Wednesday, May 19, 2021

We Marry For Unseen Reasons

Apparently, we marry the 5th-10th cousins we did not even know at a disproportionate rate, or at least the Brits do. How Greg Clark and his researchers corrected for geography and distance on this, I don't know.  When the third book of his trilogy comes out in 2022, he promises that the data will be publicly available, and if other researchers have other interpretations of the data, they are welcome to publish on it as well.

Gregory Clark is a social economist from UCDavis and  the author of A Farewell to Alms and The Son Also Rises. As he was recently cancelled at Boston University for his talk "For Whom the Bell Curve Tolls" the working title for upcoming third book, I have to give him great credit for provocative titles.  That lecture title was not a reference to Charles Murray and Richard J Herrnstein's book, BTW, but to a critical review of Clark. Razib suggested that because of the triggering of liberals, people should just use the phrase normal distribution in place of bell curve and save themselves the headache. If you think about it "For Whom the Normal Distribution Tolls" is actually pretty funny. However, it is a step away from the Hemingway puns of the other two titles.

Neither the books nor the talk touched on race in any way and very little on data outside of England, but he has been accused of racism, apologising for colonialism, and being a closet eugenicist.  The usual. The accusations that his first book was a bit lazy and sloppy I have no way of commenting about.  It could be, or it may just be scrambling to find things to throw at him. For the upcoming work he is making the data available, as I noted.

My current project is a book manuscript on the nature and implications of the determination of social outcomes in England 1680-2021, based on a complete lineage of a set of 402,000 English people across this interval. The provisional title is For Whom the Bell Curve Tolls: Genetics and Social Life, England 1680-2021. This will form the third of a Hemingway punned trilogy of books about the interconnection of modern growth, selection pressures in the long Malthusian era, and the implications for social functioning. 
The link to the working paper "For Whom the Bell Curve Tolls." I found it interesting, but it is mostly the math of comparing one model to another and seeing which fits the data better. You might find it tedious. Spoiler: the socio-genetic model accords with reality better than the economic and cultural models.

When I discuss genetics I have been leaving out assortative mating, an important piece, which is much the subject of Clark's new work. It counts both for and against the purely genetic hypothesis, as it on the one hand is genes seeking genes and reinforcing each other, yet on the other is clearly a cultural/environmental factor of its own. 

We could have system of marrying in which we chose the most attractive, or the richest, or the youngest person available, and these are all found in culture and commented on.  But marrying according to not only a current social status, but what we internally estimate will be a person's social status seems to be a far better predictor.  Couple that with the strong predictive value of genetics for occupation, literacy, and eventual wealth and genes seem to be subtly calling out to other genes.  A word on wealth: Clark notes that family size has an effect on transferred wealth initially, but this washes out when wealth at the end of life is measured. The purported effects of having better contacts and networks because of family do not show up statistically at the end. 

Clark calls social status complicated, related to "cognitive abilities, drive, self confidence, and many other factors."  Maybe so, but my bet would be that IQ explains the lion's share. We do not marry seeking maximal social status of spouse, but something most like ourselves. We may seek to marry up more than down, but something in us wants companionship as much as gain.  There are exceptions, often noted in literature and anecdote.  (Hypergamy is a thing in my family. It's real.) Yet I think we note them because we notice them and think there is something just a bit wrong about it. We want to marry each other at some deep level.* This shows up in the British data.

We share genetic material with parents and siblings approximately equally.  We also usually share an environment with siblings.  If environment mattered all that much, social outcomes would be closer to our siblings than to our parents.  The are actually less close.  We have similar percentage of genetic sharing with first cousins and grandparents, and the latter are two full generations and often quite a distance away. We should have more similar outcomes with our first cousins, but we are on average closer to our grandparents.

Our feeling is that we are closer to our cousins.  I have a male cousin six months older, raised by my mother's sister (they were only 15months apart) less than an hour away in New England. We correspond rarely, but when we get together conversation is easy and we feel drawn from the same pile. It is similar with his sister, who is just a bit younger than me. Were any of us to talk to Gramps it would be strained and difficult. My cousin and I can point to things about ourselves that are alike.  Yet those things are much the same as those we share with our grandfather: a quiet prosperousness based on thrift and preservation rather than brilliance (though each of us could display some brilliance); loyalty and reliability; avoidance of personal conflict. So when the data shows that my outcomes are somewhat more likely to be like my grandfather's than my cousin's it is not that shocking - upon further review.

