« Coupling vs. Coupling | Gene Expression Front Page | Another one bites the dust for print »
September 26, 2003

Cousin you look mighty perty

Steve Sailer links to an article that encapsulates Ahmad Chalabai's response to his piece in The American Conservative on cousin marriage-""The Jews have had cousin marriages galore, and it hasnít hurt them." When I asked a friend of mine who is half-Israel Arab (his father's family are Arabs from Nazareth who have Israeli citizenship) about cousin marriage-his response was, "Cousins are the only people of the opposite sex that you get to meet, so it makes sense...." His own sister, who moved to Israel to live with her retired mother (who is of Scottish ancestry), ended up marrying a first cousin. As for Jews marrying relatives, perhaps there is something cultural that explains it, but obviously the small number of potential mates of appropriate background had something to do with the inbreeding. The prevention of fraternization between men and women of the opposite sex who aren't related has the same practical effect-shrinking the possible options for marriage down to a small tightly knit circle.

I guess realist American conservatives are in the position of asking Arabs to be just a bit more slutty, though they wouldn't want them to be as slutty as this. The Golden Mean was forgotten long ago....

Read ParaPundit's article which spawned much of the blogospheric debate on this topic.

Posted by razib at 01:24 AM

Cousin marriage has the effect of keeping inheritance within a lineage. It was practiced by the elite in Spartan Greece (Paul Cartledge, the Spartans). It can be supplemented by aunt marriage (really), sister-marriage (ancient Egypt, Polynesia), step-mother-marriage (after the father's death), levirate marriage (sister-in-law after the brother's death). All these have the effect of preventing inheritable property or political rights to pass into the hands of a different family. It has been speculated that infant sacrifice in Carthage functionally had the effect of keeping the pool of heirs small, to keep it from being divided. In the Ottoman empire it became routine and actually authorized for the victorious heir to the throne to kill all his brothers and half-brothers. Likewise in many societies extra daughters and sons are forbidden to marry.

Here we have group entities which may be gene-rational in some ultimate way, but which are economically and politically rational in a very evident way which apparently conflicts with genetic rationality. The kin-group maximizes its power and wealth (culturally or memetically coded) rather than its fertility (biology). In some cases adoption is used to strengthen a family; genetically irrational, but economically and politically rational. An ultimate case is the ruling Mamluks of Egypt, who were of random foreign slavemarket origin and whose children were forbidden to become Mamluks; an early non-kin collective entity like a corporation. Getting away from kin organization is one of the prerequisites of progress, modernization, secularity, etc.

Posted by: Zizka at September 26, 2003 08:16 AM

I think Steve Sailer's thesis is interesting, but as usual he goes overboard with the biological determinism. Sure, nepotism is rooted in human nature, but to suggest that inbreeding over several generations may lead to stronger nepotistic drives than what you'd get from just assuming that the inbreeding began one generation ago - that's really pushing it.

A more likely scenario is that inbreeding helps foster a self-containment that can lead to an extended-family becoming a separate culture in its own right. We know that cultural traits are transmitted in much the same manner that genes are, and as the cultures of different extended families diverge, it becomes more difficult to relate to or trust people from outside the culture.

There is also the "background-check" angle to think about. If a guy is part of your family clan, its a lot easier to check out his past and his personality by simply calling up an uncle or three and asking about him. Even if you weren't inclined to nepotism by any biological urges, in a highly distrustful environment like the middle east, where secret police and religious fanatics rule the roost, hiring a relative to fill any sensitive position would likely be the safest thing to do, simply because he'd be a known quantity. I think this is one reason why the CIA has had such a hard time in penetrating the Islamic terrorist networks.

From a biological viewpoint I think there's no arguing that first cousin marriages, and even worse, uncle-neice marriages, are simply a terrible idea. Here's where Sailer's multigenerational effect really plays out, as the Saudi infant mortality figures seem to illustrate. To be honest, I'm no cultural relativist on this score, and I find the idea of marrying a close relative utterly repulsive. I'm not a libertarian either, and if I had my way first-cousin and uncle-neice marriages would be banned everywhere.

Posted by: Hanno Buddenbrook at September 26, 2003 08:36 AM

By the way, I'm wondering if someone could enlighten me as to the truth of the following, and as to why it holds, if it actually is true.

