July 14, 2007
From the desk of Jane Galt:Department of huh?
Q: Would you put on your right-thinking left-liberal educated-in-Berkeley-and-Madison hat for a moment?A: I'd find nothing easier. (You left out the dirty hippyprogressive Montessori school where they taught me Pirandello and Diderot.)
Q: Very good. (It didn't fit the rhythm, and anyway they get the picture.) How would you react to the idea that a psychological trait, one intimately linked to the higher mental functions, is highly heritable?
A: With suspicion and unease, naturally.
Q: It's strongly correlated with educational achievement, class and race.
A: Worse and worse.
Q: Basically nothing that happens after early adolescence makes an impact on it; before that it's also correlated with diet.
A: Do you work at the Heritage Foundation? Such things cannot be.
Q: What if I told you the trait was accent?
A: I'm sorry?
Q (in a transparently fake California accent): When you, like, say words differently than other people? who speak, like, the same language? because that's how you, you know, learned to say them from people around you?
A: Do you have a point to make, or are you just yanking my chain?
Q: Would you agree that accent has all the characteristics I just described?
A: Higher cognitive functions — heritable — class and race — not plastic after adolesence — correlation with diet, hah! — I guess I must.
Q: But would you say that there is any genetic or even congenital component to accent?
A: Not really. Obviously, some congenital conditions, like deafness or defects of the vocal chords, make it hard to impossible to acquire any accent. And I can imagine, though I don't know of anything, that there might be very specific mutations which make it hard to hear a distinction between a given pair of sounds, or easier to learn a specific distinction. But, in general, no, there is no non-trivial genetic component to accent.
Interpreted by LizardBreath:
So, there's this quality that's intimately connected with the higher brain functions, very highly heritable, and very strongly correlated both with race and with ethnicity on a finer-grained scale than race. It's somewhat plastic in childhood, but unlikely to change significantly after adolescence. A standard liberal reaction to that claim might be "(A) I don't believe you that those are the facts, because (B) if that were true, it'd be clearly an innate, genetic quality, and I don't believe you that there are such genetic differences with intellectual implications between ethnic groups."The quality, for those who haven't clicked through yet, is accent, which clearly has no genetic component whatsoever. The Sloth goes on to argue by analogy for the plasticity of things like IQ, and you should read the post. But I wanted to use it as a general defense for total skepticism on common-sense arguments about innate qualities of human nature.
I'm pretty uncomfortable with IQ determinism. But this is bizarrely beside the point. The serious IQ guys don't point to correlations between the IQ of children and their parents, which could obviously be either nature or nurture; they mostly rely on adoption studies, which (like studies of height and weight) seem mostly to show that adopted children more closely resemble their biological parents than their adoptive parents. No one is denying that there are plastic heritable characteristics; they're just denying that IQ is among them.
The gender evidence is harder; no matter who adopts you, you're still a girl. But there should be cross cultural variation, and there doesn't seem to be much, which is worth thinking about.
Posted by Jane Galt at July 14, 2007 6:45 PM | TrackBack | Technorati inbound links"); ?>Wow. And I thought the liberals in Washington that demand music lessons, calculus classes, and philosophy claimed that stimulating patterns of thinking increase development of the components of intellect. And team sports for building 'character', or at least, submission to social pressures.
Just think of how many hours of useless tedium we can skip and education dollars we can save.
Posted by: Brad K. on July 14, 2007 10:07 PMYes, the post clarifies the IQ guys position. but it doesn't make it any less troublesome
I feel much more comfortable with those participating in the conversation. At least they both have a role for education, positive parental roles, etc....
Posted by: lasser11 on July 15, 2007 1:09 AM"they mostly rely on adoption studies"
Really? What about twin studies?
"Really? What about twin studies?"
Many of the twin studies are also adoption studies in that they concern identical twins raised apart (in comparison to identical twins raised together and fraternal twins raised in both conditions).
Crud. I'm a left-liberal educated at both Cal and Wisco, albeit in the sciences. How offended should I be?
Posted by: Matt B on July 15, 2007 12:02 PMlasser11 writes "I feel much more comfortable with those participating in the conversation. At least they both have a role for education, positive parental roles, etc...."
If I desperately wanted to believe that I didn't descend from monkeys, so much that all other considerations were irrelevant, I might feel less queasy about the kind of religious fundamentalists who prefer to ignore uncomfortable evidence. Perhaps I might even write "At least they have a role for a essential characteristics of humanity, a loving God, etc...."
Posted by: William Newman on July 15, 2007 1:38 PMIf you follow from my post through to Shalizi's post, you'll find that what's being discussed is the Flynn effect, which shows that IQ measured by IQ tests has been increasing over time at a rate that can't possibly be attributed to changes in the genetic makeup of the populations tested. This is really strong evidence that some form of environmental change has a powerful effect on measured IQ.
