« Science is for losers | Gene Expression Front Page | AFRAID OF GROWING OLD? »
July 22, 2003

Jewish smarts

Godless stated he wouldn't post about Jews, but I'm not him, so I feel no restraint. Someone referenced Greg Cochran's new essay How The Ashkenazi Got Their Smarts, so I feel obligated to link to it on Jerry Pournelle's site. Enjoy & discuss....

Update from Godless:

Well, if it's "Jeurasian" week at GNXP, I did have a few things to chip in. First is that the Cochran title is a reference to How the Leopard got his Spots...which is a self-referentially wry dig at sociobiological just-so stories. Nice wordplay there, GC. [1]

Second, just thought I'd link to this decent summary of Jewish genetics research up to 2002:

'The authors are correct in saying the historical origins of most Jewish communities are unknown,' Dr. [Shaye] Cohen [of Harvard University] said. 'Not only the little ones like in India, but even the mainstream Ashkenazic culture from which most American Jews descend.'.... If the founding mothers of most Jewish communities were local, that could explain why Jews in each country tend to resemble their host community physically while the origins of their Jewish founding fathers may explain the aspects the communities have in common, Dr. Cohen said.... The Y chromosome and mitochondrial DNA's in today's Jewish communities reflect the ancestry of their male and female founders but say little about the rest of the genome... Noting that the Y chromosome points to a Middle Eastern origin of Jewish communities and the mitochondrial DNA to a possibly local origin, Dr. Goldstein said that the composition of ordinary chromosomes, which carry most of the genes, was impossible to assess. 'My guess,' Dr. Goldstein said, 'is that the rest of the genome will be a mixture of both.'"

So the upshot is: Y chromosomal lineage from the Middle East, mtDNA from local mothers, and autosomes are yet to be determined. A founder population of males from the Middle East came in, married local females, and practiced endogamy once the community got large enough. Another large contributor to phenotypic resemblance is natural selection effects...in irony of ironies, the Nazis exerted strong selection pressure for blond haired, blue eyed Jews.

Clearly at least some autosomes will have characteristically Ashkenazi signatures, because some of the diseases common to the Ashkenazi are located on autosomes (that is, non-sex chromosomes).

Diseases seen more frequently in the Ashkenazi Jewish population

Bloom syndrome
Canavan disease
Cystic Fibrosis
Familial Dysautonomia
Fanconi anemia
Gaucher disease
Niemann-Pick disease
Tay-Sachs disease
[From Cochran]
BRCA2 (breast cancer risk genes)
torsion dystonia
Factor XI deficiency (clotting disorder)
non-classic CAH
familial Mediterranean fever
familial hyperinsulinism
familial hypercholesterolemi
glycogen storage disease VII
maple syrup urine disease
mucolipidosis type IV.

This stuff is also useful for fact-checking Cochran's essay, of course...though I'm too lazy to look up those pathways ;)

Now, a special treat for our GNXP readers:
Intellectual property bending uploads of two papers on the possibility of selection in Ashkenazim. My professional opinion? I think Risch is wrong, though he shouldn't be derided as PC. He did write this good piece last year on the reality of race.

Selection may have happened in Ashkenazim along with response - Disagreement: selection did not happen in Ashkenazim

[1] Why, thank you, GC...says GC ;)


I don't know buy the "lower visuospatial IQ" for Jews. Supposedly it comes from Daniel Seligman's book "A Question of Intelligence", but in my opinion it's unlikely that Jews would have anything like their incredible math dominance (see here) without high visuospatial skills, unless "visuospatial" means something different than "ability to rotate/visualize three dimensional objects".

Posted by razib at 09:41 PM

Add to the list: BRCA1, BRCA2 (breast cancer risk genes), torsion dystonia, Factor XI deficiency (clotting disorder), non-classic CAH, familial Mediterranean fever, familial hyperinsulinism, familial hypercholesterolemi, glycogen storage disease VII, pentosuria, maple syrup urine disease, and mucolipidosis type IV.

Posted by: gcochran at July 22, 2003 11:01 PM

If memory serves, every one of these mutations is autosomal. As for your point about paternal and maternal ancestries, read the paper more closely. The evidence for local founding mothers is strong in some groups but not for the Ashkenazi.

