Saturday, September 22, 2012

On paternal age and IQ

Do older fathers have dumber children? For the past millennium, paternal age has been relatively high in Europe west of the Hajnal line. Yet, if anything, mean IQ is higher there than elsewhere. H/T to JayMan (source)

Greg Cochran has been running a series of posts on paternal age and IQ (here, here, and here). His argument is similar to the one he put forward for temperature and IQ:

1.   Because intelligence requires the correct functioning of large numbers of interacting genes, it is more sensitive to random mutations, perhaps more so than other genetically influenced traits.

2.   The higher the mutation rate, the greater the likelihood that something will go wrong. Mean IQ should thus be lower in a population with a higher mutation rate.

3.   As men grow older, they accumulate mutations in the DNA they pass on to their offspring. Children born to older fathers should thus, on average, be more mentally deficient.

4.   Mean IQ should thus be lower in populations where men begin to reproduce at an older age. This is notably the case in highly polygynous populations, such as the ‘female farming’ peoples of sub-Saharan Africa, Papua-New Guinea, and Australia, where women can more easily provide for themselves and their children without male assistance. In such societies, it costs less for a man to take a second wife and, typically, it is the older men who monopolize the women.

This argument is plausible. Polygyny does increase paternal age, as Pierre van den Berghe notes:

In most human societies, males marry later than females, partly because their maturation is slower, partly because older males with more resources take more than their fair share of nubile women. This is especially true in polygynous societies. Typically, the more men are polygynous in a given society, the greater the age difference between husbands and wives. Indeed, given an approximately equal sex ratio, this mean difference in age of marriage between males and females is the main factor making extensive polygyny possible. (van den Berghe, 1979, p. 51)

How large is this sex difference? Among the Nyakyusa of southern Tanzania and northern Malawi, the difference is said to be ten years or more:

[…] there is a difference of ten years or more in the average marriage-age of girls and men, and it is this differential marriage-age which makes polygyny possible. A legal marriage is effected by the transfer of cattle from the groom to the bride’s father, and the system is linked with a late marriage-age for young men and with a privileged position for the older men. Many men have not cattle with which to marry until they are well over 25, while their fathers may be able to afford more than one wife. (Wilson, 1950, p. 112)

In their review of polygyny in sub-Saharan Africa, Pebley and Mbugua (1989), found that polygynous households were headed by men who were, on average, in their late 30s to their early 50s. It is worth remembering that these figures come from the late 20th century, when male life expectancy was much longer than it had been earlier. Polygynous men were now living to much older ages. This point applies to Cochran’s observation that mean paternal age is 47 in rural Gambia (Cochran, 2012c). Not so long ago, most men died before their mid-forties  ...

It is also worth remembering that marriage is not the same thing as reproduction. In highly polygynous societies, men often begin reproducing before marriage:

The temporary celibacy of young men in polygynous societies is rarely absolute, however. While it often postpones the establishment of a stable pair-bond and the procreation of children, it often does not preclude dalliance with unmarried girls, adultery with younger wives of older men, or the rape or seduction of women conquered in warfare (van den Berghe, 1979, pp. 50-51)

But these are not the main shortcomings of Cochran’s argument. The main one is the existence of another culture area where men typically did not marry until their late 20s or early 30s. This is Western Europe, specifically Europe west of the Hajnal line, where most young men had to postpone marriage until they could inherit a plot of farmland. This pattern is attested in the earliest marriage records, notably those of an English community between 1252 and 1478:

The average age at first marriage in the Lincolnshire Fenland before the Black Death would be 24 years for the woman and 32 years for the man. The wife would die one year before her husband and the marriage would last for about 13 years. The couple could have six children, if their fertility was higher than average, of whom, judging by pedigrees, perhaps three would survive to become adults. After the Black Death the mean age would be 27 for the woman and 32 for the man. The husband would die three years before his wife and the marriage would last about 12 years. Again the couple could have six children, of whom perhaps three would survive to become adult. (Hallam, 1985, p. 66)

This pattern may go even farther back in time. Seccombe (1992, p. 94) cites a 9th-century survey of the Church of St Victor of Marseille, where both men and women appear to have married in their mid to late twenties.

So, for the last millennium, most Western European men began to reproduce at around the same age as their counterparts in sub-Saharan Africa. Over most of that period, they would have ended their reproductive careers at an older age, given the difference in life expectancy.

Yet it is in Western Europe where we find some of the highest mean IQs in the world. One might counter that IQ has been pushed upward by strong selection pressures, notably those associated with gene-culture co-evolution. Undoubtedly. But, then, why cannot gene-culture co-evolution stand alone as a sufficient explanation … not only for Western Europeans but also for sub-Saharan Africans?


