Some thoughts on the newfangled eugenics stuff

I decided that there was an unelaborated side to this recent eugenics debate that’s been cropping up on the twittersphere. As a brief summary, a few people (@matingmind [Geoffrey Miller] and @razibkhan) have been defending recent work on “China engineering genius babies”. Vice might have overstepped a bit with the phrasing—no surprise there—but reading that interview with @matingmind leaves a pretty bad taste in one’s mouth. I don’t want this to be a personal attack, so I’ll leave the immodesty stuff out of this, but I have some nits to pick.

To start off, Kevin Mitchell already has a great post on the underlying genetics of trying to select for ‘intelligence’. If you haven’t read it, you should probably go read that first (here’s another good post from Ed Yong). He describes how traits—particularly advanced cognitive ones—are likely influenced by thousands of mutations. Some of these could be quite rare, making the basic task of selecting for any of them Herculean. In response, Razib Khan did admit this by saying “It makes sense to be skeptical of the scientific possibilities in the near to medium term.”

For my first point, I wanted to add something from the scientific perspective: I’ve recently had my eyes opened on the nature vs. nurture stuff. Sure, we’ve all had it beaten into our heads that it’s not nature vs. nurture, it’s nature AND nurture and different proportions of each. But here’s the thing: there’s a (sometimes HUGE) third component. As explained by Eric Turkheimer in this fascinating paper, the ‘nonshared environment’ of an organism will have a large impact on how it turns out. Here’s a key snippet:

“Why don’t identical twins raised in the same family have identical outcomes? There are two important reasons. The first is measurement error. The second is the self-determinative ability of humans to chart a course for their own lives, constrained but not determined by the genes, family, and culture, and in response to the vagaries of environmental experience with which they are presented. The nonshared environment, in a phrase, is free will.”

And the more complicated a trait is, the more this nonshared environment variable is going to come into play. Your nonshared environment isn’t going to influence your height much. But it has the potential to have a large impact on your personality. You can certainly imagine a person that has a bunch of fluky things go wrong turning into an embittered individual. This isn’t just conjecture anymore either, as there’s a brand new study by Gerd Kempermann’s group that quantifies this in genetically identical mice that grow up in the same place (hat tip to Virginia Hughes for the explanation). In short, stochasticity plays a role in who you are. From the quantum mechanical level on up there is a lot of randomness there to screw up the best laid genes of mice and men. You might still call this ‘nurture’, but it’s nurture at a different level. Or you might just call it ‘shit happens’.

So, bringing this back to the eugenics debate, it’s not just hard to measure an advanced trait like intelligence—even if you believed in one test of it—but there are so many things that can go wrong even if you COULD select for some kind of ‘smart’ gene.

This gets at my second point: how I interpret ‘intelligence’ as a neuroscientist, and why trying to ‘improve’ humans is totally misguided.

I’m probably being simplistic, but eugenics made sense when I was 12. Back then I thought intelligence was a quantifiable thing. Some people were really smart and hey there are tests to prove it! But, as anyone who reads about these kind of things knows, it is not so simple. There are so many environmental factors and, personally, I now know how lucky I had it to grow up a white male in America. This of course leads to how intelligence testing itself is flawed. And if you don’t agree with the comic, here’s a great post from Frans de Waal about how even our studies of animal intelligence have been misguided, and their capabilities are a lot more nuanced than we realized.

Again, I don’t want to get into the Howard Gardner stuff and make a bunch of hippie arguments about ‘no one can even know anything’ but here’s a simple example I like to tell people: LeBron James is a genius. Have you seen the way he drives to the hole? No one else his size has ever moved like that on a basketball court (or maybe anywhere). Just because he probably couldn’t pass a test for MENSA, does he not still possess a staggering intelligence in his selected field? He’s also a pretty good public speaker (maybe not always content-wise), so he hardly has a one-track mind.

Now, let’s bring this back to eugenics. And we’ll use my favorite movie as an example: Gattaca. In case you haven’t seen it, the basic premise of Gattaca is that even in a world where people are selected for and typecasted based on ‘superior’ genetics, the human spirit can overcome such boundaries. This is epitomized by a few tremendous scenes where Ethan Hawke’s character, who is the natural-born runt of the litter, defeats his genetically-selected-for brother in harrowing games of chicken based on who can swim the farthest into the ocean. So, let’s say my imaginary eugenicist friend (not really my friend), points out: ‘well we can just select for the traits that gave Ethan Hawke his enhanced human spirit!’ On its face that doesn’t sound that bad. Successful people are motivated and have some gumption, heck I’d argue that that’s most of what comprises intelligence (does one read a lot because he/she is smart, or is he/she smart because that person is motivated to read?).