How does one reconcile this with the adoption studies you talk about? First, no one denies that there's a genetic component that influences IQ. It's not particularly meaningful to talk about how strong that influence is -- with a sufficiently bad environment, any child, regardless of genes, can emerge profoundly retarded. That doesn't mean that environment is the only important thing, just that you can't break it down into 40% environment/60% genes with any degree of meaningfulness. Second, I have the impression that the strength of adoption/separated twin studies tends to be overstated. Particularly separated twin studies, which if you think about it are going to have incredibly small ns -- twins don't get separated often at all.
Posted by: LizardBreath on July 15, 2007 1:51 PMWait a sec. How can a trait be both heritable and have no significant genetic component? I didn't inherit my parents' Cream City accents, I learned it from them and everyone else I knew. I could have a child with someone else raised in WI, but if we happen to live in, say, Boston then I'll bet bottom dollar that he'll pronounce the 'L' in Milwaukee.
Between the feint towards accent at the start and the multiple levels of buried links, I really couldn't figure out what the message of this tale was. That education improves IQ? Really? That seems pretty damned mundane to warrant such a convoluted post.
BTW: If IQ scores are rising, then why are SAT scores falling, or rather being inflated to give the illusion of not falling?
When you say that accent is correlated with diet in pre-adolescents, what's the connection? Is it purely socio-economic, or even more extremely are we talking about people who are basically starving as children and that affects their accent?
This has made me very curious.
Posted by: Tracy W on July 15, 2007 4:52 PMMatt B: I think the post is using "heritable" to mean "If your parents are one way, you're likely to be that way too."
Heritable like money, not like eye color.
Posted by: jb on July 15, 2007 5:24 PM"How does one reconcile this with the adoption studies you talk about? First, no one denies that there's a genetic component that influences IQ. It's not particularly meaningful to talk about how strong that influence is -- with a sufficiently bad environment, any child, regardless of genes, can emerge profoundly retarded.”
So what you're saying is that you don't believe in functions of multiple variables? f(x,y) = x*y, therefore it's meaningless to characterize f wrt to x?
“ That doesn't mean that environment is the only important thing, just that you can't break it down into 40% environment/60% genes with any degree of meaningfulness."
And you don't believe in statistics (multiple regression) either?
Moreover, your argument also applies to your own position. Surely it's meaningless to talk about the effect of environment, since the wrong genes will cause profound retardation regardless of environment.
“Second, I have the impression that the strength of adoption/separated twin studies tends to be overstated. Particularly separated twin studies, which if you think about it are going to have incredibly small ns -- twins don't get separated often at all.”
Alright, google up the studies and run the numbers yourself, then tell us what's wrong with them. Pulling out statistical significance numbers from some sample size is high school level statistics.
Posted by: random math major on July 15, 2007 7:17 PMI took my own advice and spent 5 minutes googling for some references. See http://www.loni.ucla.edu/~thompson/IQ/NRN2004_IQ.html for a short summary. Refs 82-85 in that paper address the confounding variables mentioned in the unfogged thread.
Posted by: random math major on July 15, 2007 7:27 PMNo, you misunderstand me. Sure, with the right (or, I suppose, the wrong) genes, anyone can be profoundly retarded beyond the hope of and significant environmental effect. For each person, genes and environment interact -- if the environment happens to be 'locked in a dark closet and underfed for the first twelve years of life', the genes aren't going to matter a hell of a lot, and depending on the genes, you can have someone for whom environment isn't going to be able to help much.
Saying 'genes control X% of variation in IQ and environment controls Y%' can describe a regression run on a particular population being studied, but it's not going to be meaningful either for an individual, or for a population including a different range of environments.
Posted by: LizardBreath on July 15, 2007 8:06 PMThe odd thing here is that accents are not heritable, not as I understand the term. As Steven Pinker and Judith Rich Harris have often pointed out, children learn to speak like their peers, not their parents. Sure, people often have accents that resemble their parents' accents -- but this is only when the parents have the same accent as the peers. But when those accents differ, peers win. That's why you often will meet someone whose parents immigrated from Russia (or wherever) and who speak English with a thick accent, but he himself picked up a Brooklyn accent rather than the thick Russian accent.
Posted by: Stuart Buck on July 15, 2007 9:08 PMThanks, Stuart. That's exactly what I was getting at. Besides, "parents" != the people that raised you. The accent example is not very apt.
Posted by: Matt B on July 15, 2007 9:57 PMStatistics? Studies? Bah! Let's get annecdotal:
When I was growing up in Georgia I had a pretty strong southern accent. Then I went to college up north, didn't like that my accent stood out and changed it. Now people are surprised when I tell them I'm from the south.
My mom, OTOH, grew up in Pennsylvania and moved south in her mid-20s. Though she didn't acquire a deep drawl, she definitely picked up lots of elements of the southern accent. In her case, I'm positive it was unintentional.
Posted by: Derek Scruggs on July 15, 2007 10:20 PMThere's another way the "accents" example is a dodge - accents typically change over time, based upon environment. Any claim that accents are "fixed" by adolescence flies in the face of the facts. Come from Scotland, spend twenty years in the South and you'll have a blended accent: no' one nor t'other.
