PRS and asabiya

There are all sorts of interesting possibilities generated by polygenic risk scores that I have not yet seen discussed anywhere.

There are existing tests  that you can’t fake ( without explicit cheating).  You could pretend to be dumber than you actually are on an IQ test, but it would quite difficult to score higher than you can.  Personality tests – well, you can always lie.  I’ve even heard that people lie on their resume!

As long as you administer the test yourself,  there’s no way for an applicant to cheat on a polygenic risk score.  It only gives you  statistical information, not incredibly strong prediction at the individual level, but it can’ be faked. sooo.

If you hired a CFO with a 4-std score on  honesty, he’d be somewhat  less likely to cook the books.  Although a given  company might not want that – might want the opposite.

You could pick boomer captains that were considerably less likely than most people to go insane. That could be a good thing.

You could hire a whole organization with  > 1-std asabiya PRS.  You’d use other indicators of asabiya, and you probably would consider other factors as well  – but you’d get less internal conflict, have fewer self-dealers, etc. And you know what unusually high-asabiya orgs talk about at the water cooler – the same thing they talk about every night.




This entry was posted in Uncategorized. Bookmark the permalink.

73 Responses to PRS and asabiya

  1. Anonymous says:

    I wonder how heritable asabiyya is. The usual formulation starts with high asabiyya nomads conquering a civilisation, and ends with their low asabiyya descendants being conquered by the next band of nomads.

    Perhaps nomadic life selects for asabiyya and civilisation selects against it, or perhaps the high asabiyya conqueror DNA gets diluted over the generations. Maybe Reich’s lab can take a look.

    • JP says:

      That certainly seems to be the way it works. I’ve noted that populations whose genomes have been under grain-based agricultural selection the longest, tend to be the most two-faced and self-dealing, which seems to work for them. The Great Grandsons of Arias might have had the goods to produce the capitalistic economy, but they certainly are ill equipped to navigate it over the long term.

    • Esso says:

      Ibn Khaldun talks about three generations. At that time scale it must be about family size and cultural differences between nomads and urbanites. City dwellers have fewer brothers and cousins to look out for them and have been brought up to be compliant and courteous. Nomadic life on the other hand can accommodate more conflict, also in upbringing.

      This is not to say that there isn’t a slower hereditary phenomenon with similar effects.

      • Wency says:

        I’d think 3 generations is enough time for a mostly genetic model if you assume that the conquerors are mostly taking wives from the sedentary native population immediately after the conquest, but their children are brought up into the father’s culture. Not sure how often this scenario applies. The Yuan Dynasty Emperors seem to have all had Mongol mothers, but not sure about all the lesser sons of that dynasty, for example.

        • albatross says:

          A natural guess is that we have programs available for both high- and low-trust societies/environments, and our environment determines which programs get run.

      • ghazisiz says:

        Alfred Marshall wrote about how family firms rarely last more than two generations — the founder has a special energy lacking in his sons, who dissipate their inheritance. A study I read about the Hanseatic League noted that very few merchant houses lasted as long as three generations. Vilfredo Pareto’s concept of the “circulation of the elite” acknowledges this phenomenon — there is always an elite, but in a generation or two it hands off its status to a new family.

        There does seem to be an environmental element here — the hunger of a newcomer for status, leading to extraordinary focus and effort, while those born to status take it for granted, and hang out at the Drones Club.

        But maybe there is something about the constellation of effects in a polygenic trait — that it isn’t simply the number of good alleles, but their interaction, so that sons are not likely to have the same lucky mix of alleles that led to the tremendous success of their father.

        • gcochran9 says:

          Do you think the father often marries a female industrialist?

          • ghazisiz says:

            Wealth usually marries beauty, but, more often than not, beauty of the right social class.

            But, to acknowledge your point, an exception to the Marshall-Pareto rule would be the Rothschilds. And they practiced cousin marriage.

