« Let us presuppose.... | Gene Expression Front Page | What is a Bobo? »
November 12, 2003

IQ COMPARISONS (again)

In response to some comments on my earlier post, I should make it clear that I don't deny that there are IQ differences between nations, and I don't deny that they could be important (see the last para of my post). It is also quite possible (and on a priori grounds quite likely) that there are some genetic differences in IQ between populations that have evolved in partial isolation from each other and under different selective conditions. There are genetic differences in everything else, so why not IQ?

All I say is that:

(a) it is technically difficult to compare IQ in different countries, so the figures may not be robust, and

(b) even on a strongly hereditarian view of IQ, such as Jensen's, the level of environmental differences between developed and 3rd-world countries would be expected to produce a substantial IQ deficit. I don't think anyone has yet challenged my argument on this point.

Posted by David B at 02:18 AM




david b:
> I don't think anyone has yet challenged my
> argument on this point.

And nobody can challenge those points. If any reliable measures of group IQ-differences are ever to be made, they will be based on direct genetic information (after we know "the genes for intelligence").

I'm sure this will happen within 10 years (or what do you think?). Until then, the discussion about this is mere speculation.


Posted by: antibush at November 12, 2003 02:35 AM


If any reliable measures of group IQ-differences are ever to be made, they will be based on direct genetic information (after we know "the genes for intelligence"). . . Until then, the discussion about this is mere speculation.

No, this just isn't true. There are hierarchies of certainty, not this dichotomy of total God-like understanding and armchair guessing.

Unless behavioral genetics is just "speculation".

Posted by: Jason M. at November 12, 2003 03:02 AM


My issues on this question comes from the ways in which the scientific knowledge is mmapped onto existing ethnic and class antagonisms. To me, interpreting Hispanic immigration into California (much less race relations in Alabama) primarily in terms of low-IQ vs. high-IQ gene pools is to miss a large part of the point of what's going on.

Posted by: zizka at November 12, 2003 10:50 AM


Your assertions are really not backed up with any data, and the data are either around or could be had. For example it may or may not be "technically difficult to compare IQ in different countries": one would look at the relationship between grades on the big exams leaving high school and IQ test scores, asking if the relationships are the same for, say, Blacks and Whites. My point is that you can't get away with just asserting that there is difficulty.

In the same way you assert that "the level of environmental differences .... would be expected to produce a substantial IQ deficit." Expected by whom? What is the evidence? I don't know of any evidence of environmental effects. The folks who looked at the effects of the Dutch famine were pretty surprised to find absolutely nothing.

Henry Harpending

Posted by: harpend at November 12, 2003 11:02 AM


HH: expected by whom? Well, expected by Jensen, for one, presumably, as my argument mirrors one he has used on several occasions. If you see a fallacy in the argument, what is it? Do you disagree with Jensen that environmental conditions in most 3rd world countries are at least 2 s.d.s below western standards? Or do you think that heritability of IQ is even higher than he does?

Posted by: David B at November 12, 2003 12:10 PM


antibush wrote:

And nobody can challenge those points. If any reliable measures of group IQ-differences are ever to be made, they will be based on direct genetic information (after we know "the genes for intelligence").

I'm sure this will happen within 10 years (or what do you think?). Until then, the discussion about this is mere speculation.

And nobody can challenge those points. If any reliable measures of environmentally caused IQ-differences are ever to be made, they will be based on direct biological and neurological information (after we know the "neurons which are easily influenced").

I'm sure this will happen within 10 years (or what do you think?) Until then discussion about this is mere speculation.


Posted by: sporon at November 12, 2003 12:39 PM


Has the reverse been shown -- that lifelong malnutrition plus lifelong lack of education and intellectual stimulation has no influence on IQ testing? (How long was the Dutch famine). Or that semi-literacy or illiteracy resulting from bad or no education has no influence on IQ testing?

Education vs. stimulation: some illiterates with no formal education live in stimulating environments -- e.g., multilingual kids who grow up running errands in a port city, changing currency in their heads, etc. Contrast: girls who spend their days grinding corn and pounding laundry on the rocks.

Posted by: zizka at November 12, 2003 12:55 PM


David B wrote:
>HH wrote
>>expected by whom?
>Well, expected by Jensen, for one, presumably, >as my argument mirrors one he has used on >several occasions. If you see a fallacy in the >argument, what is it? Do you disagree with >Jensen that environmental conditions in most 3rd >world countries are at least 2 s.d.s below >western standards? Or do you think that >heritability of IQ is even higher than he does?

The environment may or may not be 2 SDs worse, but we have no evidence that those environments affect IQ test scores. They may, but there is not evidence. When one does heritability estimates one gets a term that is either called "error" or "environment", and it can include anything including dominance and epistasis, developmental randomness, test-retest error, etc.

BTW I left out a crucial phrase in the first paragraph of my previous post: I said one could compare the predictive validity of IQ test scores among Black and White kids, and I should have added "in African countries where they have gone to the same schools". Plenty of such data to be had from Namibia, Botswana, Zimbabwe, etc.