I'm with Derek.
Annecdotally, both I and my mother were born and raised in NE Ohio. We both that the kind of rapid-fire, broadly neutral accent of that area, but with a couple of VW miner-family-isms - "chimley", not "chimney", "pitcher", not "picture", based on family history.
We moved to Tennessee. I did highschool there, and while I learned to slow my speech down, I credit that to debate training. I college (in CT), I corrected most of my mispronounciations. I also credit learning German, when I lived there as an exchange student, with making me listen to language more carefully. I'm now a Brooklyn resident, and sound basically the same that I did at 20.
My mother, on the other hand, stayed in TN, got involved in local politics, and picked up a southern drawl, along with some of the metasyntactic tics and social cues that I never did. ("Are ya'll liking that OK?", meaning "hurry up and finish eating". She hasn't progressed to "well that's just so inneresting", meaning, "just wait until I tell everyone at church", but she isn't a gossip.
And this happened to a well educated, smart person.
If anything, I went more formal and clipped, and tend to be considered a bit pedantic in my speaking patterns. But I have found that I swear a hell of a lot more since I moved to Brooklyn, so that probably argues in favor of some locational bias.
Posted by: fishbane on July 15, 2007 10:44 PMAnecdotally in linguistics and foreign language instruction (two fields I'm in). There is a spectrum of accent behaviors among people.
On one end you have mimics (who tend to be better at language learning). These are folks who pick up what's around them without trying (indeed, they often seem unable to not pick up accents). Even their accent and vocabulary in their native language seems to depend a lot on who they talk to.
In extreme cases, they'll unconsciously modify the morphology and syntax of their own language to match non-native speakers (if that's who they're around).
At the other extreme are what I call rocks, who get a fixed accent in whatever language they use and do not / cannot change, even if they want or try to (and usually they don't do either). Their approach to idioms and turns of phrase is similarly set in stone, they'll continue to use favorite expressions that are at odds with the rest of their linguistic environment (if they know a particular idiom, they assume everybody else in the world does to and cannot be convinced otherwise).
These are the extremes, most folks are somewhere between.
Posted by: michael farris on July 16, 2007 12:05 AMActually, there's a great twin-based research design using a massive dataset of all the twins in the UK. For any outcome variable of interest (IQ, for instance), they take each set of twins and find the difference between twin A and twin B. These differences are smaller (often much smaller) on average for the subsample of identical twins (who share 100% of their genetic material) than for the subsample of fraternal twins (who share 50% of their genetic material). The difference in variance between the subsamples, corrected for the fact that the fraternal subsample are 50% related rather than 0%, tells you the genetic component of the outcome variable. Bit more complicated than that, actually, but that's the general principle.
Google Robert Plomin, Avshalom Caspi, and/or Terrie Moffitt for a vast array of such articles.
Posted by: Charlie on July 16, 2007 12:46 AMThat's interesting, Michael. Where do people who self-conciously adopt linguistic mannerisms fit in? I guess I'm thinking of things like when I was in Germany, shortly before I managed to start thinking in German, I made myself use the "na" placeholder instead of "um" when either trying to think of what to say, or for effect. I find a similar thing happens to me with public speaking - I'm not a natural. But even in my English, I see an effective semiverbal/nonverbal technique, and I try to adopt it.
I'm now in a combination Latino/Polish neighborhood, and I passively understand Spanish well enough to know what's going on, and am working on a more active language command. Polish is a little odd to me, which I find surprising.
Posted by: fishbane on July 16, 2007 12:51 AMYou're pretty far over on the mimic side, I'd say.
When I was much better at Spanish, I wondered why so many speakers dropped syllable final -s (como etá uté?) and so I practiced doing so for a day or two and found myself with a habit I couldn't break.
And Polish is a _little_ odd to you? It's my second language and I still find it very odd.
Posted by: michael farris on July 16, 2007 4:45 AMFWIW, I grew up in Austin, Texas, and everyone around me (school and family) had a sort of Texas accent. Despite spending all my time around these people, I never got a "Texas accent", and when I tell people I grew up in Austin (all life up to 18), they typically refuse to believe me. I don't even sound like my brother, with whom I spent most of my life.
In fact, people often ask if I'm a foreigner, usually German! (Yes, I took German in school.)
I think I'm like the rocks that michael_farris referred to above, except that ... when I spent seven weeks in Germany, I adopted the local habit of rolling r's in German. (Franconia) Go fig.
(Hey, maybe my practice of putting underscores in the blanks in people's names is a written instance of being a rock.)
Posted by: Person on July 16, 2007 9:51 AMmichael_farris, when you describ mimicry in people, especially when a person is talking to a group of others, you perfectly describe Hillary Rodham Clinton. When she talks to blacks she sounds 'so black.' When she talks to a group of white southerners, she is so Cracker that the moisture dries up in my mouth. When she's in Connecticut twangy Northeasterner in me wonders at what weird place she learned her vowels.
I wonder if mimicry carries over to the body language, too? OMG, I just had a vision of Hillary doing Hip Hop and disgorged my croissant...