The North Indian Hindus prohibit marriage between kin that extends back to seven generations on the male side and five on the female side. But the South Indian Hindu tradition sanctions consanguinity.

Supposedly South Indians even approve of uncle-niece marriages, which I find extremely shocking. Why the polarity of attitudes within the same religious tradition?

Posted by: Hanno Buddenbrook at September 26, 2003 09:00 AM

Herr Buddenbrook talks himself around to understanding many of the reasons that led me to say that multiple generations of inbreeding create more intense bonds of loyalty to extended family members than does a single generation. It's a little hard to visualize precisely without drawing family trees, but the concept is simple --the more generations it goes on, the more you have _multiple_ relationship ties to your relatives. In outbreeding families, the amount of altruism you have to dispense gets dissipated quickly because you belong to so many different famiiles: Your father's family and your mother's family, but also your father's father's family and your father's mother's family, etc. In a lineage that's been inbreeding for generations, however, there's a lot less dissipation of family loyalty. Is this biological determinism or social environmentalism? It doesn't really matter. It works the same in both cases.

Posted by: Steve Sailer at September 26, 2003 05:19 PM

Hanno, Nepotistic drives certainly would be stronger over many generations. A powerful grandfather is far more likely to want to help a particular grandchild get a job if the grandchild is married to another of his grandchildren. After all, that way he's helping two of his grandchildren. Also, if his sons married his brother's daughers then his brother is going to also have a big incentive to ask him to help his own grandchildren. Plus, if he is related to his wife and if his wife's sister is married to his brother then that just makes for more family members all pushing for him to help his grandchildren.

Yes, cousin marriage certainly would increase the motive to be nepotistic across multiple generations. Think about what this means for government. If grandfather is a government minister he's going to be more motivated try to use his influence to get his children and grandchildren positions in government and industry. Also, if any family members want a government contract or permit or exception from some rule he's going to be more motivated to use his influence than would be the case in Western societies.

Posted by: Randall Parker at September 26, 2003 10:58 PM

Oriental families tend to be more cohesive, regardless of endogamy. For instance Indians do not practise cousin marriages but their culture is anchored in the extended family.

Nepotism in government is tangential to development. Singapore, arguably Asia's most successful economy, Harry Lee's family occupies positions of power. The global financial industry is defined by nepotism whereby merit is disregarded for connections and networking.

In least developed countries money goes a very long way to determine the allocation of government contracts moreso than family connections itself.

The high consanguinity of Pakistan manifests itself most prominently in the phenomenally high rates of ovarian and breast cancer. The medical effects are being countered by the fledgling practise of undergoing medical tests prior to marriage.

Cousin marriage is not responsible for the low level of development rather Hernando De Soto's explanation is more apt. The societies of Islam are continually hindered by the lack of liquidity in the economy and the rigidity of the property market (in Pakistan it takes years to accumulate the cash to buy a house because we just don't have a developed mortgage market).

Posted by: Zachary Latif at September 27, 2003 07:14 AM

China forbids cousin marriage in the male line. Cousin marriage in the mother's line (mother's sister's daughter) is mildly favored, but not terribly common. Chinese nepotism is not thereby impaired. One effect is that daughters, who will marry out, are not valued, and wives are not trusted or valued until they have produced several healthy children.

Posted by: Zizka at September 27, 2003 08:20 AM

Zachary, is it hard to establish clear title to a piece of land in Pakistan? If a piece of land is to be sold how long does it take to get the title clearly switched to a new owner? How much of the farm land is not listed in government offices as being owned by anyone even though it is being farmed?

Does Pakistan have the equivalent of America's small claims court for lower dollar value disputes? If so, is it easy to use?

De Soto's explanation is relevant for countries where a substantial portion of the population can't officially own things or where a substantial portion of the people can't bring suit in court to enforce contracts. But it is not a question of either-or with regard to De Soto's explanation versus nepotism. In fact, those are just two of several reasons why some places do not economically develop. The importance of cousin marriage is that it constitutes a much larger obstacle to reform. The loyalties that it creates can't just be swept away. It would be a lot easier to implement De Soto's reforms in a society that doesn't have a high level of cousin marriage than in one that does. It would be a lot easier to develop mediating civic institutions in a society that has little cousin marriage.

