Friday, February 25, 2011

More on Clark's model

Upper-caste Indians. Clark’s model works poorly in State societies where class divisions are rigid and where different classes operate according to very different rules.

In my last post, I discussed Ron Unz’s essay on selection for intelligence in East Asian societies. This paper, as its own author points out, makes the same point that Gregory Clark has made with respect to England.

In sum, State societies create a new environment of natural selection. The Big Man goes from hero to zero (unless he is part of the tiny ruling elite). For most people, the road to success is the market economy, and such success requires a special behavioral package:

1. abandonment of violence as a means to resolve disputes and increase personal wealth;
2. ability to plan ahead and save for tomorrow
3. general trade-related skills, notably numerical and text processing

Over time, economic success would have translated into demographic success. This nascent middle class would have grown in number, with downwardly mobile descendents spreading into the lower classes and gradually replacing them. Eventually, they and their heritable characteristics would have come to dominate the entire gene pool.

The problem with this argument is that many State societies do not seem to have followed this route. As Steve Sailer notes, “Ancient Egypt would seem like a similarly "orderly, stable, and advanced" peasant society, but the outcomes don't seem very similar” [to those of China and England].

Well, Egypt was not a closed gene pool. Its middle class may very well have been demographically successful, and their descendents may very well have spread downward into the lower classes. But that’s where the similarity ends. There was also a steady influx of foreign slaves into the lower classes (and even into the middle and upper classes through concubinage).

But, still, one can think of other examples. What about India? The main problem there would have been the caste system: the effects of natural selection on one caste could not have easily spilled over into the general population. Yes, one could change caste, but it was not an easy option.

To varying degrees, this problem would have happened elsewhere. In addition, rigid class boundaries often coincided with ethnic boundaries, the middle class being dominated by minorities of one sort or another.

What about England? Interestingly, Gregory Clark (2009) does point to social mobility as being a key element of his model. To measure social mobility, he looked at rare surnames from two points in history, 1600 and 1851, and the socioeconomic status of their bearers:

How do the descendants of these two groups look in terms of socioeconomic status by 1851? Surprisingly there seems to be almost complete regression to the mean. Table 9 shows some measures of the socioeconomic status for a sample of adult men of both name groups, taken from the names with the less frequent occurrences. While those descended from the rich show a slightly greater percentage in the top socio-economic groups, that result may well be sampling error. And at the bottom of the socio-economic scale, there are more of the descendants of the rich among “laborers” than there are descendants of the indicted.

… This implies both great downward mobility among the descendants of the rich, and modest upward mobility among the descendants of the indicted.

This can be illustrated with particular names. Clark, for example, denoted in the middle ages anyone performing clerical work, including the minor orders of the clergy who were allowed to marry. Since literacy was extremely limited in medieval England clark was thus originally an upper class name. But by 1600 0.7 percent of the indicted bear this surname, as many as among rich testators (0.63 percent). Of the 11 indicted Clarks in my sample, 7 had their occupation listed as laborer, thus illustrating the downward mobility of the medieval educated elite. There was also sign of upward mobility. Cook in the middle ages would likely not denote someone of particular wealth or status. By 1600, however it was the surname of 1.3 percent of the richest testators. Among the seven rich Cooks, five were described as Yeomen and one as a Gentleman. Even medieval and early modern England was thus a very fluid society, with families moving up and down the social scale across each generation

Thus, lack of social mobility would prevent Clark’s model from working. A related problem is that the lower classes tend to operate according to a different set of rules. A successful ability in a middle-class context might be maladaptive in another context. This is particularly the case where the lower classes are not simply less successful individuals, and where they live under a very different legal and economic regime.

Serfdom comes to mind. In this, England was lucky. The Black Death had greatly weakened serfdom, and the upper classes were unsuccessful in their attempts to re-establish it. The conditions were thus in place for the demographic revolution described by Clark.


Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World, Princeton University Press, Princeton and Oxford.

Clark, G. (2009) The indicted and the wealthy: surnames, reproductive success, genetic selection and social class in pre-industrial England,

Unz, R. (1980). Preliminary notes on the possible sociobiological implications of the rural Chinese political economy, unpublished paper.

Friday, February 18, 2011

East Asian intelligence

Pupils studying for the Chinese civil service exam.