          • Steven C. says:

            There wouldn’t be enough female industrialists; he would have to marry the daughter of an industrialist. But his children might marry solely for beauty and a pleasant personality, unless the father can force them into arranged marriages the same way that European royalty and aristocracy used to.

        • Rich Rostrom says:

          There is an old Lancashire saying: “Clogs to clogs in three generations.” (Clogs meaning the cheap wooden footwear of the working class.)

    • Johnny Caustic says:

      There’s evidence that every behavioral trait is heritable (averaging around 50%?), so asabiyya almost certainly is too. But everything we know about the arcs of rising and falling civilizations suggest that the expression of those genes is probably subject to environmental influences. I would guess that prolonged hunger and severe societal stresses like war would be two of the triggers for certain genes that influence behavior.

      I don’t think these arcs are caused by selection–I think the preponderance of these genes doesn’t change that fast. I think it’s epigenetics.

      (However, the long-term agriculturalist vs. pastoralist divide surely does involve some selection.)

  2. gyddyn says:

    How can I do “You could hire a whole organization with > 1-std asabiya PRS.”? Test for common “culture”/”way of thinking”/”clan”?

  3. marcel proust says:

    Thinking along the lines of Haldane’s observation, wouldn’t much of the genetic basis for asabiya (“Asabiyyah”?) be not so much an individual trait but a reflection of the genetic similarity or homogeneity of the group? High asabiya as due in part to inbreeding or a genetic bottleneck? So, test individuals for it but then ensure that the individuals who test high also resemble each other in many other ways (genetically) before trying to form an organization from them.

    • gcochran9 says:

      That’s not how it works.

    • The derivation of Hamilton’s rule, c<br, depends on some assumptions. Break the assumptions and the rule may not follow.

      (1) One assumption is you don’t have complicated, non-linear game theory type interactions between individuals. If all you know is that Bill is your brother, then there’s a 50% chance he’s got a gene you’ve got (over and above the chance he has the gene just by being part of the general population), and you can use that number in Hamilton’s rule to figure out how nice to be to Bill. But if you know not only that Bill is your brother, but also that he likes to play strategy 3 when you play strategy 2, then you have more information about him, and you may have to revise the 50% estimate when figuring how altruistic to be toward him. It can get even more complicated if third parties (mutual relations) care about how you treat Bill. This might matter in the case of human kinship as studied by anthropologists, where how you treat one relative is often a function of how it affects your reputation with other relatives. I have shown that in this case, kin altruism can be amplified.

      (2) Another assumption is you don’t have major changes in gene frequency due to selection between the time a gene is present in a common ancestor and the time it acts in descendants. This assumption is fairly safe in families and maybe small lineages, because selection doesn’t change gene frequencies much in a few generations. The assumption isn’t safe for larger tribes, let alone ethnic groups and races. Henry Harpending thought you could predict something about racial altruism by converting FSTs for large groups into Hamilton’s r’s. The calculation is straightforward, but the prediction doesn’t follow, for the reason given above. (This isn’t just my opinion.) There might be ways to get something like ethnic nepotism via ethnic group selection, but that’s an iffier proposition. It doesn’t just fall out of Hamilton’s rule. Race is not just family writ large.
      Some further comments, and links to an article:

      • gcochran9 says:

        Always seemed to me that the big factor would be punishment for non-cooperation. But there’s another, crude, way of looking at it, at least for humans: do we see anything that looks like a strong genetic tendency to cooperation within a tribe, race, whatever? I don’t. People turn Turk, are Medizers, Quislings, Hungarians help build guns to knock down the walls of Constantinople. Malinche. Tlaxscalans with the Spanish against the Aztecs (surely the Tlax should swallowed their irritation over being sacrificed and eaten, in the name of La Raza? ). Or you might notice all the white people talking about how awful white people are.

        Much harder ( though possible) to get people to turn on their families: even Stalin is said to have despised Pavlik Morozov

        • Bob says:

          What about Moses, Joseph, or Arminius?

