I’m more directly disturbed by the bias present in your exposition: “perfectly reasonable”, “merely expresses agnosticism”, “well documented”, “harsh and inaccurate”.
Starting off a discussion with assorted applause lights and boo lights strikes me as unlikely to lead to much insight.
What would be likely to lead to useful insight? Making use of the tools LessWrong’s mission is to introduce us to, such as the applications of Bayesian reasoning.
“Intelligence has a genetic component” strikes me as a causal statement. If it is, we ought to be able to represent it formally as such, tabooing the terms that give rise to cognitive muddles, until we can tell precisely what kind of data would advance our knowledge on that topic.
I’ve only just cracked open Pearl’s Causality, and started playing with the math, so am still very much an apprentice at such things. (I have my own reasons to be fooling with that math, which are not related to the race-IQ discussion.) But it has already convinced me that probability and causality are deep topics which it’s very easy to draw mistaken conclusions about if you rely solely on a layman’s intuition.
For instance, “the well documented IQ differences between groups” are purely probabilistic data, which tell us very little about causal pathways generating the data, until and unless we have either controlled experiments, or further data sets which do discriminate between the competing causal models (only very grossly distinguished into “nature” and “nurture”).
I don’t know if the email you quoted (thanks for that, BTW, it’s a treat to have access to a primary source without needing to chase it down) is racist, but it does sound very ignorant to me. It makes unwarranted inferential leaps, e.g. from “skin and hair color are definitely genetic” to “some part of intelligence is genetic”, omitting the very different length of developmental chains leading from genes to pigmentation on the one hand, and intelligence on the other. It comes across as arrogant and elitist as well as ignorant when saying “I think my babies will be geniuses and beautiful individuals whether I raise them or give them to an orphanage in Nigeria”.
It is not bad science to be on the lookout specifically for data that claims to be “scientific proof” of some old and demonstrably harmful prejudices, and to hold such claims to a higher standard. Just as we do hold claims of “scientific proof of ESP” to a higher standard—at least of scrutiny and replicability—than, say, claims of a correlation between apparel color and competitive performance. We have more reason to suspect ulterior motives in the former case than in the latter.
Dinnertime conversations between regular, even educated people do not contain probabilistic causal analyses. In the email Grace claimed something was a live possibility and gave some reasons why. Her argument was not of the quality we expect comments to have here at Less Wrong. And frankly, she does sound kind of annoying.
But that all strikes me as irrelevant compared to being made into a news story and attacked on all sides, by her dean, her classmates and dozens of anonymous bloggers. By the standards of normal, loose social conversation she did nothing deserving of this reaction.
I feel a chilling effect and I’ve only ever argued against the genetic hypothesis. Frankly, you should too since in your comment you quite clearly imply that you don’t know for sure there is no genetic component. My take from the reaction to the email is that the only socially acceptable response to encountering the hypothesis is to shout “RACIST! RACIST!” at the top of your lungs. If you think we’d be spared because we’re more deliberate and careful when considering the hypothesis you’re kidding yourself.
By the standards of normal, loose social conversation she did nothing deserving of this reaction.
Sure. What I do find disturbing is how, knowing what she was doing (and who she was sending it to), the “friend” who leaked that email went ahead and did it anyway. That’s positively Machiavellian, especially six months after the fact.
However, I do not feel a need to censure myself when discussing the race-IQ hypothesis. If intelligence has a genetic component, I want to see the evidence and understand how the evidence rules out alternatives. I would feel comfortable laying out the case for and against in an argument map, more or less as I feel comfortable laying out my current state of uncertainty regarding cryonics in the same format.
Neither do I feel a need to shout at the top of my lungs, but it does seem clear to me that racism was a strong enough factor in human civilization that it is necessary, for the time being, to systematically compensate, even at the risk of over-compensating.
“I absolutely do not rule out the possibility [of X]” can be a less than open-minded, even-handed stance, depending on what X you declare it about. (Consider “I absolutely do not rule of the possibility that I will wake up tomorrow with my left arm replaced by a blue tentacle.”) Saying this and mistaking it for an “agnostic” stance is kidding oneself.
Since people are discussing group differences anyway. I would just like people to be a bit clearer in their phrasing.
Inteligence does have a genetic component. I hope no one argues that the cognitive difference between the average Chimpanzee and Resus monkey are result of nurture. The question is if there is any variation in the genetic component in Humans.
Studies have shown a high heritability for IQ, this dosen’t nesecarily mean much of it is genetic but it does seem a strong position to take, especially considering results from twin studies. A good alternative explanation I can think of, that could be considered equivalent in explanatory power, would be differences in prenatal environment beyond those controled in previous studies (which could get sticky since such differences may also show group genetic variation ! for example the average lenght of pregnancy and risks associated with postterm complications does vary slightly between races).
The question disscused here however is whether there are any meaningfull differences between human groups regarding their genetic predispositions towards mental faculties.