One conundrum of human biodiversity is the high mean IQ of East Asians, specifically Chinese, Koreans, and Japanese. On average, they outclass all other human populations on IQ tests, which were originally designed by and for Europeans. This intellectual success is matched by the economic success not only of East Asian societies but also of their overseas communities, often in the face of severe discrimination (Hsu, 2011; Unz, 1980).

Recently, people have been seeking the cause in the exam culture of East Asian societies, e.g., “tiger moms” who push their children to prepare, prepare, and prepare for success on school exams.

This exam culture is not recent. Its roots can be traced back to Confucius and the introduction of the imperial civil service exam, first in China and then in other East Asian societies.

Confucius is credited with organizing China's first educational system and setting up an efficient administration system, based on the careful selection of a bureaucracy that helped the emperor and other leaders rule. Members of the bureaucracy were trained in special schools and chosen for their jobs based on their the proficiency on a civil service exam that tested their knowledge of Confucian texts. Before Confucius's time the only schools in China were ones that taught archery. (Hays, 2008)

Emperor Wu of Han started an early form of the imperial examinations, in which local officials would select candidates to take part in an examination of the Confucian classics, from which he would select officials to serve by his side. Beginning in the Three Kingdoms period (with the nine-rank system in the Kingdom of Wei), imperial officials were responsible for assessing the quality of the talents recommended by the local elites. This system continued until Emperor Yang of Sui established a new category of recommended candidates for the mandarinate in AD 605. For the first time, an examination system was explicitly instituted for a category of local talents. This is generally accepted as the beginning of the imperial examination system. (Wikipedia - Imperial examination)

East Asian exam culture and Confucianism

Was this factor strong enough to raise the mean level of intelligence? One objection is that the Chinese civil service exam was only partially adopted by Korea and Japan. Yet mean IQ is similar in all three societies.

This objection ignores the broader emphasis on education in all East Asian societies. China, Korea, and Japan have long been "exam cultures," even if we exclude the civil service exam. This exam grew out of values that were embedded in Confucianism and present throughout East Asia:

Confucius regarded government and education as inseparable. Without good education, he reasoned, it was impossible to find leaders who possess the virtues to run a government. "What has one who is not able to govern himself, to do with governing others?" Confucius asked. Under Confucianism, teachers and scholars were regarded, like oldest males and fathers, as unquestioned authorities.

The basic principal behind Confucian education is that if you work hard, endure and suffer as a young person you will reap rewards later in life. The strategy of Confucian education, used in China for centuries, is to memorize the moral precepts in the hopes that they will rub off and improve the character of the person who memorizes them and makes him or her more moral. Teachers have traditionally been held in high esteem and their power and control has been regarded as almost absolute.
(Hays, 2008)

Did only a small minority participate in the Confucian exam culture?

Candidates who passed all three levels of the civil service exam, the "Mandarins," were a tiny minority. This leads to an objection raised by Greg Cochran: the Confucian exam culture was too limited in extent to have any selective impact.

This objection implies that social benefits went only to the Mandarins. Yet social benefits accrued even to people who passed the first level (i.e., at the local prefecture). Such passage provided exemption from labor service and corporal punishment, government stipends, and admission to upper-gentry status (Britannica, 1998).

The base of the exam pyramid broadens even further if we include everyone who prepared for the first level. If the 3rd level graduates were a fraction of the 1st level ones, the latter were likewise a fraction of all Chinese children who studied with a view to taking the exam. Test-taking, selection, and elimination occurred even at this early family stage:

Preparations for the test usually began around age five when young boys were taught to bow respectfully and recite lines from classical texts. The most promising teenagers were sent to study under masters in the Chinese capital. They were taught poetry, essay writing and Confucian scholarship. (Hays, 2008)

One could object that this system, though theoretically open to everyone, was biased toward the middle and upper classes. The poor were underrepresented, and yet they made up the bulk of China’s population:

The notion that the Confucian system was based totally on merit and lacked a hereditary element is not true. Children of merchants, landowners and families with money had an advantage in that their parents could hire tutors to teach them how to properly write Chinese characters and study Confucian texts. Once they attained their position, Confucian gentlemen made sure their sons studied the classic and was prepared for the exams. There are similarities in this respect with European societies. The main difference, however, is that Europeans who studied Latin and classical literature were, until the end of the Middle Ages, mostly monks who abstained from reproduction. (Hays, 2008)

