eastern med may have had something to do with simply having better access to the the raw materials for bronze – trade with Atlantis (aka Southern Portugal and Cornwall) ftw?

  33. Jim says:

    Cyprus was the main source of Bronze age copper.

    • Greying Wanderer says:

      In the eastern med maybe but elsewhere? If there were also sources in the Balkans and Iberia and the Greeks had first access to those sources (via Minoans?) maybe that gave them an edge similar to the La Tene / Hallstatt Celts in the Iron Age sitting on all that Bohemian iron?

  34. j mct says:

    Wouldn’t a present day ‘caste’ like Jews and Gysies in Europe that one could see right now in the good old USA be the Amish? They would seem to be very much like European Jews in the Middle Ages, very little to no one converting in from the outside, one can leave and join the greater society anytime one likes if one so wishes, social practices that lead to social isolation over and above any sort of ideological reason not to mix (like keeping kosher, cannot feast with the gentiles), having a govt that is sort of a theocracy in that the clergy are the leaders of the community, having religion rather than being plundered by a common taxman as the basis of one’s primary tribal/societal membership, entirely in the case of the Amish, in part for medieval Jews and so on. The Amish don’t do much usury though.

    The Amish way of earning a living is now out of sync with the rest of American society for about 100 years now, or about 5 generations. So I guess that might not be enough time for anything noticeable to happen as far as evolving . Would that be that right?

    • reiner Tor says:

      Of course. Because they started from a slightly different and quite small founder population to begin with, and had separated their gene pool from the rest centuries ago, we could probably see some difference due to drift.
      One selection pressure that was clearly there was ethnocentrism. (If we consider the Amish an ethnic group, and I can certainly see no reason why we can’t.) Those who were predisposed to defect did so, so probably they are selected for ethnocentrism or at least for tribalism or “clanocentrism” or something like that. That’s probably the case with all of these de facto castes in a non-caste society where desertion is easy.
      In the book Ethnic Conflict and Indoctrination (ed.: Eibl-Eibesfeldt, Frank Salter) I read a survey on ethnocentrism where Jews scored significantly higher than other whites (and WASPs scored the lowest), Jews had the same score as blacks. If I recall the numbers, out of 3 maximum (and basically 1 minimum, 1 means that you are only ethnocentric if choosing your ethny costs you nothing at all) WASPs scored around 1.2, other whites and Italians around 1.35, and blacks and Jews over 1.5. I guess if desertion was easy, they were selected for that.

    • bbartlog says:

      They have some diseases of inbreeding, e.g. . Less than half of them make their living by farming these days, though. Farmland has become too expensive for them to continue expanding as they have been, even given a willingness to work extremely hard to sustain a 19th century lifestyle.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s