BUT is this really what we want? I can just imagine selecting for a bunch of Ethan Hawkes. Sure, THAT one in Gattaca was a good guy, but you know what you get when you take an average guy with a lot of gumption? A hedge fund douche. Suddenly we’d be selecting for a bunch of competitive people, and with all the different ways our environments can shape us (both by chance and by the families/places we’re raised in), you’re going to get some good ones and some bad ones. And then you’re going to get the housing crisis.

So the answer is NO. No matter what trait you come up with, there are problems. Further, that evolution stuff works prettttty well. Creating a spectrum of humans with a spectrum of strengths and weaknesses is what makes the world go ‘round. Limiting people by selecting for some kind of trait that maximizes ‘g’ is truly that: limiting.

I should point out: I’m not against trying to weed out mutations for debilitating diseases. No one should have to live through Huntington’s if it can be avoided. But when @razibkhan says “So sorry to turn this upside down, but personal eugenics may in fact be a boon for the ugly, stupid, and psychologically unstable, because it gives them a opportunity to close much of the gap with those who were lucky in the genetic lottery”, we’re dealing with some very QUALITATIVE traits that make humans what they are. Having to overcome adversity is part of what makes Ethan Hawke swim farther, and part of that adversity came because he did NOT win the genetic lottery. That was kinda the point of the movie. As for psychological instability, it is not easy to quantify, and many of the great works and creations of all time came from people on the edge. You want to select AGAINST this? If we did this already, we probably wouldn’t be the same advanced human race we are today.

So, to sum up, when @matingmind says “Obviously you should make babies genetically healthier, happier, and brighter,” I’m willing to agree with the first of those (physically, at least). Hopefully as I’ve explained here, you’ll understand why I think the latter two are fool’s errands, and ones that will probably harm humankind.


11 thoughts on “Some thoughts on the newfangled eugenics stuff

  1. “So sorry to turn this upside down, but personal eugenics may in fact be a boon for the ugly, stupid, and psychologically unstable, because it gives them a opportunity to close much of the gap with those who were lucky in the genetic lottery”, we’re dealing with some very QUALITATIVE traits that make humans what they are. Having to overcome adversity is part of what makes Ethan Hawke swim farther, and part of that adversity came because he did NOT win the genetic lottery. That was kinda the point of the movie. As for psychological instability, it is not easy to quantify, and many of the great works and creations of all time came from people on the edge. You want to select AGAINST this? If we did this already, we probably wouldn’t be the same advanced human race we are today.”

    By the time this becomes possible, we’ll be have a whole different (much more comprehensive and serious) conversation on eugenics. So this whole post is largely nonsensical.

    That’s not even to mention that intelligence is quantifiable and IQ tests do broadly measure it. The existence of special mental abilities doesn’t negate the existence, and perhaps more importantly, the utility of general intelligence.

    It may shatter your thinking to know that evolution does indeed work. So much so that it did in the past select for not only intelligence but a host of other traits, traits were instrumental in creating modern advanced civilization. This necessarily happened by leaving individuals with a variety of certain traits out in the genetic cold, so to speak. Hence, is there anything fundamentally different in attempting to deliberately guide that process?

    As I noted over at Kevin Mitchell’s, even if whatever selection system we cook up is imperfect, and even though the heritability of whatever desired traits is not 100%, it will still work to increase those traits in the target population on average. Statistics still works.

    • Thanks for the comment. Posted this on Kevin’s blog after reading through the debate in response to your points and a few others.

      I think there are a couple fallible assumptions being made by the pro-eugenics folks. Right away, I know their counterargument to this post is going to be ‘we won’t know until we try!’ But that’s the very last thing you say before Pandora’s Box opens.

      There have been a few analogies to basketball. In particular:
      Height is to winning at basketball as intelligence (g) is to winning at society.
      Here are two extensions of this analogy to consider.

      1.) There might be inherent limitations to selecting for IQ. In particular: draining other qualities at the expense of selecting for IQ. There are many 7’ people that are horrible athletes. Physical (or biophysical, if you like) limitations make it increasingly difficult for someone of that height to achieve basketball success. And if you make a 9’ tall human, he might be somewhat useful at basketball, but there would be a bunch of 6’ and 7’ guys running circles around him. A team full of 9’ players without any corresponding 6’ players could be pretty bad. The point of this analogy being that height, like IQ, isn’t the only factor in success, and when you select everyone for only one factor it is not obvious a net benefit will ensue (even when you know the parameters of the game).

      2.) So, you say, what if there aren’t inherent costs to the benefits and we can actually enhance ALL ‘desirable’ qualities simultaneously. What if somehow we can overcome biophysical limitations and make everyone 12’ tall and as fast as Usain Bolt? Or better yet: 20’ tall and as fast as a cheetah! Well, basketball has rules. And baskets 10’ tall. And 94’ x 50’ courts. Such players will quickly be constrained by the limitations of the game and start fouling out really quickly. Or become increasingly less effective as their size/speed combinations become unwieldy within the dimensions of the game.