China’s poor, however, were a population sink. As Unz (1980) points out: “In each generation, the poorest 10-15% of the population either failed to reproduce or produced only a negligible fraction of the successor generation.” This point emerges in ethnographic accounts:

How could any man in our village claim that his family had been poor for three generations? If a man is poor, then his son can’t afford to marry; and if his son can’t marry, there can’t be a third generation. (Crook and Crook, 1959)

Further down the economic scale there were many families with unmarried sons who had already passed the customary marriage age, thus limiting the size of the family. Wong Mi was a case in point. He was already twenty-three, with both of his parents in their mid-sixties; but since the family was able to rent only an acre of poor land and could not finance his marriage, he lived with the old parents, and the family consisted of three members. Wong Chun, a landless peasant in his forties, had been in the same position when he lived with his aged parents ten years before, and now, both parents having died, he lived alone. There were ten or fifteen families in the
village with single unmarried sons
. (Yang, 1958)

The ranks of the poor were constantly being replenished with downwardly mobile individuals from the middle and upper classes.

Security, relative comfort, influence, position, and leisure [were] maintained amidst a sea of the most dismal and frightening poverty and hunger — a poverty and hunger which at all times threatened to engulf any family which relaxed its vigilance, took pity on its poor neighbors, failed to extract the last copper of rent and interest, or ceased for an instant the incessant accumulation of grain and money. Those who did not go up went down, and those who went down often went to their deaths or at least to the dissolution and dispersal of their families. (Hinton, 1966)

This all sounds a lot like the scenario described by Clark (2007) for English society before the 20th century. Ron Unz says as much in a personal communication:

Overall, the model is pretty similar I think to what that Clark fellow wrote about England. However, I think the degree of genetic pressure in each generation was enormously greater, fenjia [division of land among sons] caused automatic downward mobility each generation, and I think the system remained in place for several times longer than the few centuries Clark claims for England.

Clearly, the higher mean IQs of East Asians cannot be solely or even mainly attributed to the Confucian exam culture. The main cause was the establishment of a State society, its monopoly on the use of violence, and its creation of an orderly, rules-based society. Reproductive success depended on being able to play by the rules.

The rules, however, were formalized in the teachings of Confucius. One’s knowledge of these teachings became a proxy for one’s ability to succeed in East Asian society. More generally, it became a proxy for intellectual performance, all the more so because one had to memorize Chinese characters (a minimum of 10,000 for functional fluency) and understand an archaic form of the language. Thus, Confucian exam culture might explain some of the differences between European and East Asian intellectual performance.

But why did this exam culture develop in East Asia and not in Europe? Greco-Roman society similarly valued study of classical literature and proficiency in archaic Greek and Latin (as opposed to the contemporary Koine Greek and Vulgar Latin). With the advent of Christianity, however, classical “pagan” literature became viewed with suspicion. Emphasis shifted toward study of the Bible, and such study usually involved entry into celibate religious orders. Insofar as academic success was linked to heritable predispositions, the overall impact of natural selection would have been negative.


Britannica (1998). Chinese civil service, vol. 3.

Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World, Princeton University Press, Princeton and Oxford.

Crook, D. and I. Crook (1959). Revolution in a Chinese Village, Ten Mile Inn, Routledge and Kegan Paul.

Hays, J. (2008). Chinese Education,

Hinton, W. (1966). Fanshen, Monthly Review Press.

Ho, P.T. (1959). Aspects of social mobility in China, 1368-1911, Comparative Studies in Society and History, 1: 330-359.

Hsu, S. (2011). (2011). Sociobiological implications of the (historical) rural Chinese economy? Information Processing, February 16.

Unz, R. (1980). Preliminary notes on the possible sociobiological implications of the rural Chinese political economy, unpublished paper.

Yang, C.K. (1958). A Chinese Village in Early Communist Transition, Cambridge, Mass.: MIT Press.

Friday, February 11, 2011

Rethinking intelligence and human geographic variation

Children making pillow lace for a home workshop. Germany, 1847. During the early stages of Europe’s market economy, successful entrepreneurs would expand their workforce by having more children.

There are several obstacles to our understanding of geographic variation in human mental performance.