        • Zenit says:

          Yes, Stalin was known as a family man.
          Anyways, the case of Morozov was not about ideology, but ordinary family revenge. Morozov’s father left his wife for younger woman, and Pavlik lived with his mother, whose idea it was, according to modern research.

        • RCB says:

          National anthems make people cry. I do think there is something there. But logarithmichistory is absolutely right that kin selection simply cannot favor national-level altruism at equilibrium. One hypothesis is that nations that were good at appealing to our family-level psychology are the ones that got people to go fight wars for them, etc. (Not my hypothesis.)

          • Bob says:

            Larger groups like nations and religions definitely appeal to or exploit our family-level psychology. The nation is the “fatherland” or “motherland”, the nation is conceived as a larger family, etc. Christians are “brothers and sisters in Christ”. Muslims have the “Muslim Brotherhood” and call fellow Muslims brothers and sisters.

            The mechanism may be individual selection on people who are adept at appealing to or exploiting family level psychology and then becoming leaders of larger groups. Leaders of groups generally benefit disproportionately and there can be big payoffs to the leaders from say conquest. There may be a Red Queen scenario involved with an arms race between the exploiters and exploited.

          • Woof says:

            Good organizations mimic family organizations. The best armies have family type squads, extended family type platoons and companies and clan type battalions. They are cohesive and loyal to each other. It’s only when you go beyond this that artificial means are needed to achieve cohesion, like regimental cap badges, brigade and divisional patches and uniforms.

      • “With but very few exceptions, authorities have shied from describing the nation as a kinship group and have usually explicitly denied any kinship basis to it. These denials are customarily supported by data showing that most nations do in fact contain several genetic strains. But . . . it is not what is but what people perceive as is which influences attitudes and behavior. And a subconscious belief in the group’s separate origin and evolution is an important ingredient of national psychology. In ignoring or denying the sense of kinship that infuses the nation, scholars have been blind to that which has been thoroughly apparent to nationalist leaders. In sharpest contrast with most academic analysts of nationalism, those who have successfully mobilized nations have understood that at the core of ethnopsychology is the sense of shared blood, and they have not hesitated to appeal to it.”
        Walker Connor. Ethnonationalism: The Quest for Understanding. 1994, p197

  4. Jacob says:

    It’d be more accurate with whole exome sequencing, more accurate still with whole genome.

    When do you guess it’d be feasible, if someone were actually working on it now?

  5. JP says:

    I’d use these scores to find tall basketball players for NBA teams.

    • Leonard says:

      You joke, but testing children for predicted adult height could be very useful for them and to the sports world. Not to mention strength.

      • dearieme says:

        The leading footballer (soccerballer) Messi was identified young and then given human growth hormone.

        • David Chamberlin says:

          Messi was prescribed HGH very legitimately because he was diagnosed with Human Growth Disorder, people with HGD are not only very short like Messi was going to be but have all kinds of other physical problems caused by the body not growing like it would normally.

          On the subject of HGH prescribed for the express purpose of making a good athlete a successful pro one. It is going to happen if it hasn’t happened already, but that’s another subject for another day and points directly at the NBA.

        • Esso says:

          Just looking at him, Messi has a very low center of gravity. That allows very quick changes of direction and overall agility. And he can use that potential for extreme movements with his beefy high-torque legs. Normal players get clumsier in puberty (change of scale), whereas Messi can probably use his childhood intuitions, with the difference that he is stronger and shoots faster.

          He wouldn’t be the best if he was normal.

  6. JayMan says:

    Here’s the kicker: a world where genetic screening is a routine condition of employment would have test results that pick up different things (at least somewhat so) than said test would in a world that didn’t have such practices. People’s behavior would be affected as they negotiate the world that screens your genes. Not that tests would be rendered meaningless – far from it – but they may need to be recalibrated once and awhile if implemented this way.

