I had promised a couple of posts to bring those not necessarily familiar with HBD (i.e., Human BioDiversity) up to speed, but as I’ve not been getting around to those transitional posts, I’ve decided to skip that step and go on ahead.  So here is my first semi-original contribution to the world of HBD.  I’m going to take a look at the global distribution of average IQ scores in light of what we know about recent human evolution:

The two general things one notices is that first, average IQ scores cluster by race; that is European peoples all have average IQ scores around 100, Black Africans all around 70, Native Americans in the 80s, etc.

(IMPORTANT DISCLAIMER: It’s very important to note, for those who aren’t familiar with the statistical nature of IQ and other biological traits, is that in most every population, you have individuals at all levels of IQ; that is, there are plenty of smart Africans and plenty of not so smart Europeans, for example, even though the group averages are different.  By no means is every last European smarter than every last African.  A good analogy is height among the sexes.  Men are, on average, taller than women.  But that surely doesn’t mean that there aren’t tall women or that there aren’t short men, as one can clearly see here. Rather, the average of a trait tells you something about the frequency of that trait in a given population, as for example, there are many more men than women at a height of 6’2″, just as there are many more women than men at a height of 5’2″.  Group averages don’t necessarily tell you anything about a given individual.)

The second thing one notices is that average IQ increases with latitude. The traditional explanation, advanced by Richard Lynn and J. Philippe Rushton, is that racial differences in IQ are very ancient and go back to when humans migrated out of Africa into Eurasia and experienced the two previous ice ages.   The theory goes that cold winters select for higher intelligence as these conditions present challenges to survival not found in the tropics (such as storing food, keeping warm, or hunting big game).  As such, Eurasians evolved higher IQs than Africans to cope with these conditions.

(A note on the low average IQ of some groups, like sub-Saharan Africans and the Australian Aborigines: those scores are not in error. Multiple lines of evidence confirm these low scores, especially for Africans.  You may have heard that IQs ≤ 70 signifies mental retardation, but that is not exactly the case.  That number was selected as a somewhat arbitrary marker of retardation and is only obliquely related to true retardation in the sense that most people think of it.)

However, this theory is problematic for several reasons.  For one, the most northerly group, the Inuit, do not have the highest average IQ (though it is higher than most peoples further south).   Going further south in the New World, the Native North Americans do not have IQs comparable to Europeans, despite having lived in very similar climatic conditions.  Additionally, although the average IQ of Mongolia is listed as “unknown” on the map, IQ data from Mongolians living in China find that their average IQ is about 100, slightly lower than the Han Chinese who live further south.

As well, we have Gregory Cochran’s & Henry Harpending’s 2010 book The 10,000 Year Explosion: How Civilization Accelerated Human Evolution which demonstrated that the pace of human evolution has in fact been speeding up and has been since the advent of agriculture and the rise of civilization.  This suggests that much—or perhaps all—of the global distribution in IQ is recent, within the last few millennia or less.  To see this, let’s take a closer look at Europe (click to enlarge):

This is a map that I have made of the average IQ of the European nations based on data assembled by Richard Lynn and Tatu Vanhanen (also here) as well data from the PISA test as complied by A Reluctant Apostate and hbd* chick.

As we can see, there is considerable variation in the average IQ of the different parts of Europe, even within nations, as we see in the U.K., Spain, and Italy.  It is not at all true that the average IQ of all European populations is 100 (indeed that number is based on the average score of White British), because several populations score below it.  In general, we see a decline in the south and southeast.  But also standing out as distinctly low is Ireland (Lithuania’s score is probably spurious due to bad sampling).

But more interestingly, there seems to be a relationship between modern IQ levels and fairly recent history in Europe, since the Middle Ages, as we see here:

This is a map of the Hajnal line, which is the red line seen here.  West of this line, we have a unique pattern of marriage, as Wikipedia states:

West of this line, the average age of women at first marriage was 24 or more, men 26, spouses were relatively close in age, and 10% or more of adults never married. East of the line, the mean age of both sexes at marriage was earlier, spousal age disparity was greater and marriage more nearly universal.

The blue lines mark areas west of the Hajnal line where nuptiality was high and the above pattern didn’t necessarily hold.

But we see something more interesting when we overlay the Hajnal line with my map of European IQ, as hbd* chick did:

It can be seen that aside from Finland, these exempt areas correspond to the regions of lowered average IQ in Western Europe.