We know quite a bit from genetic analysis about where people with certain markers have spread and which groups have been isolated. Therefore the real question we face is twofold:
Just how really evolutionary recent is abstract thinking and other mental tricks the IQ test measures? The late advent of behavioral modernity compared vs. the early evidence of anatomically nearly modern could be considered for example. Some claim it was an evolutionary change following the well documented recent bottleneck of the Human species others say the advent of modern behaviour was a radical cultural adaptation to a abrupt environmental change or just part of a long and slow progress of rising population density and material culture complexity we haven’t yet spotted. Considering how sketchy the archeological record is we can’t be suprised at all if it turns out we’ve been wrong for decades and modern behvaiour isn’t recent at all.
Is the selective value of inteligence compared to other traits identical in all environments econuntered by Homo Sapiens? Remember we may already have some evidence that sometimes inteligence may not be that usefull for hominids depending on how we interpret the fossiles of Homo Floresiensis. Could this also be true of Homo Sapiens population as well?
The answers to these two questions would tell us how likley it would be to see these differences appear and how noticeable they may be in the time window current biology estimates we have for differences between populations to occur.
If genetic differences in intelligence could not be relevant to reproductive success within a single generation it is difficult to see how human intelligence could have evolved.
I don’t know if the email you quoted (thanks for that, BTW, it’s a treat to have access to a primary source without needing to chase it down) is racist, but it does sound very ignorant to me. It makes unwarranted inferential leaps, e.g. from “skin and hair color are definitely genetic” to “some part of intelligence is genetic”, omitting the very different length of developmental chains leading from genes to pigmentation on the one hand, and intelligence on the other.
Let’s be careful here. The letter does not assert baldly that “some part of intelligence is genetic”. Rather, the letter asserts that some evidence “suggests to me that some part of intelligence is genetic”.
Furthermore, that particular inferential leap does not begin with the observation that “skin and hair color are definitely genetic”. Rather, the inferential leap begins with the claim that “Women tend to perform less well in math due at least in part to prenatal levels of testosterone, which also account for variations in mathematics performance within genders.” Therefore, at least with regards to that particular inference, it is not fair to criticize the author for “omitting the very different length of developmental chains leading from genes to pigmentation on the one hand, and intelligence on the other.”
[ETA: Of course, the inference that the author did make is itself open to criticism, just not the criticism that you made.]
I say all this as someone who considers Occam to be pretty firmly on the side of nongenetic explanations for the racial IQ gaps. But no progress in these kinds of discussions is possible without assiduous effort to avoid misrepresenting the other side’s reasoning.
I’m more directly disturbed by the bias present in your exposition: “perfectly reasonable”, “merely expresses agnosticism”, “well documented”, “harsh and inaccurate”.
Starting off a discussion with assorted applause lights and boo lights strikes me as unlikely to lead to much insight.
What would be likely to lead to useful insight? Making use of the tools LessWrong’s mission is to introduce us to, such as the applications of Bayesian reasoning.
“Intelligence has a genetic component” strikes me as a causal statement. If it is, we ought to be able to represent it formally as such, tabooing the terms that give rise to cognitive muddles, until we can tell precisely what kind of data would advance our knowledge on that topic.
I’ve only just cracked open Pearl’s Causality, and started playing with the math, so am still very much an apprentice at such things. (I have my own reasons to be fooling with that math, which are not related to the race-IQ discussion.) But it has already convinced me that probability and causality are deep topics which it’s very easy to draw mistaken conclusions about if you rely solely on a layman’s intuition.
For instance, “the well documented IQ differences between groups” are purely probabilistic data, which tell us very little about causal pathways generating the data, until and unless we have either controlled experiments, or further data sets which do discriminate between the competing causal models (only very grossly distinguished into “nature” and “nurture”).
I don’t know if the email you quoted (thanks for that, BTW, it’s a treat to have access to a primary source without needing to chase it down) is racist, but it does sound very ignorant to me. It makes unwarranted inferential leaps, e.g. from “skin and hair color are definitely genetic” to “some part of intelligence is genetic”, omitting the very different length of developmental chains leading from genes to pigmentation on the one hand, and intelligence on the other. It comes across as arrogant and elitist as well as ignorant when saying “I think my babies will be geniuses and beautiful individuals whether I raise them or give them to an orphanage in Nigeria”.
It is not bad science to be on the lookout specifically for data that claims to be “scientific proof” of some old and demonstrably harmful prejudices, and to hold such claims to a higher standard. Just as we do hold claims of “scientific proof of ESP” to a higher standard—at least of scrutiny and replicability—than, say, claims of a correlation between apparel color and competitive performance. We have more reason to suspect ulterior motives in the former case than in the latter.
Dinnertime conversations between regular, even educated people do not contain probabilistic causal analyses. In the email Grace claimed something was a live possibility and gave some reasons why. Her argument was not of the quality we expect comments to have here at Less Wrong. And frankly, she does sound kind of annoying.