One more anecdote. My wife spent six months in China, during which I started working closely with a new business partner. When she came back from China, she noticed that I'd picked up a new verbal tic from him. At first I argued that I had always talked this way, but she was adamant that it was something new.
It wasn't a phrase like "yadda yadda yadda" from Seinfeld, more of a variation on "uh" or "um," and not something I would ever comment about around the water cooler. That to me indicates it was totally subconscious, not an attempt to keep up with pop culture.
Posted by: Derek Scruggs on July 16, 2007 12:21 PMThe serious IQ guys don't point to correlations between the IQ of children and their parents, which could obviously be either nature or nurture; they mostly rely on adoption studies, which (like studies of height and weight) seem mostly to show that adopted children more closely resemble their biological parents than their adoptive parents.
So, wouldn't that mean that IQ had a high degree of heritability? So, those criticizing that idea are wrong, despite the orthodox buzz I hear (like from S. Gould) is that inheritance of IQ is close to bunk, and dangerous to purvey as well?
I beg to differ.
No matter who adopts me (assuming my parents died and I was adopted as an adult), I'd never be a girl.
Posted by: Sigivald on July 16, 2007 1:45 PMTracy, I think the correlation between accent and diet is that they're both related to some third factor (socioeconomic status or location). Something like: if you're from the South, you eat grits. If you're from the south, you have a southern accent. Therefore there's a correlation between eating grits and having a southern accent.
Posted by: sasha on July 16, 2007 3:58 PMIt's pretty evident that other physical characteristics differ between ethnic groups and between the sexes (height, for example), which is an interaction between genes & environment (I'm not short because I'm underfed or had a diseased childhood, though that was true for many of my ancestors). So it's not beyond the pale to think there could be some cognitive differences as well -- there may not be, but it's also possible there will be for reasons of both genetics and environment.
If you've decided what you want the answer to be before you look at the data, or you decide not to look at the data at all, then you're not doing science but religion. And don't think anybody is fooled. The people who decide on Correct Opinions may think it Not Nice to notice that, say, women in general are shorter than men; but pretending the difference does not exist does not make it go away or make it such that no one sees it.
Posted by: meep on July 16, 2007 4:13 PMIn the discussion, be careful to differentiate between "inheritability" and "heritability". The guy seems to be using the latter (roughly) as a synonym for "communicability", that is, can you catch it from your friends. IQ and accent are closely related this way-- no one would claim that peer group has no effect on your ability to solve problems.
One's ability to solve problems, whether real life ones or symbolic ones, has got to be correlated rather highly with the degree to which solutions to those particular classes of problems have been modeled for you. (Someone who has never had toolworking modeled for them would show really bad performance on a carpentry IQ test.)
Posted by: Twill00 on July 16, 2007 5:45 PMGood point Twill00. I believe that is why there is something called a 'Discovery' that has, historically, been associated with moving forward in an art, industry or even geographically speaking. If you are with someone who is a raw genius and he stumbles on something you never really paid attention to, or didn't realize one small step would get you to that point...it helps to see or hear that it happened.
So, that does it for me!!! Those of you who do not have IQ's of at least 140, don't bug me! Why would I want to spend time with sub-, normal, or even moderately intelligent people, when I can find all the super-Mensas in the area and rub synapses with them?
Posted by: falkoyn on July 16, 2007 6:21 PMIt's really easy for smart people to pick out other smart people. However, to say who is smarter is usually more difficult, because we all think in different ways.
I think the limits of your I.Q. are indeed inherited, but whether you press those limits or not are affected by your environment, and other exogenous sources.
Posted by: Stan on July 16, 2007 10:22 PMI just got here and I'm a little confused. Are there actually people who think IQ doesn't have a genetic connection? How do they explain the complete dominance of Ashkenazi Jews in top-level chess?
Posted by: Gary Kasparov on July 16, 2007 11:21 PMHow does IQ differ from height? Heights are undoubtedly inherited, but of course there is a drastic environmental effect, mostly at the extremes. Heights have their own Flynn effect in that they have certainly increased over time. Heights vary widely between the sexes and among different ethnic groups. What's the scientific rationale for accepting the heritability of height but not intelligence?
Posted by: Josh on July 17, 2007 7:35 AM"Ashkenazi Jews in top-level chess"
Self fullfilling prophecy? They tell themselves "our kind" has always been the best at chess, and they pride themselves on it so thats what they focus their lives on and voila they become chess wizards.
Not saying that is a perfect explanation, but it would stand to reason that if you have a certain culture that valued chess highly that culture would be better at it and the fact that they're better at it does not mean we should run around declaring that its in their genes.
Americans consistently create and consume more wealth. How can you claim that it's not the result of our genetics?
Germans in southern Baveria consistenly chop more wood than everyone else. How can you claim its not in their genes?
Culture, circumstance, emphasis, focus, all play a tremendous role. Not to say their brain capabilities/limits aren't passed down, but if all you do is focus on playing chess then over a lifetime you better be among the best in the world.