First, the subject is taboo. When people do discuss it, they often resort to euphemisms, “code words,” and the like. The result, all too often, is confused thinking. When ideas cannot be expressed clearly, the resulting theory is likewise obscure.

Second, there has been a tendency to imitate the physical sciences by seeking a “unified theory,” i.e., Phil Rushton and r-K life history theory (Rushton, 2000), Ed Miller and parental investment (Miller, 1994), etc.

There is indeed a unified theory. It’s called evolution by natural selection. But natural selection acts in many different ways in many different environments. This could hardly be less true for our species, which has had to adapt to a wide range of physical environments—from the tropics to the arctic—and a wide range of cultural environments—from nomadic hunter-gatherers to socially complex urban civilizations.

Third, until recently, human evolution was supposed to have happened a long time ago, surely no later than the Pleistocene. Our mental traits, like all other traits, are thus a product of ancient environments, when humans lived by hunting and gathering. As for all of the later developments—agriculture, civilization, literacy, State societies, class stratification—these things have shaped us culturally but not genetically.

Well, they have shaped us genetically. Changes to the human genome have accelerated over the past 40,000 years and especially over the past 10,000 years (Hawks et al., 2007). We are a young species. Human nature—or rather natures—is largely post-Pleistocene.

Recent selection

Thus, one theoretical model cannot account for all or even most variation in cognitive capacity among present-day humans. The higher IQ of East Asians, for example, almost certainly came about during historic times and was probably favored by the widespread use of exams as a means of social advancement. Likewise, the higher IQ of Ashkenazi Jews and other European populations is probably post-medieval in origin and driven by the high fertility of successful entrepreneurs, particularly those in cottage industries who could expand their workforces only by having larger families (Frost, 2007).

Another relevant factor is the rise of the State, particularly its monopoly on violence (Frost, 2010). This is discussed with respect to English society in Gregory Clark’s Farewell to Alms. Clark (2007) argues that the slow but steady demographic expansion of the English middle class from the 12th century onward gradually raised the population mean for predispositions to non-violence, deferment of pleasure, and other future-oriented behavior. Although the embryonic middle class was initially a small minority in medieval England, its descendants grew in number and gradually replaced the lower class through downward mobility. By the 1800s, its lineages accounted for most of the English population.

This natural selection came to an end with the Industrial Revolution. Previously, successful entrepreneurs expanded their workforce primarily by having larger families. With the decline of cottage industry and the commodification of labor, it became possible to hire workers on a large scale. Henry Ford was a successful businessman, but his economic success did not translate into reproductive success. He had only one child.

No-so-recent selection

Even further back in time, neither Rushton’s model nor Miller’s fits all of the facts. Yes, cognitive capacity does seem to show some kind of relationship with family structure, specifically low incidence of polygyny and high paternal investment. I am not convinced, however, that this relationship is best understood in terms of K-type versus r-type reproductive strategy.

Today, most of the human gene pool is derived from populations that only 15,000 years ago were confined to the northern tier of Eurasia. These populations have since expanded southward into temperate and even tropical Eurasia, as well as Oceania and the Americas. In the process, they have displaced other populations that were nonetheless better adapted in terms of climate and ecology.

What was their competitive advantage? It could not have been a K-type reproductive strategy. If we look at present-day hunter-gatherers from the northern arctic and sub-arctic, we find that they pursue a moderately r-type strategy despite high levels of paternal investment. Traditional Inuit, for instance, have short inter-birth intervals, with menstruation being a rare occurrence.

The competitive advantage seems to involve three characteristics of ancestral northern Eurasians:

1. A predictable yearly cycle, which favored the ability to plan ahead and make future decisions in the present. Indeed, early modern humans had more complex tools and weapons at arctic latitudes than at tropical latitudes, apparently because of the yearly cycle of resource availability: “Technological complexity in colder environments seems to reflect the need for greater foraging efficiency in settings where many resources are available only for limited periods of time.” (Hoffecker, 2002, p. 135)

2. A low incidence of polygyny, which reduced male-male competition for mates and the consequent disruptive effects on social organization.

3. A high level of paternal investment in the family, which in turn emancipated women from food provisioning and enabled them to develop a ‘family workshop’ of garment making, structure building, food processing, etc. (Kelly 1995, p. 262-270).