    (Yes, you read right, JayMan made an “environmental” argument. I’ve long said technology is the big factor driving short run social change).

    • Phille says:

      “I’ve long said technology is the big factor driving short run social change”

      Along the same line, fertility is pretty heritable and must currently strongly be selected for. But if the environment keeps changing, what does that heritability even mean? Maybe fertility in a society where women are starting to participate in the labour market is something quite different from fertility in a society with internet and smartphones and tinder. And it will be something else again in the near future.

      As soon as the environment stabilises the heritability of fertility should kick fertility up quite noticeably every generation. There doesn’t seem to be a sign of that yet. And maybe cultural selection will turn the tide much earlier.

    • Zenit says:

      In a world, where blood and genes decide all, there would be no such thing as “employment” in our sense.
      In this world, there will be no place for such nonsense as “hiring” “firing”, “building a career” or “following your dreams”.
      In this world, you you will know, and everyone will know, your future, your fate and your destiny written in your genes. Since you were born you will know whether you are Priest, born to teach and advise, Soldier born to fight and rule, Merchant born to trade, or Laborer born to serve, Since you were born, you will learn and prepare to fulfill your duty.
      So says both the most modern science and the most ancient religion.

  7. Cloudswrest says:

    “If you hired a CFO with a 4-std score on honesty, he’d be somewhat less likely to cook the books. Although a given company might not want that – might want the opposite.”

    They may want PRS for taqiyya instead.

    • David Chamberlin says:

      Tests, lies and Cochran smirking somewhere….

      “You say taqiyya, I say asibiyya, let’s call the whole thing off.” It’s a terrible song and this is a worse analogy. Yea well a CFO that has saint like honesty is going to still be fired multiple times until he picks a job outside of corporate land. Saints never get as far as the corner office, they can share a cubicle now and then but they never get farther.

  8. Eugine Nier says:

    The question is which group is the candidate’s asabiya directed towards?

  9. David Chamberlin says:

    a hundred Cochrans are going to come and go
    time is going to roll on
    we will throw up our hands in despair
    and raise them in thanks

    ain’t shit but the wallpaper gonna change
    a thousand clouds deemed crucial will blow away
    and then we reach in to ourselves
    tweaking the wrenches the right way
    till DNA is tweaked and evolution works

    suddenly as if we were stupid and didn’t expect it
    it works
    and evolution works again
    only this time by design and not by random misery

    • David Chamberlin says:

      Not a poem worth rewriting and polishing up. But as poetry usually is, it isn’t clear. My guess is we won’t have genetic engineering that has the impact of increasing human intelligence significantly until the end of this century and even then it will take a world wide effort of hundreds of brilliant scientists and tens of thousands of normal folk. It might be a simpler problem to solve than I suspect, for instance if intelligence was simply the result of genes that positively impact intelligence and genes that are deleterious but I’ll bet that it’s not that simple. Everybody gets wrapped up in the drama of the moment and very very rarely does anybody look at the big picture. Three fundamental revolutions in human evolution down, one to go.
      1) modern inventiveness 50,000n years ago
      2) agricultural revolution 10,000 years ago
      3) industrial revolution 1850, 168 years ago
      4) genetic engineering of higher intelligence kickstarting late in this century

      thanks for putting up with me Greg

  10. +4SD honesty scores are considered “lying” on the MMPI. They have specific questions (in a pattern!) like “Would you attend a movie without paying if you could be sure you wouldn’t get caught?” They expect a mix, that people would not do most of them even if they know they wouldn’t get caught. But no one answers them all in the Completely Upright category. Everyone has a place or two where their honesty wanes. Those who claim it are putting on the shine. It is important to note that this is not merely theory, but empirical results compared back to actual behavior. They knew from other sources that Jack, who went 30-for-30 on the honesty scale, had done two years for armed robbery.

  11. Maciano says:

    Maybe I’m thinking out loud too much, but would a company which selects for people who physically look alike have higher levels of asabiyah also?

