All right, Hamster King, that’s perilously close to an accusation of trolling. As you know, such is not permitted in Great Debates. Please be careful going forward.
No warning issued.
Everyone: this is going to be one of those threads that can go off the rails in a damn hurry. I’ll do my best to monitor it but let’s try to keep the discussion civil and on a higher plane.
Ignoring all the other problems with the question, people do not have an IQ score midway between their parent’s scores. There’s no reason to expect they would, or to have that extend over any population group.
There are too many confounding variables to simply try to mathematically average IQs of source populations for offspring with genes mixed from both groups. For one thing, “race” is a poorly-defined construct biologically. It’s better to think of source population haplogroups if you are looking for biological definitions. Alternatively, you can let an individual self-define into a label such as “black” or “white” or “hispanic non-white” or whatever. These types of Self-Identified Race/Ethnicity (SIRE) groups do correlate with various biological haplogroups in the sense that various ways to measure proportion of genes derived from source populations tend to correlate with recent geographical ancestry, which in turn correlates with biological haplogroups. As humans came out of africa around 70kya and filled the planet, various descendant haplogroups have become separated from one another (including within africa itself, of course) to drive disparate gene frequencies by haplogroup.
For example, you are much more likely to see genes for fine light-colored hair and pale skin in someone who self-identifies as “white” than if they self-identify as “black.” Similar average phenotypic and physiological differences are noted for a myriad of “race” differences. What is really happening is that a self-identification for a given “race” group basically correlates with the odds that you’ll draw a gene for a given characteristic from the source population underpinning your SIRE group.
The polygenic nature of “intelligence” coupled with the difficulty in deciding how to measure it, along with any number of confounding nurturing variables, makes it silly to assume one can simply average source population IQs for mixed pairs to come up with an average IQ for the offspring. Worse, average IQ scores of US (and other) SIRE groups are are changing faster than it seems evolution can change them, according to James R Flynn and others who study IQ averages over the last hundred years or so.
IQ tests are not meaningless but neither are they the be-all and end-all for measuring mental function. Flynn’s data assumes IQ is a legitimate measurement, and that the current average black adult IQ in the US is 85; his extrapolation puts average black IQ closer to 70 or even lower in the last 50 or 60 years. That seems untenable to me, at least.
One would expect the IQs of mixed-SIRE groups to regress toward a mean of both SIRE groups, but I don’t think anyone thinks you can draw some sort of exact percentage out of that concept. For example, you can’t say, “Populations that are 80% west african haplogroups and 20% european haplogroups will have an average IQ 80% of the way toward the west african average IQ from the european average.” It’s not even impossible (theoretically) for a hybrid vigor to show up unexpectedly and to get a net benefit from mixing two populations.
Having said all this (poorly summarized, for which I apologize), I comment to the posters above who want to pooh pooh the entire “race-based” intelligence concept that the easiest way to destroy arguments promoting racial differences in intelligence is to produce studies which show there are no differences. To date, all studies–even ones accounting for nurturing influences–show exactly the opposite: we are products of our genes, and our average success at various skillsets correlates with the odds that we received genes from a given haplogroup set. If you want to argue that nurturing accounts for observed differences in “intelligence,” you need only to produce data that shows differences are erased when putative nurturing influences are erased. I am not aware of any such studies. For example, children of wealthy and well-educated black parents markedly underperform children of poorer and less-educated white parents on SAT and other standardized exams. Instead of critiquing studies which purport to show race-based differences, why not simply do better studies? The answer is that no studies of any kind have been able to show we don’t have race-based differences. Even at a SIRE group level, we do have measurable average differences.
What we need to promote, I think, is that for any given individual, group averages by SIRE or biological definitions don’t make a patoot of difference, and that our primary goal as a society should be to drive opportunity for success for everyone, regardless of mother nature’s disregard for fairness.
That’s a new one for me. lol. I guess we learn something everyday.
Well, if this is all simple, linear math (which it isn’t), it wouldn’t be half-way. It would be 1/5 the way between 70 and 100 ( that is, 76) since the average AA is ~ 1/5 white.
However, we don’t know how genes control IQ (much less “intelligence”, whatever that is) but it’s unlikely to be one gene or even a suite of genes that operate linearly like that. Not to mention that there are almost certainly environmental factors involved, too.
We’ve done this before, so no need to rehash all the old arguments (I’ll just rehash some of them), but most of this is false. No studies have eliminated all “nurture” influences (this might even be impossible, short of some sort of biosphere-like experiment), and there is no genetic evidence for any difference in intelligence among various groups. Further, there is more genetic diversity in sub-Saharan Africa than the rest of the world combined, and there are individual SSA populations that are more closely related, genetically speaking, to non-black populations than to certain other far-flung African populations- there is no possible grouping based on a genetic “family tree” that would encompass all sub-Saharan African populations that would not include much of the rest of the world as well.