Posted by: cdub on July 17, 2007 10:30 AMLB,
So your mention of n was a red herring – even if n were infinity, that wouldn’t be enough to satisfy you. No doubt I’m misunderstanding you again, because it seems to me that you’re saying it’s “not meaningful” to talk about something if it can’t be characterized over the set of all possible circumstances. That would make all social science research meaningless, not to mention a lot of hard science research too.
If your objection is merely that we don’t know with absolute certainty what would happen in circumstances we’ve never encountered yet, then I agree, but I don’t think anyone is saying otherwise. If you’re a Bayesian, surely collecting information about one environment tells you something about other environments. If you’re a frequentist, I’m sorry :-) For instance, you mention the Flynn effect, but studies of g have been done over the past couple of decades, and the heritability of g hasn’t really changed over that period of time. Ditto for cross-cultural studies. If you believe that the Flynn effect is not just a statistical artifact, but a real change in g, doesn’t that say something about the robustness of the heritability of g?
Posted by: RMM on July 17, 2007 1:15 PMSo what you're saying is that you don't believe in functions of multiple variables? f(x,y) = x*y, therefore it's meaningless to characterize f wrt to x?
I think you are oversimplifying matters so far that you lose the point.
First, AFAIK, "intelligence" (the 'x') does not seem to be widely agreed upon: memory/retention (short and long term), language processing, speed of uptake, ability to transfer knowledge, ability to make indirect ("aha") leaps, etc are all part of the gig.
Second, environment includes a huge number of variables, including (off the top of my head):
nutrition,
education of parents,
income of parents,
number of teachers,
education level of teachers,
teaching ability of teachers,
commitment level of teachers,
classroom sizes,
number of siblings,
intelligence of siblings,
number of peers,
intelligence of peers,
parental emphasis on education,
cultural emphasis on education,
time spent in school,
time spent on education (out of school),
(famously) number of books in household,
# of books read for leisure,
geographic location,
salience of education in relation to other tasks,
number of drops on head as child,
etc etc etc.
In a general sense, we all agree that "quality of education," and "support for education" are probably the most important factors and that many of the above factors are coincidental with each other, but they are impossible to measure as independent units.
(It doesn't matter how committed your teacher is if they never finished jr high school, and it doesn't matter how educated they are if they don't make an effort or aren't available. Same with income as a proxy for parental learning, or parental learning as a proxy for resources available, etc.).
So, you tell me:
if x = a * b * c * d * e (where the relationships are actually unknown)
and y = aa * bb *... ss
The question folks like myself would as is how well are we going to be able to come up with that f (x,y) function? What would that look like in the end?
(BTW, this is an honest question.)
Posted by: Brad L on July 17, 2007 1:57 PMMy only point about separated twin studies is that as finding a whole bunch of separated twins to study is a non-trivial task, we're not talking about a great mass of data of here. Here's a paper discussing some of the methodological problems with the Minnesota study of separated twins.
More generally, the point of the post is that this is not an area where relying on common sense is going to be terribly useful. Someone brought up height as a heritable characteristic above, and it certainly is. Nonetheless, if you go back a couple of hundred years, you could have found a strong correlation between height and residence in a rural, rather than an urban, environment -- country people were less likely to be malnourished. Environment is going to have a stronger, and genetics a less strong, influence on final height where the range of environments includes those where malnourishment is a possibility, and measurements of the heritability of height are going to be uninformative unless they're done across the full breadth of environments being discussed.
Posted by: LizardBreath on July 17, 2007 2:11 PMAccent will seem to be highly inheritable if you only follow traits through biological families, but if you study twins that were separated at birth, or even unrelated adoptees, it will be obvious that it's a matter of environment.
Twin studies aren't quite the gold standard for inheritability, either. It's entirely possible that the 9 months in the womb has as large an environmental effect as all the years after birth, but twin studies can't distinguish that from a genetic effect. There might be a way around this, for effects common enough to show up in small samples: couples having fertilization difficulties often have eggs fertilized in vitro (outside the body) and then implanted. There are often extra eggs left over. If these were implanted in other women, and the scientists then followed those children... I just don't see how you're going to get many volunteers for the parents, though.
One other thing I've never seen a scientist acknowledge is that twins raised together aren't in exactly the same environment, because usually one establishes a permanent dominance over the other.
Posted by: markm on July 17, 2007 3:39 PMThe fact that faternal twins are essentially no more alike than normal siblings would seem to discount the idea that the shared 9 months in the same womb is a large factor.
Posted by: Roy on July 17, 2007 4:02 PMMarkm, the dip of your toe into the pool of in vitro has some interesting possibilities. I'm wondering if we have the biological technology to bring a human baby to full term, ethics and morality notwithstanding.
If so, we could then control for having to be in a woman's womb, which would allow the researcher to test for many different types of outside/transitional environments, e.g. dark and quiet, dark and classical music, dark and rap, light and show tunes, etc. Of course, minor changes in temperature could allow for a check on whether there is a significant different in a baby whose human mother had a core temperature be 2-3 deg F lower than the 'average' core temp.