Also, cousin marriage works in more ways than just in how much corruption it encourages. Certainly cousin marriage turns loyalties inward and therefore decreases the energy that people put into secular civic organizations that create mediating political entities between government and the people.

Regarding the Chinese: the turn family business bonds was the consequence of a lack of a properly functioning court system for adjudicating commercial disputes. People tended to keep business in the family as a way to be able to bring pressure on people to honor deals they made. But cousin marriage is rare among the Chinese. Go to consang.net and see just how more cousin marriage there is in Muslim countries.

Posted by: Randall Parker at September 27, 2003 12:16 PM

I have a link to the latest NY Times John Tierney article on consanguineous marriage in Iraq and the political effects of this practice. The Iraqis obviously understand what so many American commentators seek to deny. In that post I include links to my previous relevant posts.

Posted by: Randall Parker at September 27, 2003 03:35 PM

Partition helped solved the land titles problem because of the exchange of properties between East and West Punjab (approximately an equal exchange of Hindus & Sikhs with Muslim Punjabis).

In Pakistan ownership can be easily established but red tape, especially in the rural areas, is a hinderance. Anyway most of the Pakistani rural population are landless peasants and can't even concieve of owning property. Interior Sindh is an excellent example where the landless Haris are at the mercy of their Wadhera landlords (Punjab is slightly better because of the canal colonies, which armed the population with property rights).

Enfranchisement can only come when people have access to property and land ownership.

Hernando de Soto touches upon the more important issue, which is generating wealth by establishing property rights and generating wealth & liquidity. A government can print as much money as it wants if the wealth of the nation is rapidly growing.

Pakistan's black economy is phenomenally large and indeed the past few years have shown extremely high economic growth because it is being reintegrated into the economy. It is the systematic incorporation of property rights and the incorporation of the illegal economy, which is critical for development.

The impact of cousin marriages is rather limited and indeed everyone has extensive ties with the rest of society (a normal Pakistani family can have over 30 first cousins and innumerable second cousins). Our growth has been so explosive that though we now number 140mn our ancestry has been confined to a historically limited stock.

Naturally connections and "rishtidar", family ties, exist. Indeed the system of patronage is particularly apt for the Muslim model, perhaps an echo of the "mawali" being a client of the Arabs. It is arguable whether they are a hindrance or an aid to our societal development. Indeed there is a tolerance for intermarriage with other ethnicities and religions if cousins are few and far between. Some ethnicities, like the Muhajirs who don't practise cousin marriage, are rapidly absorbing into mainstream society because of extremely high intermarriage rates.

To be honest as for corruption it is favourable if it siphons funds from white elephant projects. The money is inevitably recycled back into the economy and has a multiplier effect.

South Asian and Islamic governments have had a trend toward extreme centralisation, which has been extremely harmful. This has been to compensate for the uneasiness of the state at the inception. Accountability just doesn't exist as a concept because of the removal of executive and legislative power from that of the people.

Development has always followed the European "top down" model rather than the "bottom up" path adhered to by the United States.

Posted by: Zachary Latif at September 27, 2003 05:05 PM

Zachary, Are family connections important for getting jobs in government or in private industry? How do most people get jobs?

Posted by: Randall Parker at September 27, 2003 11:57 PM

Hanno, Nepotistic drives certainly would be stronger over many generations. A powerful grandfather is far more likely to want to help a particular grandchild get a job if the grandchild is married to another of his grandchildren. After all, that way he's helping two of his grandchildren. Also, if his sons married his brother's daughers then his brother is going to also have a big incentive to ask him to help his own grandchildren. Plus, if he is related to his wife and if his wife's sister is married to his brother then that just makes for more family members all pushing for him to help his grandchildren.
I can buy that a grandfather might be more willing to help out a grandchild to whom he was related many times over, but that doesn't mean that this nepotistic effect would necessarily extend beyond living links of this sort.

I don't buy that just because two people have shared ancestry at some time in the recent past, they'd feel any sort of biologically driven nepotistic tendencies towards each other. For one thing, how would they possibly know they were related? It isn't as if they could sniff each other's armpits and say "We share 40% of our genes in common!"