Cochran, C. (2012a). Gerontocratic polygyny, August 26, West Hunter

Cochran, C. (2012b). Obvious! Yessss! It was obvious! September 5, West Hunter

Cochran, C. (2012c).  Gambia, September 17, West Hunter

Hallam, H.E. (1985). Age at first marriage and age at death in the Lincolnshire Fenland, 1252-1478, Population Studies, 39, 55-69.

JayMan (2012). More on farming and inheritance systems – Part 1: IQ, June 14, JayMan’s Blog

Pebley, A. R., and Mbugua, W. (1989). Polygyny and Fertility in Sub-Saharan Africa. In R. J. Lesthaeghe (Ed.), Reproduction and Social Organization in Sub-Saharan Africa, Berkeley: University of California Press, pp. 338-364.

Seccombe, W. (1992). A Millennium of Family Change. Feudalism to Capitalism in Northwestern Europe, London: Verso.

van den Berghe, P.L. (1979). Human Family Systems. An Evolutionary View. New York: Elsevier.

Wilson, M. (1950). Nyakyusa kinship, in Radcliffe-Brown, A.R., and Forde, D. (Eds). African Systems of Kinship and Marriage. (pp. 111-139), London: Oxford University Press.


Sean said...

Life expectancy of an adult is very different to that at birth which is a much lower figure, and more often quoted. Life Expectancy at Older Age. Male polygynists would not have the problem of their wives being too old while they were still fertile. So I still think GC has made a very good case that there will be more germline mutations in polygynous societies. CH Waddington said mutations get canalized out of expression in normal circumstances. Looking at Black Aficans they don't look like they are loaded with mutations, not ones that are being expressed anyway.

RE the argument that "Because intelligence requires the correct functioning of large numbers of interacting genes, it is more sensitive to random mutations". People with intellectual handicaps invariably move oddly. Even schizophrenics have a staggering walk.

Chris Crawford said...

Disentangling the many variables at work here seems impossible given the paucity of data. For example, older fathers can provide better nutrition for their children. Different cultures have different selective forces at work for males, implying different intelligence levels for fathers in different cultures. Polygyny could well be correlated or anticorrelated with lots of other variables that could influence intelligence.

Extracting a signal from all this noise seems impossible just yet.

Anonymous said...

Presumably there would be differences between older fathers in a monogamous environment versus older fathers in a polygynous one.

Anonymous said...

My wiki-only reading of the hajnal system is that the average age for men was 26 whereas the average age in the polygynous cultures seems to be higher so far?

I'd have thought a better test of the idea would be (from the wiki):

"to the east of the line and in the Mediterranean and select pockets of Northwestern Europe, early marriage was the norm and high fertility was countered by high mortality"

If paternal age was the *only* factor then the non-hajnal parts of Europe should historically have had the higher IQs (and if the hajnal system started at some point then the pre-hajnal population in the same regions should have been smarter before it began).

I don't know if the genetic load idea is correct or partially correct but if it is then it seems that might provide a possible explanation. If different marriage systems provide different levels of genetic load shedding then that will have an effect on average IQ also.

If so the cumulative effect would be the combination of the load shedding effect and the paternal age effect.

If polygynous marriage has a low genetic load shedding ability then the two negative effects would be combined.

Even if polygynous marriage is better at genetic load shedding in particular environments it might not or only partially outweigh the extra load from the higher paternal age.

Or in other words

Total load = additional load per generation minus load shed per generation.

(The thing i'm particularly curious about in this regard is if it effects the mutational clocks people use to estimate the dates of prehistoric events.)

Anonymous said...

Thought you might like this:

Common Parasite Linked to Personality Changes

Ben10 said...

The average IQ for an entire country doesn't make sense anymore.
France has well different populations with strong ethnic components: germanic in the north east, celtic in the west, you could still say northman in Normandy, latin in the south, basques, alpin...and now, some areas are 90% composed on african or arabic background, not a single white head can be seen. It's probably not much different in England now, so, are we averaging apple and oranges?
My math teacher told me you could never do that.
We could we argue, for example that based on the 'average' IQ of the australian continent, the IQ of english migrants has fell 20% by coming to Australia.

JayMan said...

Thanks! I've updated the map used on this post with Lynn's & Vanhanen's 2012 IQ data. See here:
More on Farming and Inheritance Systems – Part I: IQ | JayMan's Blog

Peter Fros_ said...


Part of the increase in life expectancy is due to lower infantile mortality. But there has been a real increase, even if one ignores younger age groups. This is particulary true in sub-Saharan Africa, where colonial pacification put an end to the situation of endemic war.


There are some ways of controlling these variables. One could, for instance, look at IQ within families. Are younger siblings less intelligent on average than older ones? That would be expectable if paternal age influences IQ.


