29 Jan Kong et al on the Nature of Nurture
You might be getting old, if….
Half your posts are on the theme of, “Nobody reads the classic literature anymore.”
Kong et al.’s The nature of nurture: Effects of parental genotypes is a clever, original analysis, but I assumed that the title was an hommage to Plomin and Bergeman’s no-questions-asked classic paper in BBS (1991), The nature of nurture: Genetic influence on “environmental” measures. But neither Kong et al, or Koellinger and Harden, in their comment on Kong et al, even mention it. BBS papers can be hard to find in full text (Why?) and in fact I can’t find a full text version, and I don’t want to post one on my own. The abstract is here.
(And it didn’t end there. In 1994 Theodore Wachs wrote a book, very influential at the time, called The Nature of Nurture. It was also the title of Plomin’s 2009 contribution to the Festschrift for Sandra Scarr.)
Plomin and Bergeman, published at the height of the twin study era, was concerned with the realization that the old distinction between nature and nurture was starting to break down. All of the usual BG suspects– intelligence, depression, you name it– had turned out be heritable, but once people (mostly Robert Plomin) started looking, it turned out that phenotypes that were supposed to be on the environmental side, like socioeconomic status and parenting practices, were pretty much as heritable as everything else. The paper was extraordinarily influential. Paired with Plomin’s other classic BBS paper from a few years before, Why are Children in the Same Family So Different from Each Other? it marked the beginning of the end of the classic BG era of just quantifying the heritability of things into its integration with social science, under the heading of “gene environment interplay.” (See Plomin’s 1994 book, Genetics and Experience: The Interplay of Nature and Nurture; and my (2006) Inter: Action and Play.
The paper was important for me personally as well. Irv Gottesman was asked to comment on it, he invited me to write it, and we came up with, Is h2=0 a null hypothesis anymore? in which we said,
Our concern is about where all this will lead. Behavior is influenced by genotype and environment. The environment provided by a parent is influenced by the parent’s (not to mention the child’s) genotype, and the parent’s rearing environment, which had its own tangle of reciprocal genetic and environmental influences. Every thing is intercorrelated; everything interacts. Where does this leave the columns of model-fitting heritabilities, meticulously computed to two decimal places and starred for statistical significance on the basis of path models that cannot hope to keep pace with the reciprocal causal structures described in the target article?
Having established beyond any reasonable doubt that some’ of the variance of every characteristic is influenced by genes, it may be time to enshrine h2=0 as the first law of behavior genetics, and concede that H2=0 is longer an interesting null hypothesis.
I think we are seeing the same thing happening now, with DNA and GWAS in place of twins. If history repeats itself, the results will be a mixed bag. You can see what is coming down the road: transmitted and untransmitted maternal and paternal GPS, combining with GPS for other traits in the parents and children, correlated with and interacting with observed phenotypes in the parents and kids. Everything is correlated with everything, everything interacts with everything: interplay. What this means is that Kong et al’s talk of variants with “direct” vs. “indirect” effects is a little misleading. Once we are in the domain of the tiny effects of unspecified variants on distant social outcomes like educational attainment there are no direct effects. (Just by the way, there are no “causal variants” either.) There is just a complex network of genetic and environmental inputs into developmental systems. In human beings, where controlled experimentation is impossible, it is very difficult to make substantive scientific progress.
Gene environment interplay has produced a mountain of social science, all sorts of more or less interesting models of what is correlated with what and how. I have produced a bunch of that sort of stuff myself. Seen from the inside, as an exercise in genetically informed social scientific methodology, it is fascinating. I have spent half my career doing that kind of thing. But at the end of the day, genomics won’t shine the bright light of biological genetics on the social sciences; instead it will ensnare genetics in the methodological briar patch of social science.