3/4ths of the Ashkenazi conenxin-26 deafness mutations are a common Middle Eastern mutation, one of the two factor XI mutations is shared with Asian jews, familial mediterranean fever is also found in Sephardic/Asian Jews and some other middle easterners.

Reasonable people estimate that around 80% of Ashkenazi ancestry is Middle Eastern and I have no trouble with that.

Posted by: gcochran at July 22, 2003 11:15 PM

Why is Israel's IQ only around 95?

And surely the higher IQ of Ashkenazi Jews can be explained by the practice of usury, which gave them more financial resources - thereby allowing the men to pick and choose the best local Slavic women. In other words, the high Ashkenazi IQ comes from Slavic genes.

Posted by: fredrik at July 23, 2003 05:07 AM

Even if your slightly lurid theory is basically correct, it begs the question of why the Slavs don't also have average IQs of 115. There might be some heterosis at work, but I still think the primary mechanism is brutal selection pressure for high intelligence plus probably some luck in terms of the genetic material that survived the early medieval population bottleneck. Of course once we have a better understanding of the genetics, we can find out whether the slavic and ashkenazi populations share some IQ-boosting variation that isn't found in other Jewish populations. I bet if they do find such a thing it will be more thoroughly distributed in ashkenazim, though.

Posted by: bbartlog at July 23, 2003 07:06 AM

One other disorder to add to your list: Crohn's disease. More common in caucasians and especially eastern european jews, and clearly has a genetic component.

Posted by: bbartlog at July 23, 2003 07:14 AM

One thing I've noticed about the racists who inhabit the web: They don't seem to understand even the basics of gene frequencies. It is possible that alleles for high intelligence in jews came from Poles. It is also possible that gene frequencies in a population can change over time. So it is possible that, no offense meant to anyone at all, the modern Greeks, absent any population inflow/outflow have very different gene frequencies than ancient Greeks, ditto Italians and Romans. I sometimes wonder if selection in Spain for risk-taking, violent men to go to the New World had much of an affect on the societies there.

Posted by: rob at July 23, 2003 08:27 AM

So it is possible that, no offense meant to anyone at all, the modern Greeks . . .

heh. who ever could you have in mind with such a disclaimer? :)

I sometimes wonder if selection in Spain for risk-taking, violent men to go to the New World had much of an affect on the societies there.

Hmm? Could be.

Posted by: Jason Malloy at July 23, 2003 09:03 AM

I still think Ashkenazic Jews have more 'German' then 'Slavic' genes, the mixed up Russian Jews notwithstanding. This is supported by classic physical anthropology.

I'd love to see a lot more research done on Jewish maternal DNA.

Posted by: Melnorme at July 23, 2003 09:37 AM

Mmmmmm, Slavic Jewish women......

Oh sorry, I'll stop drooling now over hot Russian Jewish chicks. (who, by the way, likely look more "Aryan" than you)

"usury"--funny. You wouldn't be on the internet if it weren't for "usury"--capitalism and the corporate structure wouldn't exist, and the world would be stuck in the Dark Ages.

Your local quadroon.

Posted by: David at July 23, 2003 10:28 AM

Admixture references:

Proc. Natl. Acad. Sci. USA, Vol. 97, Issue 12, 6769-6774, June 6, 2000
Jewish and Middle Eastern non-Jewish populations share a common pool of Y-chromosome biallelic haplotypes
M. F. Hammer et al

They get 23% +/- 7% from Y-chromosome data - less than 0.5% a generation.

back in 1982 Motulsky estimated 12.5% based on 18 classical genetic markers.
8. Motulsky, A. G. (1980) in Population Structure and Disorders, eds. Eriksson, A. W., Forsius, H. R., Nezanlinna, H. R., Workman, P. L. & Norio, R. K. (Academic, New York), pp. 353-365.

The right approach is use a bunch of microsatellites, but as far as I know this has not yet been done.

In terms of selection for IQ, I doubt if ancestry matters much: rather the social niche and low gene flow during the period of selection.

As for Risch: in general he must be wrong. Does he think that _drift_ increased Ashkenazi intelligence? You can't drunkards-walk your way to high IQ.

Posted by: gcochran at July 23, 2003 11:06 AM

You're wrong. Just about every Jewish IQ study shows visuospatial skills equal or lower to general European scores, and if you were really tuned in, you'd know this already. It fits the occupational/professional patterns. Moreover, it fits my selective model - they were financiers, not engineers or architects or bowmen. 'g' is one principal component of intelloigence, but visuospatial is another principal component, statistically independent.