Hbd* chick has just written a glorious summary of why this may be so:

 the populations behind the hajnal line (i.e. the core of europe) are characterized by:

- late marriages (present since at least the early medieval period)
- small family sizes (nuclear or stem families versus extended families; also present since at least the early medieval period)
- higher average iqs, in general, than populations in the periphery of europe (see map)
- strong future time orientation, strong societal collectivism, strong preference for rules and order (Ordnung!), strong drive to succeed
- being more civic than populations in the periphery of europe

why?

well, maybe it’s just ’cause these populations are mostly germanic, or at least had a strong-ish germanic presence in their territory at some time in the past. maybe this is just an example of ice peoples who evolved high iqs and a lot of other neat traits ’cause they survived for a long time in adverse conditions.

but’s it’s hard to ignore how the Type A Personality areas of europe coincide with the hajnal line. at least, i find it hard to ignore. what happened behind the hajnal line?

at the risk of repeating myself (is there an echo in here?), what happened behind the hajnal line starting in the early medieval period was:

- changes in mating patterns (thanks to the church) from close relative marriage to more distant marriages, thus breaking down clans and tribes
- changes in the economic structure from whatever the h*ll went before (i have no idea) to manorialism
- changes in family structures (thanks to both the increased outbreeding and manorialism) from extended families to smaller nuclear or stem families

all of these would’ve changed the selection pressures on the populations in the areas where these practices were adopted.

A look at Italy illustrates what hbd* chick is talking about.  Here’s a map made by M.G.:

This is a map of consanguinity—that is, the rate of cousin marriage in Italy during the early 20th century.  This map shows that Southern Italians are fairly inbred (no, it’s not just Appalachia—more on that in a later post).  What’s more fascinating is that this map corresponds very well to my map of the average IQ of the various Italian provinces (and a whole host of other things, see M.G.’s blog). Italian stereotypes notwithstanding, especially of southerners, (including unfortunate and tragic demonstrations of Italian incompetence such as the recent cruise ship disaster), this shows the impact of more recent history on the evolution of Europeans.

What I found fascinating however is that people east of the Hajnal line still managed to evolve high IQs, such as the northern Slavs and the Finns.  People in this part of the world lived in a family system that was vastly different than that of westerners, as hbd* chick discusses here.  This caused Easterners to go down a considerably different social and political trajectory than did Westerners, and the system found in Eastern Europe was common in the rest of Eurasia going as far east as China.  This is displayed geographically here:

http://www.scribd.com/doc/72270658/Todd-s-Family-Systems-Map-1500-1900

However, despite these different systems, Easterners still managed to evolve IQs as high as Westerners.  Perhaps the IQ differences were pre-existing in the various populations, and I believe they most likely were—to an extent.  But one cannot escape certain recent historical correlates with modern average IQ levels.  For example, most of Southeastern Europe was dominated by the Byzantine Empire and then later the Turks.  One has to wonder if centuries of Muslim domination (and perhaps Muslim family structures) have had something to do with the lowered IQs of SE Europeans.

In any case, despite all this evidence for the recent evolution of modern IQ levels, there are the facts that average IQs cluster by race and correlate strongly with latitude (and skin color, but that’s not important at the moment).  If IQs reached their modern levels due to the fairly capricious events of history, then why are they not more haphazard across racial groups?  Why the consistency?  Perhaps this is evidence for a prehistoric origin.  Or perhaps this is result of several factors acting together to produce this effect, which is what I propose.

What if the reason that average IQ correlates with latitude is because climate imposes a ceiling on average IQ, not necessarily selects for a particular IQ level?  There is some evidence that head and brain size faced biological constraints in the tropics, due the problem of overheating.  And indeed, average head size increases with latitude, as can be seen here on this map of the average cranial capacities of indigenous populations:

Human variation in cranial capacity. Black, 1,450 cc and over; checkerboard, 1,400-49 cc; crosshatching, 1,350-99 cc; horizontal striping, 1,300-49 cc; diagonal striping, 1,250-99 cc; dots, 1,200-49 cc; white areas, under 1,200 cc (Beals et al., 1984)

Head size (and hence brain size) does indeed correlate with IQ, both between individuals and between groups, but it’s not a perfect correlation.  Far northerners, such the Inuit, have the largest heads (and largest brains) but not necessarily the highest IQs (as well, peoples with the smallest heads don’t necessarily have the lowest IQs, but as one can see from comparing this map to the first, the pattern is fairly solid overall).

Especially interesting are the Native Americans.  All Native Americans descend from Asians that crossed the Bering land bridge into the Americas.  Hence, all have had to  have been fairly cold-adapted people.  But Natives that hail from the American tropics seem to have regressed a bit in terms of head size and average IQ (as well as evolved darker skin).  I propose in that the tropics, the ceiling on average IQ is lower (not necessarily maximum individual IQ, as there are plenty of smart individuals who originate from tropical climates).  This may be due in part  to physiological constraints (heat stress), but also survivability.  Tropical living is overall easier than living in temperate and polar climates (in pre-modern times anyway, when we didn’t have the modern conveniences that makes winter living more bearable), since for one, food is typically available all year around.  Being exceptionally smart is not as much of a reproductive advantage (especially facing tropical diseases); or perhaps more accurately, being not very smart is not that much of a disadvantage.

