Posted at 15:36h in Uncategorized by Eric Turkheimer 0 Comments
One of the peculiar things about Robert Plomin’s recent, Blueprint, is that it seems to point in both directions. It is impossible to read the book without coming away with the impression that genes are the most important, in some ways the only, determinant of human individual differences. In the Prologue, he says, “This DNA fortune teller is the culmination of a century of genetic research investigating what makes us who we are. “. The last chapter is titled, Our Future is DNA.
Yet many other statements in the book could lead one to wonder why Plomin would say this. In the exercise about guessing the heritabilities of different behaviors at the beginning of the book, the examples don’t turn out to have heritabilities of 90%, they are all from 40% to 60%. Especially in the first half of the book, Plomin often says that environment is important too, or that genes don’t determine outcomes, they only determine probabilities. So how is it that “genetics”, a probabilistic cause that accounts for about half the variance, comes to be the great fortune-teller that shows us who we really are?
There are several reasons. The most obvious is book-selling hype. A book that led with the real complexities of nature-nurture would be much harder to sell to the general public. I don’t take this seriously as an excuse. A related reason is polemical. Plomin puts all the big hereditarian claims up front, and then qualifies them or even takes them back in subsequent chapters. This allows defenders of the book to accuse critics of not giving Plomin a fair reading– What do you mean the book is hereditarian? He says the environment is important too! Sort of. Like I say, there isn’t any serious doubt about what point of view Plomin is pushing, and even more important, Plomin never makes any serious effort to reconcile the importance of the environment with DNA as the infallible fortune-teller. The two points of view just sit there in separate chapters, unresolved.
An even more important reason for the ambiguity about the supremacy of DNA over the environment is that Plomin knows better. I say this as a compliment. Robert Plomin is the no-questions-asked most important behavior geneticist of his generation. The first half of the book is a review of his many, many accomplishments in the field, from the Colorado Adoption Project to TEDS, from the first serious description of gene-environment correlation to the nature of nurture to generalist genes and on and on. This work constitutes the great majority of Plomin’s career, and while I have disagreed with parts of it, it has unquestionably made an enormous contribution, and has never been particularly hereditarian in outlook. Robert Plomin has always emphasized the symmetrical interplay of nature and nurture; it is only now that he has decided to pick sides. How did this happen? How do you get from 50% heritability to DNA being the essence of what people are?
The answer lies in a theoretical sleight of hand in the way Plomin considers genes and environment. Originally, heritability was a variance component estimated from twin, adoption and family studies. Scientists could estimate heritability before they knew what DNA was, much less how to measure it in individuals. Heritability estimated as 2*(rMZ – rDZ) is a coarse sum over a great many– only today do we appreciate how many– unknown alleles. The beauty of quantitative genetics is that it allows us to capitalize on known variability in genetic relatedness to estimate heritability without reference to actual DNA.
Contrast this situation to the environment. There is no way to estimate how similar my environment was to yours or to one of my siblings’ environments. As a result, there is no equivalent way to estimate the environmental component of variance– it is perennially an error component after genetic effects have been accounted for. As I said in the Three Laws paper in 2000:
Thanks to the fact that identical twins are on average exactly twice as similar genetically as nonidentical twins, one can use straightforward statistical procedures to estimate the proportion of variability in complex outcomes that is associated with causally distant genes, all the while maintaining a state of near-perfect ignorance about the actual causal processes that connect genes to behavior. This methodological shortcut is not available to rivals of behavior genetics who seek to measure the effects of families on behavior. How similar was my rearing environment to that of my siblings? And how similar was it to the environment of my adopted sibling, if I have one, or to the environment of my biological sibling who was raised by someone else? The apparent victory of nature over nurture suggested by the first two laws is thus seen to be more methodological than substantive. In a world in which there were occasional occurrences of “identical environmental twins,” whose experiences were exactly the same, moment by moment, and another variety who shared exactly (but randomly) 50% of their experiences, environmentalists could re-produce the precision of their ri-vals,andlikethebehavior geneticists could measure with great precision the total contribution of the environment while knowing almost nothing about the developmental processes that underlie it.
So the difference between genes and environment is not that genes are somehow more important, it is that their effects are more easily quantified in the absence of any knowledge of what they are doing causally. Plomin doesn’t address this issue. Instead he finesses it with a distinction between environments that matter and environments that (don’t) make a difference. Environments matter in that 50% of the variability is environmental; they don’t make a difference because you can’t identify any particular environmental variable that has a measurably large causal effect.
The irony of this argument is that the exact same problem exists for genetics. Despite the success of the twin and family genetics project a couple of decades ago, and the contemporary success of the genetic risk score project, the candidate gene project– in which individual alleles were evaluated for causal effects one at a time– was a massive and embarrassing failure. Individual genes don’t have detectable causal effects any more than individual environmental variables do. But once again, the quantifiability of genes– this time in the form of genetic risk scores– came to the rescue of behavior genetics when it was reeling from the inability to find individual genetic causes for anything behavioral.
All of this is related to the idea of The Gloomy Prospect, a phrase from Plomin and Daniels (1991) that I picked up on and wrote about for many years. The Gloomy Prospect is another instance, not quite as egregious as the last one, where Plomin carefully rewrites history to make it seem as though an important idea in behavior genetics, one that he was originally wrong about, was actually his idea all along. That will be the topic of the next post.
(Thanks to Jonathan Kaplan, who discusses some of these issues in an [upcoming?] review of Blueprint.)