Reinforcing this skepticism is the fact that cousin marriages between people who have grown up together should even be palatable in the first place; weren't sociobiologists theorizing that we had an incest avoidance taboo not so long ago? If we aren't actually biologically driven to avoid sleeping with close relatives, why should we be biologically driven to help others more, just because we happen to share more alleles in common with them than is the norm amongst relatives? I'd bet that if you met a man on the street, who, to your ignorance, was a long-lost brother of yours, you wouldn't be pulled to respond to his pleas if he happened to be a drunken smelly wino begging you for a dollar.

The nepotistic influence that you and Sailer argue for does exist, but you push things too far in assuming that it must continously increase in strength past 2 or 3 generations.

I have a link to the latest NY Times John Tierney article on consanguineous marriage in Iraq and the political effects of this practice. The Iraqis obviously understand what so many American commentators seek to deny. In that post I include links to my previous relevant posts.
Seeing as Tierney explicitly acknowledges Sailer's influence on him in the article, he isn't exactly an unbiased observer, and anyway, this is merely a New York Times article, rather than a carefully designed study anybody is duty-bound to respect. At best, it indicates that Sailer has the contacts and the know-how to push his stuff in front of journalists looking for a story, which is hardly news, is it?

Posted by: Hanno Buddenbrook at September 28, 2003 06:01 AM

The health effects of first-cousin marriage are often exaggerated. There is a real adverse effect, but it is minor compared to many other hazards.

I can't help recalling that Charles Darwin married his first cousin, Emma Wedgwood. They had 10 children, 7 of whom survived to adulthood. 3 of them became eminent scientists.

Even uncle-niece marriage isn't that bad - see A. Bittles et al. in 'Minority populations: genetics, demography and health', ed. A. Bittles and D. Roberts, 1992.

I'm not advocating incest, just pointing out that it ain't all bad.

Posted by: David B at September 28, 2003 07:17 AM

Before anyone else chips in, I forgot to say that 7 out of 10 was not bad in Victorian times. One of the 10 (born when Emma was in her late 40s) may have been mentally defective, and died before he was 1, but Randal Keynes thinks he may have been Downs, which would have nothing to do with cousin marriage. One of the others (Bessy) is sometimes said to have been slightly retarded, but the evidence suggests she was just quiet and shy - she learned to read French as well as English, which is hardly compatible with any severe deficiency.

Posted by: David B at September 28, 2003 07:24 AM

also remember, the deleterious effects of inbreeding can be counter-acted by one generation of out-breeding....

Posted by: razib at September 28, 2003 11:12 AM

I'm not advocating incest, just pointing out that it ain't all bad.
Well, Arabs and South Asians aren't exactly endangered species, so you're likely correct. Still, the thought of some guy looking lustfully at his niece just creeps me out!

Posted by: Hanno Buddenbrook at September 28, 2003 11:44 AM

Clan systems have an authority which is independent of individual feelings. The big Chinese clans were enormous corporate economic and political entities controlling large businesses and whole districts of territory. The senior male of the senior clan was boss (with adjustments to allow for incompetence and talent). The seniors (clans and individuals) exploited the juniors. Confucius worked this up into an idealized system, though he also had a large number of (ineffectual) warnings against nepotism and graft.

When I was in Taiwan in 1983 I was told that family businesses would never give any responsibility to anyone from a different family (except that a powerful family might favor someone from a dependent family). I also heard a story of an employee caught walking out the door with a list of all his company's bids; he was going to take the bids to his cousin, who would underbid everything by 1/4 % or so.

The power of kinship (~ genes) is enormous in many societies, but progress and modernization in Europe depended heavily on reducing that power and producing individualist non-kin institutions. This was as important as religious secularization. Taiwan is only part way there, or was in 1983.

Posted by: Zizka at September 28, 2003 09:26 PM


Pakistan has the civil service examinations for graduates for the public sector.

Private industry is dominated by a highly successful and relatively large elite with family connections, networks and contacts providing avenues for employment (anyone who's anyone knows everyone).

This is similar to the financial industry in the City of London where contacts provide upto 50% of the jobs obtained.


Posted by: Zachary Latif at September 29, 2003 04:27 AM