Big brains come with drawbacks.  For one, larger heads make giving birth more difficult.  As well, brains are very metabolically expensive (your brain consumes about 25% of your resting calories).  Larger brains only make sense evolutionarily when these drawbacks are outweighed by the advantage increased intelligence affords.

Populations in different latitudes thus face different limits on how far they can advance in average IQ, both because of limits on those at the top, as mentioned, and because of the level of relaxation of limits on the bottom.  That is, even if the smartest members of society reproduce well, the society’s average IQ will not increase if the least intelligent individuals also reproduce as much.

For long periods of time in Europe, there was a strong limit on the bottom.  The highly stratified nature of European society meant that the poorest (and on average, least intelligent) individuals faced poor prospects for surviving and raising children.  Over time, that meant that almost all people in many European societies were descended primarily from the upper classes (which is why when many of you trace your ancestry back that far, you often find European nobility/royalty).

This is, by the way, an example of how I think this ceiling effect can operate: perhaps in lower latitudes, the constraints on those on the bottom were not as tight, and less intelligent people were able to survive and reproduce more.

However, the important point—and the part that I think has been missing from the puzzle—is that regardless of the ceiling given by climate and geography, it is not a given that a population will evolve to reach that ceiling.  Looking at Europe again, we can see that the Irish have a lower mean IQ than the rest of the inhabitants of the British Isles; they have not reached their region’s ceiling.  I contend that in pre-civilized times, no Europeans did.  The earliest Europeans (or perhaps more accurately, the earliest farmers) probably did not have the equivalent of an average IQ of 100 (though it was probably higher than modern Africans).   Instead, I contend that their IQs increased to their current levels through the effects of civilization, mostly during the Middle Ages, as described by hbd* chick’s hypothesis.

Over in Asia and the Americas, the same phenomenon likely applied.  The early Chinese/Koreans/Japanese probably did not have the same average IQs as their modern counterparts.  Their average IQs may have instead been more like the Mongolians or the Inuit, and only reached their modern levels through the effects of millennia of civilization (and likely heavy attrition in the lower classes).

So it seems that different latitudinal regions have different ceilings.  Broadly speaking, the IQ ceiling in sub-Saharan Africa is lower than it is North Africa, the Middle East, and South Asia, which is in turn lower than it is in Europe and Northern Asia.  Average IQ is allowed to rise somewhat in these northern regions, as the Mongolians, Siberians, Inuit, and Native Americans demonstrate. Civilization then takes it “all the way”.

Even in Africa, farming/herding Africans, such as the Bantu and Nilotic groups, seem to have higher average IQs (≈ 70) than the more primitive hunter-gatherers like the rainforest Pygmies or the Khoisan (≈54).  The evolutionary/societal changes brought about by farming (e.g., population growth), as well as tools/ideas acquired through outside trade, allowed various Bantu groups to at times form regionally powerful empires in Africa.  The fact that many relatively unrelated groups, such as the Bantu and Nilotic peoples, have the same low (by Western standards) average IQ (70) indicates the rather low IQ ceiling there.

Another interesting fact is that it seems that the global average IQ ceiling may be 100.  While East Asians are noted as having an average IQ of 105, their mental profiles (along with everyone along the Pacific Rim, from Mongolians, Inuit, Native Americans, and Australian Aborigines) is such that their visuospatial reasoning is stronger than their verbal reasoning.  This may be because the plain, featureless terrain of the Asian interior may have selected for better navigational ability, or perhaps these abilities developed over time considering that these people have had the longest migration distance out of Africa (or both).  In any case, the average verbal IQ of East Asians is around 100, whereas their visuospaitual IQ is about 109-110 (which averages to 105).  Could it be that, for whatever reason, average verbal IQ peaks at 100 globally?  (It is verbal reasoning that correlates most strongly to most of the life success measures in the modern world.)

Indeed, the only group to exceed an average verbal IQ of about 100 are the Ashkenazi Jews, who seemed to have evolved an average IQ of 110 (which is strongest in verbal and mathematic ability but weaker in visuospatial ability) during the middle ages in Europe.  But the interesting fact about the Ashkenazim is that they did not evolve this high IQ in isolation, but by living among gentile Europeans.  As the Europeans Jews were always a minority group in a larger population, they were free to specialize in cognitive occupations (or more accurately, forced to specialize, since they were often barred from many occupations) in a way no other population could (because every other society needed laborers, farmers and other people do the grunt work).  This may have allowed them to break through the ceiling imposed on all other peoples.  But even this may have come at a price, because the mutations that lend themselves to Ashkenazi intelligence came with the side-effect of genetic diseases.  This could be a result of the necessary inbreeding that Jews practiced based on their restrictions on marriage outside the group; the high levels of exogamy among other Europeans may have distributed deleterious mutations rather than allow them to accumulate as they seem to have done among the Ashkenazim.

Our friend Occam’s Razor leads one to prefer explanations that are as simple as we make them, but no simpler.