But that all strikes me as irrelevant compared to being made into a news story and attacked on all sides, by her dean, her classmates and dozens of anonymous bloggers. By the standards of normal, loose social conversation she did nothing deserving of this reaction.
I feel a chilling effect and I’ve only ever argued against the genetic hypothesis. Frankly, you should too since in your comment you quite clearly imply that you don’t know for sure there is no genetic component. My take from the reaction to the email is that the only socially acceptable response to encountering the hypothesis is to shout “RACIST! RACIST!” at the top of your lungs. If you think we’d be spared because we’re more deliberate and careful when considering the hypothesis you’re kidding yourself.
Sure. What I do find disturbing is how, knowing what she was doing (and who she was sending it to), the “friend” who leaked that email went ahead and did it anyway. That’s positively Machiavellian, especially six months after the fact.
However, I do not feel a need to censure myself when discussing the race-IQ hypothesis. If intelligence has a genetic component, I want to see the evidence and understand how the evidence rules out alternatives. I would feel comfortable laying out the case for and against in an argument map, more or less as I feel comfortable laying out my current state of uncertainty regarding cryonics in the same format.
Neither do I feel a need to shout at the top of my lungs, but it does seem clear to me that racism was a strong enough factor in human civilization that it is necessary, for the time being, to systematically compensate, even at the risk of over-compensating.
“I absolutely do not rule out the possibility [of X]” can be a less than open-minded, even-handed stance, depending on what X you declare it about. (Consider “I absolutely do not rule of the possibility that I will wake up tomorrow with my left arm replaced by a blue tentacle.”) Saying this and mistaking it for an “agnostic” stance is kidding oneself.
Since people are discussing group differences anyway. I would just like people to be a bit clearer in their phrasing.
Inteligence does have a genetic component. I hope no one argues that the cognitive difference between the average Chimpanzee and Resus monkey are result of nurture. The question is if there is any variation in the genetic component in Humans.
Studies have shown a high heritability for IQ, this dosen’t nesecarily mean much of it is genetic but it does seem a strong position to take, especially considering results from twin studies. A good alternative explanation I can think of, that could be considered equivalent in explanatory power, would be differences in prenatal environment beyond those controled in previous studies (which could get sticky since such differences may also show group genetic variation ! for example the average lenght of pregnancy and risks associated with postterm complications does vary slightly between races).
The question disscused here however is whether there are any meaningfull differences between human groups regarding their genetic predispositions towards mental faculties.
We know quite a bit from genetic analysis about where people with certain markers have spread and which groups have been isolated. Therefore the real question we face is twofold:
Just how really evolutionary recent is abstract thinking and other mental tricks the IQ test measures? The late advent of behavioral modernity compared vs. the early evidence of anatomically nearly modern could be considered for example. Some claim it was an evolutionary change following the well documented recent bottleneck of the Human species others say the advent of modern behaviour was a radical cultural adaptation to a abrupt environmental change or just part of a long and slow progress of rising population density and material culture complexity we haven’t yet spotted. Considering how sketchy the archeological record is we can’t be suprised at all if it turns out we’ve been wrong for decades and modern behvaiour isn’t recent at all.
Is the selective value of inteligence compared to other traits identical in all environments econuntered by Homo Sapiens? Remember we may already have some evidence that sometimes inteligence may not be that usefull for hominids depending on how we interpret the fossiles of Homo Floresiensis. Could this also be true of Homo Sapiens population as well?
The answers to these two questions would tell us how likley it would be to see these differences appear and how noticeable they may be in the time window current biology estimates we have for differences between populations to occur.
Note: This from Razib Khan’s site (Gene Expression), I’m reposting it here so you don’t need to hunt it down in my other post. http://www.gnxp.com/wp/wp-content/uploads/2010/02/PIIS096098220902065X.gr2_.lrg_.jpg
If genetic differences in intelligence could not be relevant to reproductive success within a single generation it is difficult to see how human intelligence could have evolved.
Group selection may help you imagine more.
Isn’t group selection largely discredited?
Let’s be careful here. The letter does not assert baldly that “some part of intelligence is genetic”. Rather, the letter asserts that some evidence “suggests to me that some part of intelligence is genetic”.
Furthermore, that particular inferential leap does not begin with the observation that “skin and hair color are definitely genetic”. Rather, the inferential leap begins with the claim that “Women tend to perform less well in math due at least in part to prenatal levels of testosterone, which also account for variations in mathematics performance within genders.” Therefore, at least with regards to that particular inference, it is not fair to criticize the author for “omitting the very different length of developmental chains leading from genes to pigmentation on the one hand, and intelligence on the other.”
[ETA: Of course, the inference that the author did make is itself open to criticism, just not the criticism that you made.]
I say all this as someone who considers Occam to be pretty firmly on the side of nongenetic explanations for the racial IQ gaps. But no progress in these kinds of discussions is possible without assiduous effort to avoid misrepresenting the other side’s reasoning.