If you select for 'g' you get less of other things: this is inevitable.

If you want high viz skills, I'd look at the Mongols.

Posted by: gcochran at July 23, 2003 11:22 AM

If you want high viz skills, I'd look at the Mongols.

Really? Like the Mongolians in Mongolia?
I know the eskimos have high visuospatial ability, and its often cited as an environmental influence on IQ, cuz the people citing it aren't very smart I guess. Why don't we use these hbd hotspots to draw students into engineering programs?
Not on the subject of this thread, but has anyone tested native americans for differences in gene frequencies? They are such a small population, probably from only a few waves of migration, that if one tribe/ethnic groups had a seriously different selective environment, we should be able to clearly see adaptations.

Posted by: rob at July 23, 2003 11:57 AM

Even if your slightly lurid theory is basically correct, it begs the question of why the Slavs don't also have average IQs of 115.
Why does the theory require modern Slavs to have an IQ of 115?

Oh sorry, I'll stop drooling now over hot Russian Jewish chicks. (who, by the way, likely look more "Aryan" than you)
Perhaps, but what's your point? Do you think Aryan means better?

You wouldn't be on the internet if it weren't for "usury"--capitalism and the corporate structure wouldn't exist, and the world would be stuck in the Dark Ages.
Again, what is your point?

Posted by: fredrik at July 23, 2003 01:35 PM

It doesn't, I suppose, if the theory is that the Jewish group got some intelligence-boosting genes from the surrounding population and then spread them throughout their community due to the selection pressures already mentioned elsewhere. The point I was trying to make is that you still have to have differences in selection pressure to explain the results, since the influx of local genes was not that large or long-lasting. Eventually genetic analysis will show whether the relevant variation is a lucky mutation, a complementary one from the local population, or a previously rare allele that got widely spread. Or some combination, as seems likely. But in all these cases the critical element is the selection pressure on the community.

Posted by: bbartlog at July 23, 2003 01:44 PM

The time depth of separation between sub-Saharan Africans and anybody else is much, much greater than the time depth of separation between Europeans and Middle Eastern populations. So the fact that AShkenazi SNPS are a lot closer to European than to African tells you nothing about the extent to which Ashkenazi ancestry is European or Middle Eastern.
Again, the mtDNA study (Thomas et al, AJHG, 2002) clearly shows that many Jewish populations had very few female ancestors. In some populations, like the Bene Israel from India or The Falasha, the mtDNA is clearly local.

But in the case of the Ashkenazi, mtDNA gene diversity is not particularly low (0.973 versus 0.988 among Germans). Thomas says "The pattern in Ashkenazic Jews is of particular interest. Despite the common opinion that this population has undergone a strong founder event, it has a modal haplotype with a frequency similar to that of its host population (9.0% vs. 6.9%), providing little evidence of a strong founder event on the female side." Nor can they tell _where_ Ashkenazi mtDNA comes from. Could be European, could be Middle Eastern.
On the topic of whether the Ashkenazi went through a bottleneck: neither Y-chromosome or mtDNA evidence suggests that. A bottleneck should decrease heterozygosity: I ran 88 loci out of Ken Kidd's database, and the Ashkenazi show no decrease at all; just the same as Russians or Druze. . The Finns show a decrease, Samaritans do - not the Ashkenazi.

Posted by: gcochran at July 23, 2003 04:26 PM

Do women from Aleppo Syria or Lebanon whose background is Christian, Aramaic/Syriac speaking recent ancestors, match mtDNA with most Ashkenazi women?

Posted by: Haddad at August 27, 2003 11:24 PM

I think the Thomas mtDNA study reported no exact matches between Ashkenazi mtDNA haplotypes and those from either Europe or the Middle East. This might very well be the result of genetic drift. At the mtDNA haplogroup level, there is apparently not enough discriminatory power to differentiate between European and Middle Eastern origins.

Haddad, is your reasoning that Aleppo is near Urfa (in Southeast Turkey), the biblical Haran, where Isaac and Jacob are reported to have taken cousins for wives? Interestingly, I have read that Muslim tradition identifies Urfa/Haran as the birthplace of Abraham.

Posted by: anon at October 5, 2003 12:05 PM