Heck, I can see where we can eventually correct factors that truly do allow us to breed (perhaps the wrong choice in words here) those who are Stronger, Faster, Higher, Smarter, etc., than those of us who just...happen.
Penultimately, there would be no need for a woman to carry a baby to term, just like there really isn't any need for a sperm-donor to inseminate the woman. Heck, a good technologist could just get rid of Mom and Dad for real (a teenager's dream?) and become either a Creationist or a Darwinist (depends on your PC reference):
♂ + ♀ > ☺ ☻ ☺ ☻ ☺ ☻ ☺ ☻ ... ∞
After a few hundred thousand of them, you could get some real good statistically normed test results.
"they mostly rely on adoption studies"
Um, really? The major IQ work is on "heritability" which is defined as Cosma suggests -- correlations between parents and children. The adoption studies (like the major one in France described by Lewontin) seem to find that being raised by wealthier families gives you a huge boost in IQ.
Posted by: Aaron Swartz on July 18, 2007 12:03 AMAaron,
Follow-ups to those adoption studies which show that being raised in a wealthy family gives the adoptive child a boost in IQ strongly suggest that that boost is only temporary. The same is true with the boost blacks get by being adopted by whites. Google the Minnessota Scarr Interracial Adoption study and follow-up. Blacks adopted by well-to-do white families scored significantly higher than the black average (though not as high as their white adoptive brothers and sisters) as children, to the delight of the researchers. To their horror, however, as the black children passed through adolescence and into adulthood, they regressed back to the black mean of 85.
Posted by: MarcZ on July 18, 2007 6:52 AM"How do they explain the complete dominance of Ashkenazi Jews in top-level chess?"
There is no such thing. The highest rated Jewish chess player, Teimur Radjabov (Elo 2746; Jewish father), is placed no.10 in the world.
Posted by: ~~~~ on July 18, 2007 7:03 AM"The major IQ work is on "heritability" which is defined as Cosma suggests -- correlations between parents and children."
Except that heritability already has a technical meaning in biology, (phenotypic variance that is due to genetic effects) no matter how a statistician might want to treat it. It's nonsense to talk about accent (just as, for example, "power", or "money") as a heritable trait. It's meant to spread confusion.
Posted by: eudoxis on July 18, 2007 10:27 AMI'd say that the intent, rather than spreading confusion, is to clarify the fact that studies described as measures of 'heritability' often in fact measure correlations between parents and children, rather than directly measuring "phenotypic variance that is due to genetic effects", because we don't have techniques for measuring heritability (as strictly defined) directly.
Posted by: LizardBreath on July 18, 2007 10:46 AMThis is a pretty retarded post and discussion.
Is there anyone who has not noticed that certain people are smarter than others? If there is, please leave the room, you're clearly a hermit, raised by wolves.
Is there anyone who has not noticed the same differences in intelligence clearly demonstrated by the youngest of children (say, down to the age of 2?)
Is there anyone who has not noticed that these same intelligent children tend to turn into intellectually superior adults?
Do you really think that there is enough time for those 2 year olds' differences in intelligence to be solely due to the environment?
Try this thought experiment:
Have you ever noticed that some people are more "athletic" than others? I sure have. With the same amount of training, there were plenty of guys on the team who were faster, more agile, better coordinated, etc than me. These differences were manifest at an early age and continued throughout high school and college.
Do you think that athletic ability is solely due to environment?
Please.
Clearly a large component of intelligence and athletic abilities are inherited. Period.
Posted by: David on July 18, 2007 1:44 PMDavid,
I agree that this discussion is retarded. But p.c. ideology has a way of knocking a good 30 points off of someone's IQ. So what do you expect?
As for Jewish chess champions (I have no idea how to address someone whose moniker is a bunch of tildes), here are the undisputed world chess champions since the late 1800's and their ethnic backgrounds:
Wilhelm Steinitz 1886–1894. Jewish
Emanuel Lasker 1894–1921. Jewish
José Raúl Capablanca 1921–1927 Cuban
Alexander Alekhine 1927–1935 Russian
Max Euwe 1935–1937 Dutch
Mikhail Botvinnik 1948–1957 Jewish
Vasily Smyslov 1957–1958 Jewish
Mikhail Tal 1960–1961 Jewish
Tigran Petrosian 1963–1969 Armenian
Boris Spassky 1969–1972 Russian
Robert J. Fischer 1972–1975 Half Jewish
Anatoly Karpov 1975–1985 Russian
Garry Kasparov Jewish
That's 7 out of 13 who are of Jewish descent. Pretty impressive. No, I am not a Jew.
Posted by: MarcZ on July 18, 2007 2:49 PMThis is a pretty retarded post and discussion.
Either much more than you think or absolutely not.
Clearly a large component of intelligence and athletic abilities are inherited. Period.
The "period" doesn't deter someone like Cosma Shalizi to sputter contrived arguments to deprecate this evidence.
Yet Cosma himself cannot be seen as retarded or even average intelligence, isn't that amazing/amusing?