No, in high-polygyny societies, men usually marry around 26. Of course, they often procreate before that age. A Cameroonian classmate told me that many children in polygynous households are fathered by young lovers.


Yes, I'm familiar with Toxoplasma gondii and how it seems to manipulate the behavior of its hosts. I suspect there are many other pathogens like T. gondii. Vaginal yeast, for one.


Yes, there is probably a lot of significant variation in mean IQ within countries. In France, Brittany seems to stand out as a higher IQ region.


Great! I'll update the map on this blog.

Anonymous said...

"No, in high-polygyny societies, men usually marry around 26."

But what is the average age of fatherhood if a man takes a first wife at 26, a second at 34 and a third at 40?

Anonymous said...

No, in high-polygyny societies, men usually marry around 26. Of course, they often procreate before that age.

I meant in terms of traits.

Peter Fros_ said...

"But what is the average age of fatherhood if a man takes a first wife at 26, a second at 34 and a third at 40?"

In the past, the average age of fatherhood was the same in low-polygyny and high-polygyny societies. For one thing, most men were dead by their mid-40s. If their wives were five years younger, there would have been little or no "fertility wastage." The same was true for widowers, who typically remarried much younger women.


Yes, older men in highly polygynous societies are typically "big men" (socially charismatic, domineering, extravagant in giving away gifts and favors). Quite the opposite of old codgers in monogamous societies.

Tomasz R. said...

Mutation number is not only dependant on age, but also on exposure to chemicals or radiation via environment or lifestyle. This means people in industrialized societies have their genes mutating faster than tribes from jungle, because of all the chemicals in food, water, air, as well as higher radiation emmissions from medicine (CT-Scans, X-rays), or telecommunication / broadcasting.

There are also regions, like Scandinavia, which have very high natural background level. said...

Paternal age in lower temperature climates could affect the total genetic load?

The surviving children in a tropical setting could be different from the children surviving in temperate/cold setting?

Cold and germs kill in a different way and they should select for different traits.

Sean said...

Greg Cochran maintains that in the brain there is More to go wrong. So intellectual function is affected while movements, requiring the action of fewer genes, are not.

OK, but why did Jensen believe that blacks equaled whites at rote memorization of simple facts and skills.

How come mutations only affect conceptual learning? As Jensen said, one reason teachers doubted his figures for black IQ was that the average Black child seemed able to learn things like the names of everyone in their class or school as quick or quicker than whites.

Jensen's own experience of testing was that, in person, the black children seemed smarter in some (perhaps superficial but nonetheless real) ways than white children of the same IQ. He retested some for that reason.

I just don't see how mutations could drag down conceptual functions, but leave associative learning completely unaffected.

Sean said...

In his latest Cochran cites better outcomes for low birth weight, preterm birth and stillbirths in white mother and black father couples than black mother white father couples as suggesting magnified maternal-fetal conflict plus genetic load in blacks. A commenter mentions Flynn's German mothers of black children and their surprisingly high IQs. Nemesis did not have long to wait .

Crimson Guard said...

jayman is posting nonsense from Richard Lynn's made up figures. He tried trolling those maps he had made and spreading mis- information around on various blogs and forums related to Race and IQ to attack Italy/Italians/South Italians.

Lynn's figures are garbage and cannot be trusted:

Refuting Richard Lynn's Italian "IQ" Study

Richard Lynn Further Refuted

More Criticism of Richard Lynn

Another Richard Lynn Refutation

Rebuttal to Richard Lynn's Reply

Sean said...

In his recent posts GC says that 'foragers' have "quite a few guys [who] lived well over 50".

In forager society there would be no rich old men who were trading on their family status; hunter gatherers are egalitarian. There, a man needs to be top quality to get a fertile wife at 50; an exceptionally able hunter, or perhaps a war leader. It seems to me that if mutations make a difference to brain functioning the old fathers would have fewer mutations.

Germline mutations occur in the testes, and one would assume they would tend to be expressed there. So do black Africans have lower sperm quality? No, they have better at all ages (ie not a result of the germline mutations that increase sperm quality and are selected for in older men). Moreover a man who can get a fertile women when he's 50 years old, in a polygynous mating system (where women do the choosing don't forget) would tend to be in much better shape than his contemporaries; in other words he would have better genetic quality, and fewer mutations reflected in extra social intelligence or charisma. Health too; there is a "clear link between semen quality and the rate of death." Here. So older men in good enough shape to become fathers would tend to have fewer total mutations. To suggest otherwise is to suggest mutations don't make a difference to the qualities that women look for in men. But navigating a social environment and being a good hunter surely require relatively good brainpower.