Henry

Posted by: harpend at November 12, 2003 01:42 PM


A general comment: Human behavior is vastly complex so we, rightly, use a vast number of conceptual tools to explain why different people behave differently: e.g., age, language, religion, culture, ethnicity, health, nutrition, education, class, sexual orientation, parental upbringing style, height, metaphysical outlook, wealth, etc etc etc etc. These are all useful to one extent or another.

The crucial point about two categories popular on GNXP -- race and IQ -- is not that they are magic bullets that explain everything. They are simply currently underutilized conceptual tools. They explain only small fractions of all human behavior, but those small fractions are much bigger than the tiny fractions of the intellectual attention they are given at present. Their statuses in 2003 are analogous to the undervalued role that biological sex was given during the Dark Ages of Feminist Fundamentalism only a decade or two ago. Fortunately, the curtain of censorship has lifted significantly on discussing male vs. female differences since just, say, 1991-1992, and our ability to think sensibly about humanity has correspondingly enjoyed a small, but important, increase.

Posted by: Steve Sailer at November 12, 2003 06:03 PM


I think that if you go to Beijing or Shanghai you will find similiar "low-IQ / high-IQ" high/low crime ethnic contrasts without a racial difference. Possibly a dialect difference. Germans I know speak ill of the East Germans.

What I object to is your insistence on interpreting these situations primarily racially when there are historical non-genetic factors. Genetics is your stock in trade but you often seem historically uninformed, sometimes in principle (since all historical sciences are soft sciences to you).

In American politics you are taking sides in old ethnic antagonisms with a long history as such, which are in turn based on historical processes which go back centuries. I have never seen you display any interest, understanding, or very much knowledge of these factors.

In the case of immigration into California from Mexico and Central America, it goes along with a deliberate low-wage policy which goes back intermittently to the fifties. A whole economy of sub-minimum wage no-benefit part-time temp jobs has been created for people who will never have the chance to enter American life. This may support your anti-immigration view, but it undercuts the idea that IQ is the problem. How much difference would it make if the illegals speaking no English were from one of the poor provinces in China.

My argument was not specifically against the concept of race, but your explanation of everything by race and lack of knowledge or curiosity about the other factors.

Posted by: zizka at November 12, 2003 07:50 PM


My argument was not specifically against the concept of race, but your explanation of everything by race and lack of knowledge or curiosity about the other factors.

I'm not sure what information exactly we have omitted, but I think I have always tried to evaluate the case (especially re: black-white differences) using all the most reliable data available. I read sociological lit, and left-leaning stuff. I read things that explain differences from socio-historical and economic perspectives. I've read Oliver and Shapiro's Black Wealth, White Wealth, which theorized wealth differences underlie b-w social differences, and I read John Hoberman's (unconvincing) socio-cultural response to Entine, Darwin's Athletes. I read Ogbu (one of the few people who thinks to actually test [instead of assert] theories of caste). I read the left-oriented work of Harvard sociologist Christopher Jencks. (Such as his book on the test-gap, and this wonderful synopsis of it published in The American Prospect. And on and on.

The point is, is that I am familiar (as far as I know) with much of the data pool, and with alternative ways of looking at this data pool. I want to be familiar with it, because I'm trying to understand the problem. I say this with intellectual humility, if there are books you think we should read, or data you think we are ignorant of, you can recommend where to find it, and I will probably seek it out. But genetic theories have a lot going for them, in the way of parsimony (why certain differences hold relatively constant through cultural, geographic, and historical space), that alternative explanations for the same phenomenon (Diamond, Landes, Sowell) seem to lack. This is without even touching upon the psychometric and behavioral genetic evidence, which pretty much seals the deal in my mind.

--------------------------------------------------

[additional comment - added after comment close]

but your explanation of everything by race and lack of knowledge or curiosity about the other factors

I dunno, this just doesn't seem defensible, zizka. I mean how many people do you know who consider this question, with as diverse a range of sources as we consider? Do they even leave room for genetic possibilities? Do they leave room for, much less read Jensen, or do they just know all that psychometric stuff is "psuedo-science", just like Gould promised? Many liberals (a lot of them friends) who I discuss this with take an almost mystical approach to this topic (much in the same way some Jewish religious fundamentalists reject any attempt to understand the causes of the Holocaust in any way, as a way of excusing it ["There is no 'why' here"]. Also similar to the way some hyper-patriots condemned any consideration of root-causes for 9-11 as an apology for terrorism). All cultural considerations are considered reactionary and racist. Thomas Sowell is dismissed as a right-wing apologist. Even Ogbu's work, much of which is meant to show how racism affects blacks, is considered suspect, because even acknowledging any black dysfunction (in order to explain it) veers too close, in their minds, to "blaming the victim". That is where the real ignorance and lack of curiousity exists. If you think I'm straw-manning ask Atrios what he thinks of Sowell or McWhorter. Ask him what he thinks of psychometrics (I recall him comparing TBC to The Protocols of the Learned Elders of Zion). See what kind of a range of opinions and evidence is allowed to be discussed in his comment box. I read and sympathize with a wide range of authors with explanatory theories, but at the end of the day it is only theories which are (and can be) tested that matter.

Posted by: Jason M. at November 12, 2003 10:46 PM