Therefore what can seem "retarded" is to expect rational arguments to bring any consensus over any kind of compromise.
The self proclaimed rationalists themselves aren't rational, how nice!
That's why at the end of the day snake oil salesmen and politicians always win.
Good luck naked monkeys!
Posted by: Kevembuangga on July 18, 2007 3:00 PMthe primary reason people can continue to argue against significant, heritable genetic effects on intelligence is that brains are just not as perceptible as muscles or height. everyone in the thread is likely at least one to two (or more) sd's above the iq mean (quick -- anyone with less than a 700 SAT verbal here?). If iq were height, this would be like a conclave of 6' and 6'2" plus individuals arguing over whether their height was entirely the function of milk doing a body good or whether DNA might have something to do with it.
Posted by: blah on July 18, 2007 8:21 PM
Actually, what do you mean that adoption studies show that "adopted children more closely resemble their biological parents than their adoptive parents [regarding IQ]"? I don't think that is wholly correct unless you have empircal evidence to suggest it.
Now, it is true that adopted children's IQs are more *correlated* with their bio parents than their adopted parents but it would be misleading to say that they are closer in IQ to their bio than their adoptive parents. This bit of double- speak has been exploited by racist and other charltans to "prove" that IQ is deterministic.
I think there are studies that suggest that in many adoptive households, adopted children are closer in IQ to their *adoptive parents* rather than their bio parents. What this means is that their *average IQ* is closer to their adoptive parents than their bio parents. But it still can be (and is) true that their IQs correlate with their bio parents more than their adoptive parents.
What "correlated" means in the statistical sense is this: Say we have the population of bio parents which we call X and the population of adoptive parents which we call Y and the population of adopted children which we call A.
Now, let's say that the average IQ of X is 88 and the average IQ of Y is 117 (yes, adoptive parents do tend to much have higher IQs than average). Let's also say that the average IQ of A is 107. This means that the adopted children are closer in IQ (more similar) to their adopted parents than to their bio. But their IQs might be more correlated with their bio parents. That is, higher IQ individuals in X (relative to that population) tend to have higher IQ children in A (relative to that population). But, there is little correlation between higher IQ parents in Y (relative to that population) and their chilren's IQs. That is, their IQs are no more correlated than any individual in the genral population with another. In this case, there is a higher correlation (i.e., a higher positive pearson product moment coefficient) between X and A than with Y and A. But obviously, A's average IQ is more similar to Y's than X's so it is very misleading to say that X's IQs are more similar to A's than with Y's.
Posted by: JuJuby on July 18, 2007 9:59 PMThis bit of double- speak has been exploited by racist and other charltans to "prove" that IQ is deterministic.
Let's see, take some hypothetical non-IQ traits and assume that some have zero correlation from parents to siblings while some have nearly perfect correlation and others (as many as needed for having a nice "smooth" curve) lay in between.
Where does it gets "deterministic"?
MarcZ, I know there have been many great Jewish chess players, but there is no "complete dominance" that needs a genetic explanation. (You have a few of the players wrong by the way; Smyslov is half Jewish, it was recently revealed that both of Fischer's parents were Jewish, Kasparov had a Jewish father and his mother is Armenian, and you forgot Kramnik (undisputed since October 2006), Russian.)
Posted by: ~~~~ on July 19, 2007 6:07 AM>>Let's see, take some hypothetical non-IQ traits and assume that some have zero correlation from parents to siblings while some have nearly perfect correlation and others (as many as needed for having a nice "smooth" curve) lay in between.
Where does it gets "deterministic"?
I don't know,... why don't you ask the "racists and other charletans" I was explicitly referring to? *eyeroll*
And what do you mean by "smooth curve"?
Posted by: JuJuby on July 19, 2007 11:41 PMI believe some of the strong evidence for genetic basis of a bunch of stuff (IQ and personality, but also weird stuff like religiousity) is established based on looking at the difference between the correlation between fraternal twins, and the correlation between identical twins. That gives you a shared womb, and some pairs with almost identical genes, and other pairs with 50% shared genes.
Posted by: albatross on July 20, 2007 12:51 AMI have a question. *raises hand*
Cosma says:
That is, a gene might find it in its interest to make the body it is in do things which depress its reproductive prospects, if by doing so it raises those of other bodies which contain the gene. In fact, if the benefits to the other bodies are sufficiently large in relation to the costs, it makes sense for the gene to do this even if the other bodies only probably contain the gene.
How the hell is the gene supposed to "know" about the other bodies - your brain might know your cousins in the sense that you've hung out with them a few times, but are there secret codes or signals that would tell your genes that it's okay not to breed 'cause Cousin Johnny'll pass on my inheritable characteristics (whichever ones you want to label as such)?
Posted by: arby on July 20, 2007 1:03 AMJuJuby : I don't know,...
Are you saying that you don't know what you mean when using the word "deterministic"?
why don't you ask the "racists and other charletans" I was explicitly referring to? *eyeroll*
Because you posted here, not them, and since you are arguing against the "charlatans" I assume (wrongly it seems!) that you know what you are talking about.