      Just change the rules then! This is where things get bad. The analogy to changing the rules of ‘winning at society’ include such things as genocide, social injustice, dystopias and so many more things I can hardly imagine.

      You might think this is a dumb analogy: but I implore you that it is not. We are constrained by the physical world we live in. We SHOULD BE constrained by making humanity better in an ethical way. Maybe you want to roll the dice and possibly incur the Wrath of Khan—but it’s not so easy to close Pandora’s Box once it’s open.

      PS—I’ll admit I am poorly read on the relationship between psychophysical measurements and intelligence. But I believe equating reaction times and inspection times with “intelligence” is flawed on many accounts. One, the limitations of measuring and utilizing intelligence itself (as I go on about above) is flawed. Two, processing speed is just one of MANY components of making beneficial decisions for a person (or a society). And really this just collapses into the analogies I wrote out above.

  2. Pingback: Problems with using “IQ” in…anything | Too long for twitter...

  3. You people are a hoot. It’s one thing to worry about the questionable morality of human eugenics, or even to warn your peers about the adverse social consequences that may result from breeding a race of super-humans — some of which may be entirely unpredictable, despite the best efforts of geneticists and bioengineers.

    But no, just like many a confused simpleton before you, you’re not only saying that eugenics is inadvisable, or that it is of dubious ethical value — you’re saying that it wouldn’t work whatsoever on traits like beauty, intelligence, or personality. Why? Because of the made-up achievements of a fictional character in a friggin’ Hollywood movie, that’s why. You can’t even parody this sort of chicanery. Star Trek was a wonderful television series, but shit, even I know that Khan Noonien Singh doesn’t exist in reality, and that “extreme ambition” (or whatever label you affix to that cluster of psychopathic traits) is a heritable trait, just like height and IQ scores.

    Hate to break it to you, but there are almost no Ethan Hawkes in real life. In reality, children with two-digit IQs who are born to privileged parents almost never grow up to become physicists. Achondroplasic dwarfs never become Olympic shot-putters, gymnasts, or marathon runners, and neither do the vast majority of people within the normal range of the spectrum. And what a shame that nearly all of these traits are heritable, no matter how ill-defined they may be — for the twin and longitudinal adoption studies demonstrate once and for all that just about any phenotype in the human species may be tinkered with through selective breeding. Even the silly ones like what side you like to sleep on at night, the number of times you snore between 3 AM and 6 AM, and your taste for mystery novels.

    You, of course, are totally ignorant of quantitative genetics. In order to tinker with any phenotype from skin color, to height, to muscularity (and beyond), all you need to do is to demonstrate that there is genetic variation in the population which contributes partially to the total phenotypic variance (otherwise known as heritability), and that statistical gene-environment interactions are of relatively little importance. Both are true of IQ and height. Animal breeders have managed to select for all sorts of unusual traits in domesticated animals, ranging from ear shape in cats, to herding ability in dogs, to egg laying capacity in hens (which ranges from over 200 per year among domesticated layers, to just a few per year among their wild counterparts), to speed and trotting patterns in horses, and beyond. Some of these qualities have no objective definition that holds true across cultures, or even between two breeders within the same country. We may not understand the genetic architecture of any of these traits today, but that didn’t stop anyone from succeeding, did it?

    As it turns out, on a functional level, there’s not much of a difference between athletic ability in the movie GATTACA and horn length in longhorn cattle. Both are heritable, and both can be altered in a predictable manner by tinkering with the genome. If you think this is morally questionable, well, hat’s off to you, but if you insist that it won’t work because of some silly objections based on bad science — may you be one of the first to be enslaved by an augment chieftain during the inevitable Eugenics Wars.

    (EDIT: Oops, I missed it in your reply above, but you did mention the Wrath of Khan. How predictable.

    From now on, I recommend that you restrict your observations to domains where you are not entire blind, deaf, and dumb. Not science.)

  4. As for IQ testing — if Lebron James really were such a talented public speaker, the odds of him scoring below average on a “MENSA-like test” would extremely low.

    Similarly, students who do well in physics almost never do that poorly in English, French literature, or any subject to which they apply their best efforts. It’s almost as if all of these cognitive abilities, no matter how unrelated they may be, are positively correlated.

    Why is this?

    You would know, if you took the time to learn anything about psychometrics.

  5. @misdreavus With your reliance on IQ, I implore you to see my new post from tonight 😉

    You’ll also note that I did think selecting against diseases was okay by me. It’s more the idea of thinking we know what to select for, and wouldn’t be sophomorically doing so before we knew the potential drawbacks of our path. If your breed of cattle doesn’t work out, no biggie. It’s not so easy for humans…

    • Measure the heritability for a certain trait in any domesticated species; say, “nervousness” in chihuahuas. (Consult a book on agricultural science sometime, there are all sorts of methods.) As long as it is above zero, you are good to go.