There’s no more reason to believe that differences in outcomes today are based on genetics than the differences a hundred years ago, in which Jews and the Irish may have been at the bottom of the social pyramid with regard to test scores and crime stats. There’s nothing magical about test scores and other outcome differences today.
So, to sum up, the Genetic Explanation is just one hypothesis to explain disparate test-scores (and other disparate outcomes), but there is no genetic evidence (yet, anyway) that would support this Genetic Explanation.
Without addressing the implied substance of Mr. Pedant’s remarks, I’d ask the Board’s genetics experts whether the term “haplogroup” is commonly applied to any genotypical data other than on the Y-chromosome or mtDNA. If the answer is (as I think it will be) No, then I respectfully submit that Mr. Pedant is wrong and/or misusing his terminology. The correlation between sex-linked haplogroups and autosomal genotypes is relatively weak.
Independent of the lack of support for the primary argument about race, genetics and IQ, IQ measures are good measures of intelligence.
Isn’t it also not impossible that it could have an opposite effect and have a net negative (ie: in the other direction)? Or do small quantitative changes in genes affecting intelligence tend to produce large qualitative changes in IQs? I’m not familiar with any such theory.
I agree that better studies can be conducted to get closer to resolving the matter (nature or nurture). For example, if someone or some organization would be willing to fund an experiment involving four randomly selected infants from each of the major races, raise them all on a remote island in such a way that they have no awareness of racial prejudice or stigmatization, feed them all on the same diet, educate them all the same and, finally, when they reach a certain age, give them an IQ test. Control experiments like that would settle the issue once and for all. But that is not going to happen anytime soon.
Really? You think that experimental dssign would resolve anything?
Before explaining the effect, though, can we actually get a cite to support the claims made in the OP.
I was wondering the same thing. Among the many, many severe problems with the design is the miniscule sample size.
And I don’t even want to touch how a randomly selected individual could even be taken as a representative member of each of the “major” races.
Science it certainly is. But good (or even below average) science it most certainly is not.
The OP purports to merely be repeating the arguments made by Rushton et al to see how they stand up to the least scrutiny.
Maybe that’s a fine hair to split, but there it is.
The best ways to isolate nature / nurture analysis is with the various datasets of separated twins. I am not aware of a dataset that has a significant number of African Americans in it.
http://www.brookings.edu/views/papers/dickens/20060619_IQ.pdf
The number isn’t 85, hasn’t been for some time now. It’s closer to 90.
I’m sorry. I forgot about the possibility of prenatal mental defects caused by pollution or bad nutrition. Well there are ways of controlling for things like that.
Regarding the sample size, I was only giving a brief example – a small model – of an experiment that would resolve the issue. Of course the actual experiment would have to have a much larger sample size.
I’m actually surprised that some people are asking for citations of those IQ figures I gave in the OP. I thought they were pretty much common knowledge in the race-iq discourse.
Actually, the phenomena of epigenetics is raising some interesting questions about how the environment (nurture) impacts one’s genetics (nature). We’re accustomed to seeing these two things as independent factors, when really it is likely they shape one another.
The number may have come from Herrnstein and Murray and may have even been accurate for the year they wrote the book.
My point was that reliance upon The Bell Curve for any extrapolation from that data, (from a belief in the magical g to comparisons of groups based on disparate and uncontrolled test groups), can only result in junk. People who criticized The Bell Curve dismissed it on the grounds of general incompetence. I am not sure how many people addressed particular numbers, but there was no point, given the shoddy methodology employed.
In Kenya among tests for rural schoolchildren, IQ grew by 20 points between 1984 and 1998.
I don’t recall the exact numbers, and they were adjusted due to the flynn effect but I think IQ went from the high 70s to the low 90s during that period. So whatever Kenya did to achieve that should attempt to be replicated. It was supposedly a mix of better nutrition, parenting, environment, etc.
Kenyan DNA didn’t change during that transition period, it was probably all environment.
Look at SAT prep courses. They raise a student’s average score by about thirty points. They obviously do not do this with genetic modification. They just teach students test taking skills.
An IQ test is a written test. Setting aside innate intelligence, a person who’s familiar with the procedure will do better than a person who isn’t. As the joke goes, an IQ test is a way to measure how good a person is at taking an IQ test.
So whenever I hear about a significant disparity between the test scores of two groups of people, my first question is going to be to look at their social and educational background and see if they have different amounts of test-taking experience.
Which IQ test are you referring to? Which subscale(s) are you referring to as “written”?