    In high school I was always struck how much kids self-select in groups of similar looks, age, status, etc.

  12. J says:

    Since there is no rule that comments have to be topical, wish to ask the host’s opinion on the evolutionary meaning of Dr Christine Blasey Ford’s trauma caused by a very short and a very incompetent apparent sexual attack. I was impressed by her testimony. Yet fertile females of all mammals are constantly exposed to male interest, and evolution has equipped them with defenses, like instant vagina lubrication, etc. thus the “attack” almost always ends without physical injury. Evolution should have equipped the female with mental defenses against these common life events and protected her against lifelong, paralyzing agony and trauma. Where did evolution fail with female WASPs?

    • gcochran9 says:

      There is such a rule, but it is very loosely enforced.

      There’s enough political excitement that you’d expect a few people to exceed the threshold and make stuff up: several obviously have. Before we’re done, I fully expect someone to claim that Kavanaugh cooked her in a microwave.

    • dearieme says:

      I liked the bit where the accuser said that after the attack she made her way home by hitching a lift on a drone.

  13. RCB says:

    So what we care about here is phenotype. Which means measuring the phenotype is generally better than measuring the genotype underlying it, if you can do it accurately. The argument here is that the phenotype in question (personality, integrity, etc.) can be easier to fake than a genotype score, so the latter might actually be more informative, even given the inherent loss when inferring phenotype from genotype. Maybe if we get to the point where the majority of the variance can actually be predicted by genetic information, then that will be true. Not yet, though, for any trait. Psychopaths aside, I still feel like you can get a good feel for a person if you put them through a stressful interview – hard to fake “cool under pressure.”

  14. To construct a polygenic score, you first need a good way to measure the trait itself, but how do you measure honesty? You could imagine an experimental setup with the test subjects left in a room to find someone’s “missing” wallet. You’d need a lot of test subjects, though, and it would be ruined if word of the experiment got out, as if too many people could see through it it would just reduce to measuring intelligence. Doing it on kids would be easier,(though harder to get past the IRBs) as they’d be less likely to see through the experiment. You could even imagine polygenic scores constructed from the results of testing on children, and then applied to the population as a whole. Instead of having to wait for children to grow up to determine if the marshmallow test correlates with income, you’d construct a marshmallow test polygenic score and then see if it correlates with the permanence of adults when other polygenic scores are controlled for.(Not saying this will actually work.)

    Initially it will be much easier to construct polygenic scores from demographic variables, avoiding the need for anything more than genetic sampling. Has anyone tried to construct a polygenic score for divorce? I’m sure the data to do so is out there. Similarly, if you work with a background check company you could create a polygenic score for the probability of getting arrested. In the farther future, if you had access to the genetic information of millions of parents and their children, you could use a simple paternity test to find out which kids weren’t really fathered by the putative father, and thus construct polygenic scores for being a cuck and cucking your husband.

  15. Rich Rostrom says:

    Hiring for high asabiya works only if those hired regard the workplace as the target of their loyalty. Otherwise the enterprise will be staffed by people for whom advancing the interests of their relatives, or co-religionists, or co-ethnics, outweighs serving the interests of the enterprise.

  16. enkypala says:

    And here is what you get when you assemble these together:

    • random observer says:

      A parsing/fisking of that would not go amiss, should anyone care to undertake it or know of one that already exists.

  17. R. says:

    Anyone more knows about the claim that the eternally falling sperm counts keep decreasing due to epigenetic effects somehow translating the decreased count to next generation, which then starts from an even lower start and so on?

    It’ll be interesting to see where this ends up, if it keeps up a few generations from here there won’t be a need for male birth control.

  18. We sleep much less. That may account for half of the decrease all by itself.

  19. Warren Notes says:

    Just for the record, the law that would have to be changed to implement what’s suggested in the post is Title II of the Genetic Nondiscrimination Act of 2008 (GINA), which prohibits the use of genetic information in making employment decisions.

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