And what do you mean by "smooth curve"?
A seemingly continuous monotonous curve approximated by as many points as necessary, no mathematical nitpicking here, just fancy whatever curve shape you like and tell me at which point along the axis of increased correlations the traits the curve is about can be deemed "deterministic" while below this point the traits are not "genetically determined".
"How the hell is the gene supposed to "know" about the other bodies - your brain might know your cousins in the sense that you've hung out with them a few times, but are there secret codes or signals that would tell your genes that it's okay not to breed 'cause Cousin Johnny'll pass on my inheritable characteristics (whichever ones you want to label as such)?"
It's called kin selection. Of course the gene doesn't "know" and it is possible the other body might not contain the gene, but that's why C. writes: "it makes sense for the gene to do this even if the other bodies only probably contain the gene".
Posted by: ~~~~ on July 20, 2007 6:57 AMarby,
Evolution is cool that way. There's not an engineer building something that meets the requirements, but there's a very broad random search going on for ways to improve fitness, simply because genes that do a better job of causing copies of themselves to come into the world tend to spread faster, avoid extinction by random bits of bad luck, etc. So you'd expect signals that indicated kinship over most of human history to be used. Like, kids raised together will interact like siblings, even if they don't share any unusual number of genes. Adoptive parents are probably running much the same parenting program as biological parents, even though there is no kinship there, either.
Posted by: albatross on July 20, 2007 10:59 AM>>Are you saying that you don't know what you mean when using the word "deterministic"?
I don't know what you are trying to get at. You are being too vague.
I am not saying that IQ is "deterministic". I am saying that it is not purely genetically deterministic as some people think and that the racists and the Charletans have got it *wrong* (they misinterpreted the evidence).
You need to express yourself in a clearer and more complete manner if you expect people to respond with informed responses.
>>A seemingly continuous monotonous curve approximated by as many points as necessary, no mathematical nitpicking here, just fancy whatever curve shape you like and tell me at which point along the axis of increased correlations the traits the curve is about can be deemed "deterministic" while below this point the traits are not "genetically determined".
Correlation is measured by the pearson product moment coefficcient, a score between -1 and +1. 0 means there is no correlation. 1 means perfect positive correlation. -1 means perfect negative correlation. Now, any score on the ends (-1 or +1) may be considered "deterministic" in some senses of the word granted it is a cause and effect relationship and not one of epiphenomon or multiple effects from the same cause (V-correlation). But this is a more technical definition of "deterministic". Obviously, many racists and charletans (or some other common people) have much less rigorous (and more confused!) definitions of that term and any high or moderately high correlation between the vairance in populations might be (falsely) attributed as significantly "genetically determined"
http://en.wikipedia.org/wiki/Pearson_product-moment_correlation_coefficient
Posted by: JuJuby on July 20, 2007 9:08 PMI don't know what you are trying to get at. You are being too vague.
Do you mean you want to guess my agenda before cooking up a duplicitous response?
I am clearly asking you to define the meaning you ascribe to the word "deterministic".
Correlation is measured... blah... blah... blah...
I know, I am only asking you beyond which value of a correlation (positive or negative) is the observed effect deemed to be "deterministic".
To me the correct answer is +1 and -1 solely.
Therefore even the "charlatans" don't claim for a deterministic incidence of genetics on IQ.
Could you explain then what your critiscism is?
re: kin selection -
Ah, ok. So then adoption would throw the whole thing off, then, wouldn't it? If you're getting your kin recognition from who you're raised with, you're acting on false information and your genes could be jeopardized. Or am I missing the point.
Sorry for the basic questions, my gene theory education is v. spotty as you can all probably tell. I've read Dennett but not Dawkins.
Posted by: arby on July 21, 2007 1:16 AMYes, from the gene's point of view parents who care about their adopted children are mistaken.
Posted by: ~~~~ on July 21, 2007 7:39 AM>>I know, I am only asking you beyond which value of a correlation (positive or negative) is the observed effect deemed to be "deterministic".
To me the correct answer is +1 and -1 solely.
Therefore even the "charlatans" don't claim for a deterministic incidence of genetics on IQ.
Could you explain then what your critiscism is?
Hey, guy, the charletans don't have a philosophical and correct definition of determinism. THAT'S THE PROBLEM! My original post was about the (False)way that some people can interprete the adoption studies on IQ. You then moronically asked what *I* thought determinism meant (a philosophical question). I gave you a philosophically correct response. Now, obviously, just because *I* have a rigorous definition of what "determinism" is doesn't mean that the "racists" and "charletans" I refered to in my original post does too. In fact, I implied in my original post that they don't, that they have a *incoherent* idea of what it means.
Also, many racists and charletans might interprete a very high correlation as *essentially* evidence for genetic determination. This might be because they don't ahve a very clear idea of what correlation is or what it is actually measuring in these twin adoption studies.
Posted by: JuJuby on July 21, 2007 9:53 PM