      Now, devise some sort of test to measure this nebulous quality (“nervousness”) that has a sufficient degree of test-retest reliability — by that, I mean that the same animal measured multiple times by different people, following the same rubric, under slightly different environmental conditions, should achieve roughly similar scores. Of course, Pepe the Chihuahua might score high on one day, but only medium-high on another. He might be more nervous around Chloe than around Pedro, but as long as the test works reasonably well, and the sources of error are not largely unidirectional, it should do the trick.

      (By the way, the test-retest reliability for any IQ exam is roughly 0.9. And no, practice doesn’t do that much — not even if you can afford fancy SAT tutors.)

      Next, you must alter the allele frequencies of the genes responsible for the total genotypic variance. Do you want chihuahuas who are easygoing around strangers? Well, for starters, you could prevent the bottom 30% of scorers on your “nervousness” metric from breeding. (This is called “truncation selection”.) If this process is repeated, your pack of chihuahuas should grow less and less nervous with each passing generation. Presto! Before you know it, you will have an entirely different breed of dog. One that isn’t yappy and skittish around every new friend that you introduce to your apartment.

      It does not matter if there is no uniform definition of “nervousness” that holds constant across all cultures. It does not matter if some people may actually desire timidity in their precious little lap dogs, or if genes for “nervousness” may have pleiotropic effects on other phenotypes, which will inadvertently be selected against during the process. It will still work. Period! There’s almost never any instance where it doesn’t.

      Do I have any reason to believe that traits like “beauty”, “IQ”, or even “piano playing ability” are any different? No.


      Hell, most farmers from 8000 B.C. onwards dispensed with ALL of the steps I articulated above, and yet artificial selection still worked. How does this even happen?

      You can even domesticate wild foxes in less than 60 years, if you try hard enough.

    • Sterilizing people because they are judged to be immoral, stupid, ugly, etc. is of course immoral. And most civilized people oppose this. (I don’t know about Geoffrey Miller, but I sure hope he and I are in agreement here.)

      But what if people were granted the ability to alter the genetic potential of their own progeny?

      Imagine if a 5’3″ man married a 4’8″ woman, and that neither of them had any autosomal dominant disorders that result in shortness, or any other alleles of large effect. (Also, assume that they were reared under environmental conditions that are typical of most developed countries. )

      If they could somehow manage to have 1000 children, by pure chance, a few of them could end up six feet all — drawing exclusively from the additive genetic variance in the two parents alone! You wouldn’t need to introduce any new strange genes. And neither would the poor lady have to suffer through so many pregnancies. Now that rapid genomic sequencing has come of age, only the few embryos with those desired characteristics would need to be implanted. Whether Ethan Hawkes is real or not, GATTACA is indeed feasible in real life.

      Would it be perfect? Hardly, a star olympic athelete might turn out to win a silver medal, instead of a gold. But could we make it reasonably accurate?

      Of course we could.

  6. Thanks for the comments, you bring up interesting points, which I think are very much what I’m trying to address.

    Yes I understand basic genetics and breeding. But most of those miss the point. For one, you mentioned like a dozen traits and only 2 were cognitive. Yes I’m fully aware you can get taller people by breeding taller people. This doesn’t bother me so much from a eugenics standpoint.

    What bothers me is basing cognitive eugenics on an arbitrary test. The key sentence being “It does not matter if there is no uniform definition of “nervousness” that holds constant across all cultures.” Why do you think this doesn’t matter? And what proof do you have that selecting FOR something arbitrarily “good” and AGAINST something arbitrarily “bad” is DEFINITELY a good thing?

    Geoffrey Miller pushes for the GATTACA method you described–making many iterations of natural kids and selecting the best embryo. For height and eye color or non-cognitive traits I don’t have a huge problem. But for cognitive traits you’re measuring something very complicated with greater environmental impact (to the negative, as well) that is frankly arbitrary.

    And here’s one of the biggest problems with that: there’s no way we’re patient and do it right. Let’s say you’re right about the genetic stuff above and there really is an objective ‘motivation’ trait to select for that makes everyone work harder with no apparent downside. Do you really think we’re going to wait until we have that perfectly parsed out? Nope. We’re gonna dabble with it before we know what we’re doing and screw things up. I repeat, if you mess up your cattle progeny and they don’t have longer horns: not gonna matter much. But cognitive traits in humans are hardly something to be messed with willy-nilly.

  7. Pingback: There are no easy shortcuts to ending obesity. | Too long for twitter...

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 )

Google+ photo

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

Connecting to %s