These northern Eurasians were thus mentally pre-adapted, despite their simple social organization, for later technological developments, even though such developments were possible only in more southern environments for which these populations were less ecologically adapted. It is perhaps no surprise that they were able to expand southward into the temperate and tropical zones, eventually peopling almost all of Eurasia, Oceania, and the Americas.


Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World, Princeton University Press, Princeton and Oxford.

Frost, P. (2010). The Roman State and genetic pacification, Evolutionary Psychology, 8(3), 376-389.

Frost, P. (2007). Natural selection in proto-industrial Europe, Evo and Proud, November 16.

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution, Proceedings of the National Academy of Science USA. 104, 20753-20758.

Hoffecker, J.F. (2002). Desolate Landscapes. Ice-Age Settlement in Eastern Europe. New Brunswick: Rutgers University Press.

Kelly, R.L. (1995). The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, D.C.: Smithsonian Institution Press.

Miller, E. (1994). Paternal provisioning versus mate seeking in human populations, Personality and Individual Differences, 17, 227-255.

Rushton, J.P. (2000). Race, Evolution, and Behavior, 3rd edition, New Brunswick, NJ: Transaction.

Friday, February 4, 2011

Five years later ... still no study

Tay-Sachs child. In the 1980s, there was concern that Tay-Sachs heterozygotes might be mentally handicapped to a lesser degree.

What is the evidence that Tay-Sachs heterozygotes have higher-than-normal intelligence? To date, it is inferential. Ashkenazi Jews have unusually high incidences of several genetic diseases that have similar effects on neural tissue.

[…] the sphingolipid mutations look like IQ boosters. The key datum is the effect of increased levels of the storage compounds. Glucosylceramide, the Gaucher storage compound, promotes axonal growth and branching (Schwartz et al., 1995). In vitro, decreased glucosylceramide results in stunted neurones with short axons while an increase over normal levels (caused by chemically inhibiting glucocerebrosidase) increases axon length and branching. There is a similar effect in Tay-Sachs (Walkley et al., 2000; Walkley, 2003): decreased levels of GM2 ganglioside inhibit dendrite growth, while an increase over normal levels causes a marked increase in dendritogenesis. This increased dendritogenesis also occurs in Niemann-Pick type A cells, and in animal models of Tay-Sachs and Niemann-Pick. (Cochran et al., 2006)

Such “diseases” may have beneficial effects in the heterozygote state:

Heterozygotes have half the normal amount of the lysosomal hydrolases and should show modest elevations of the sphingolipid storage compounds. A prediction is that Gaucher, Tay-Sachs and Niemann-Pick heterozygotes will have higher tested IQ than control groups, probably in the order of 5 points. (Cochran et al., 2006)

That paper was written five years ago. Since then, no one has tested the IQ of Tay-Sachs heterozygotes in relation to non-carriers (after controlling for differences in ethnicity and socioeconomic status). Yet such a study would hardly be rocket science.

Ironically, this kind of study had been done previously, back in the 1980s. At that time, the concern was that Tay-Sachs heterozygotes might have lower IQ. The study was consequently set up to look for some kind of mental deficit:

Two groups of heterozygotes, one for metachromatic leukodystrophy (MLD) and the other for Tay-Sachs disease, were given a battery of neuropsychological tests, a standard neurological examination, and an EEG. Neurological and EEG findings were unremarkable for both groups. The MLD heterozygotes showed deficits in the neuropsychological tests involving spatial or constructional components, but not in tests involving language skills. The Tay-Sachs heterozygotes showed no consistent deficit on any component of the neuropsychological tests. (Kohn et al., 1988)

Yes, but how well did the Tay-Sachs heterozygotes do on those tests? The authors say nothing on that point, other than making the following observation:

However, there is a significant difference in the educational level achieved between the two groups. About two-thirds of the Tay-Sachs heterozygotes have education beyond high school, whereas only one third of the MLD heterozygotes have gone beyond high school. (Kohn et al., 1988)

It would be interesting to dig up the raw data and have it reanalyzed …


Cochran, G., J. Hardy, and H. Harpending. (2006). Natural history of Ashkenazi intelligence. Journal of Biosocial Science, 38, 659-693.

Kohn, H., P. Manowitz, M. Miller, and A. Kling. (1988). Neuropsychological deficits in obligatory heterozygotes for metachromatic leukodystrophy, Human Genetics, 79, 8-12.