Interesting side notes: Reproductive fitness - have more descendants - was much greater among those who went to Oxford from 1680-1850. It was better than 2:1.  Then that reversed and those who went on for advance education in England have had fewer children than the others. In America much is made of the fact that women with less education have more children.  Yet there is an interesting twist to this which I did no even hear a whisper of until Clark alerted me to it.  Among those women who have their first child after age 22, those with more education have more children.  So the overall trend is driven by women having children young.

Clark thinks assortative mating in America is somewhat different and thus reasons that other cultures will show even more variance. But that doesn't move away from genetics, but towards it. Social status is even more deeply tied to cognitive skills in America, making those interactions even more prominent.  And if there are gene-gene cognitive interactions, as the epigeneticists assure us will be so, that will heighten rather than lessen the effect. (I may have to do an updated epigenetics post.  I have been considerably anti-, but interesting data is coming out that there are threshold points of turning genes off or on during development or trauma that are usually invisible but in extreme situations can have large effects.)

Clark does this research to investigate which interventions, especially government interventions, actually have a positive effect on outcomes. He will measure at the break point of when increased mandatory education was required - the three years earlier versus the three years after - to see if there were any measurable differences in occupational or economic outcomes decades later.  There weren't. So why do we spend the money on that if it doesn't actually create a benefit?

*In one of the Madeline L'Engle books the blinking of fireflies is discussed in terms of human sorting. Like kinds give off the same number of blinks. There was a moment in college when I recognised that the woman who became my wife gave off the same number of blinks that I did.  Other girlfriends had been similar, but not the same.  (I suppose if I thought about it I could identify an earlier girlfriend who also gave off the same number of blinks, but that seems an unhealthy speculation.) My second son's courtship of our newest daughter-in-law has this mark, that they almost instantly recognised that they were of the same number of blinks and spent the next 18 months making sure there were no deal-breakers to call the thing off. Well, children's librarian. That's a pretty good first clue.



David Foster said...

"Among those women who have their first child after age 22, those with more education have more children"...well, *that* is more point supporting the idea that you can drown in a lake whose average depth is 8 inches.

David Foster said...

"But marrying according to not only a current social status, but what we internally estimate will be a person's social status seems to be a far better predictor"...there are certainly a significant number of women who marry according to actual or perceived future status of a man, but are there many men who do the same? In American, I mean?...Marrying for female status/wealth does seem to have been a thing in in various aristocracies, such as the British, but I'm not perceiving it as much of a factor here. Although there are indeed a lot of men who want to avoid marrying someone with a negative dowry, ie, a student loan balance.

Assistant Village Idiot said...

I went to college in search of a wife. That was an era when there was still some sneering at women who went to college to get her MRS, but I always thought that made complete sense.

I think also, we seek to marry "our own," or something better, and thus status ratings are not absolute. People value similar levels of ambition, physical activity, seriousness, or piety, and are willing to sacrifice in other areas to get those. As all of these are somewhat heritable and may prove out to be even more so, we may find increasing examples of genes calling out to genes.

Sam L. said...

I did not go to college to find a wife. That's because I had a sister a year older than me, and she was bossy. I met my first wife in the service. I'd just come back from 2- weeks leave, was told there were four new female Lieutenants, and one I liked. We dated. Her dad came half-way across the continent to meet me, and approved of me.

Assistant Village Idiot said...

She and he may have been the last two, Sam.

Erick Arnell said...

My wife is my ninth cousin, once removed. Our common ancestors were born in the mid 1600s. That doesn't seem like a lot of consaguinity to me.

Compared to wild animals, even fifth cousins is probably less close than would be common just based on geography. They mate with the population available, and that usually means descendants of a relatively small founding group.

Assistant Village Idiot said...

@ Eric - I don't know how they separated out the factors you mentioned. The founding population of New England was relatively small and the fact that I can trace back to Stephen Hopkins by at least five routes means there are cousins of some sort marrying, even though I can't identify them specifically. If I cared I could track it down. Somehow they thought it was still an elevated percentage.

I actually have one set of first cousins marrying in Nova Scotia around 1800, shortly after the English colonists were given the land that the French Acadians had been kicked off. I suppose choices were few.

Donna B. said...

When my youngest daughter married, her mother-in-law and I added each other to our Ancestry files. Her mother-in-law texted me a 'thank you' for adding a photo of an historical marker from the 1600s about one of their ancestors. It was one of ours as well. This particular great-great+ grandmother of our children was married numerous times, named in a lawsuit for not marrying someone she was supposedly promised to, and quite notorious in her time.

Donna B. said...
This comment has been removed by the author.