A paper in The New England Journal of Medicine, Cell-free DNA Analysis for Noninvasive Examination of Trisomy, reports on the effectiveness of a new proprietary method to screen for Trisomy 21, which is the cause of 90% of cases where individuals exhibit Down Syndrome. This issue is well known at this point. There are many methods to screen and diagnose Down Syndrome prenatally, but they all suffer from drawbacks, from invasiveness to false positives. Currently the vogue is for methods which analyze maternal plasma. That is, all that’s needed is a blood draw from the mother. This is one such method.

Their data set was large, N > 15,000, and not particularly old, with a mean maternal age of 31. Here’s a key point: “Among the 11,994 women with low-risk pregnancies on the basis of a maternal age under 35 years, cfDNA testing identified 19 of 19 women with trisomy 21, with 6 false positive results.” This is actually a good result, as older methods have a much higher false positive rate. But many people will not be reassured when they see that the true positive and false positive ratio is so high. Base rate neglect is always going to crop up. For many couples who get positive results confirmatory evidence is essential. That is why it is crucial that the mean gestational period of detection has to be pushed back beyond the beginning of the second trimester if these tests are to cause the minimum amount of discomfort for the couples. The later the abortion, the more psychologically and physically stressful the process.

But, I put a question mark in the title of this post because there is often an assumption that widespread screening will result in a massive decline in the number of individuals with Down Syndrome, to the point of extinction. Actually, I’m not sure about that. It seems that widespread adoption increases the pool of people who are getting tested, and fewer of these in the United States terminate pregnancies through abortion that the early adopters. As mean maternal age creeps up the number of people with Down Syndrome in the United States may not decrease nearly as much as we expect.

Whenever I talk about The Nurture Assumption there are a minority of angry and peeved comments. Usually they’re not too coherent, but they don’t get me down. The reality is that the basic message of the book is very important to get out to the American public, by which I mean upper to upper middle class Americans (since these are the target of “think pieces”). The reason is that today the “nurture assumption” reigns ascendant, and makes superhuman demands on parents, especially mothers. This explains some of the reaction to a new paper, Does the Amount of Time Mothers Spend With Children or Adolescents Matter? (ungrated). For a representative example, see in The Washington Post, Making time for kids? Study says quality trumps quantity. Quality as in quality of time.

But one thing bothers me about these treatments in the press: the totally confounded nature of causality. Consider:

The one key instance Milkie and her co-authors found where the quantity of time parents spend does indeed matter is during adolescence: The more time a teen spends engaged with their mother, the fewer instances of delinquent behavior. And the more time teens spend with both their parents together in family time, such as during meals, the less likely they are to abuse drugs and alcohol and engage in other risky or illegal behavior. They also achieve higher math scores.

The implication above is that “time engaged with mother” → “less delinquent behavior.” But we don’t know that that’s causal at all. Rather, it could be a correlation between a third factor, term in “prosociality,” and these two variables. More generally if you look for references to genetics in the original paper you won’t find it. It strikes me that one of the reasons that parental investment doesn’t seem to matter so much is that there are many outcomes they just aren’t effecting, because their primary contribution in heritable, with a major secondary contribution to the environmental context in which children grow up (the “non-shared environment”).

High parental investment also isn’t about the children in the proximate sense (though parents sincerely believe that it matters ultimately). Rather, it’s a form of inter-familial status competition. The “best” parents are those who invest the most, and achieve the best outcomes in their offspring.

Unlike some books I did not read Strange Parallels: Southeast Asia in Global Context, c.800-1830 (vol. 2) in one sitting, or even over a few days. Part of the reason is length, at 977 pages in the print edition. Another aspect is the frankly verbally baroque writing style of the author. Very rarely do I encounter terms such as “architectonic” or “mycelium” in historical narratives. This is not the first time I’ve tried to read Strange Parallels (vol. 2). In 2010 I read Strange Parallels (vol. 1) while flying back and forth across the Atlantic for a wedding. Naturally a few weeks later I decided to tackle volume 2…but was daunted by the length, and the surprising substance of the sequel. I recall stopping somewhere at the point where the author was expounding at length on the nature of early state formation in Kievan Rus.

Not that I have a problem with an exploration of a topic like the genesis of the precursor of the modern Russian state, but the survey seemed ridiculously expansive. In Strange Parallels the author leverages his area knowledge of early modern mainland Southeast Asian state formation, and applies those insights more generally to arc of French, Russian, Japanese, Indian, and Chinese socio-political development. These specific cases are used to explore a descriptive observation about Eurasian polities over the past ~1000 years: many of them developed in a synchronistic manner in regards to economic growth and political robustness, or lack thereof.

The author explores a variety of hypothesis to explain this pattern, but more important is the massive “core dump” of specific detail which surveys the histories of these disparate nations. Though Strange Parallels is conceptually fertile, with ideas such as “charter states” and “protected vs. exposed zones” (in relation to Inner Asian nomads), it is in the empirical richness that the work justifies a close reading. It’s one of those books where you inspect the footnotes! Though the prose and the length of the work are somewhat of a slog, going through the whole book is worth it to gain a broader understanding of Eurasian political and social history over the past 1000 years.

Addendum: There are many similarities with William McNeill’s The Human Web, though Strange Parallels is a much more expansive book that focuses on a narrower time period and a more finite set of societies.