What Galton Wrought (Part 2)
Nature-Nurture-Noise (11)
Let’s assume for the moment that Galton’s Nature-Nurture dichotomy wasn’t fundamentally wrong-headed in ways well covered in a number of previous critiques (http://hdl.handle.net/10822/546678; 4827368; 10.1111/j.1532-7078.2011.00079.x), and that Nature and Nurture can be well defined. For Galton, Nature represented something that is innate (present at conception) and heritable, that is transmissible from parent to offspring. Nature came to be identified with genes, and later, genomes. Nurture was particularly ill-defined and amounted to anything that was not Nature. Nurture comes by way of the environment, influences from without.
Since Galton was also the father of eugenics, you know on which side he was wagering. Eugenics approximately means “well born”. Well born, not in the sense of the advantages you might enjoy because you were born into a wealthy, aristocratic family. But well born in the sense that, as a result of your parents superior genetic endowment, you will make an ideal British citizen. As it happens, the two ways of being well born are not easy to separate. Galton thought he had a pretty good answer as to why that might be. But he needed proof. Identical twins provided the opportunity to obtain it, or so he thought.
Galton attributed all of the uncanny similarities of twins to their shared (genes) nature, naturally. His advocacy of eugenics was founded on the strong belief that it wasn’t so much Nature-Nurture as NATURE-Nurture, and those “identical” twins seemed ample demonstration.
That was the state of play in human genetics from the turn of the 20th century through much of World War II. It was also a common view among non-geneticists, including the prominent psychologists, Edward Thorndike and Robert Yerkes. Moreover, it was common view within the United States Government and influenced public policy. Laws were passed, first in Virginia, then by the federal government, to enact forced sterilization of the “genetically inferior”, and to restrict immigration to all but a few countries, in which the populations were deemed “genetically superior”, mostly based on racial criteria. Data—questionnaires to diagnose feeble-mindedness-- was collected by an offshoot of the Carnegie laboratories at Cold Springs Harbor, which later came to be called the Eugenics Record Office. The data were invaluable for the sterilization campaign.
The Nazis were inspired by the Eugenics Record Office and other American advances in implementing eugenics. They were to build on these foundations the notion of the risk to Aryans of racial degradation.
World War II did not serve the eugenics cause well, primarily due to rapidly increasing anti-German sentiments. Eugenics and the genetic determinism that scaffolded it briefly fell out of favor. America’s formative role in nurturing eugenics was largely glossed over.
Even before the war there were scientists who emphasized the role of the environment in shaping individual differences. Some went to the opposite extreme with respect to Galton’s scheme, most famously, John Watson, the father of what came to be called Behaviorism. Watson famously claimed that given a dozen healthy infants, he could, through environmental manipulations, make one a doctor, another an artist and another a “merchant-in-chief”. On his view, infants were “blank slates” (tabula rasa) on which a good psychologist could write their fates.
B.F. Skinner was Watson’s post-war heir with respect to environmental determinism. He advocated for a socially engineered utopia in which efficiency and harmony could finally prevail. His Beyond Freedom and Dignity culminates with: “we have not seen what man can make of man”. He was not talking about eugenics—genetic engineering—but rather social engineering. In 1971, when the book was published, Skinner’s views, though controversial, were well received in many Psychology departments where the Behaviorist philosophy still reigned. But things have changed markedly since then. You’d be hard pressed to find a self-proclaimed Behaviorist now.
By the time Steven Pinker wrote The Blank Slate in 2002, he was flogging a dead horse, more aptly, the bones of a horse carcass that he went to great lengths to disinter. By then there existed a consensus that it wasn’t one or the other—genes or environments—but both, the more ecumenical going so far as to claim that, for complex traits such as cognitive capacities or personality, genes and environment were about equally responsible. As I will argue later, this is a pseudo sophisticated pseudo solution to a question that was poorly framed. We would all be better off if it had never been posed.
But the reality is that Galton’s dichotomy seems quite natural for non-scientists and scientists alike. Moreover, for most non-scientists some degree of naïve genetic determinism prevails as both Folk Psychology and Folk Biology. Hence the credulity with which reports like this one is met: Two identical male twins, separated at birth: both were named Jim; both married a woman whose first name was Linda, subsequently divorced, then remarried, their second wives both named Betty. Are we to conclude, then, that not only is the name of your first wife, but your second as well, encoded in your genes? What if one of the Jims had moved to Mexico? Would he have still managed to track down a Betty? Then there is the matter of their own identical first names. Was that written in their genes as well? If so, it’s a good thing they were separated. Oh, and they both had a dog named Toy.
Folk genetic determinism is pervasive. It has always been so. Galton was merely trying to provide a framework that would irrefutably demonstrate these pervasive intuitions, which he clearly shared. For the last 50+ years The Nature-Nurture debate has been fueled by Galton’s heirs, that is by those with a bias toward demonstrating the primary role of Nature. This bias is clearly evident in Behavior Genetics, for which twin studies are the sine qua none. This is not disinterested science, it is science with an agenda.
Insightful critiques of twin studies are legion. Here, I will only focus on the most fundamental defect, one that can be traced all the way back to Galton: the notion that genomic effects can be measured independently of environmental effects and vice versa. For Galton and his heirs, it seemed perfectly reasonable to measure a genetic contribution to individual differences and a separate environmental effect on individual differences. When these two quantities are summed (added) you have a complete account of those individual differences.
In science, as the philosopher Alfred North Whitehead wisely advised, we should “seek simplicity and distrust it”. Galton, like many scientists involved in the nature-nurture dispute, didn’t get past the first part of this injunction. He certainly kept it simple. The assumption that the relationship between genetic and environmental effects is additive is a magnificent simplifying assumption (Conventionally denoted as G + E). Additivity is the simplest possible relation that any two causes can have. But the additivity assumption is not just simple, it is simplistic.
The overwhelming majority of statistical tools used in the measure of heritability assume that you can just add genomic effects to environmental effects to account for whatever trait—from aggression to lung cancer—is to be explained. As such they are insensitive to non-additive effects, which fall under the broad banner of genome-environment interactions (henceforth G x E). By genome-environment interactions, I simply mean that the genomic effects on, say, aggression, depends upon the environmental conditions experienced during development. And vice versa. If such interactions exists, it makes no sense whatsoever to calculate separate genomic and environmental effects and add them up.
Statistical models that assume you can add things up are called linear because they can be graphed as a straight line in Cartesian coordinates. Graphically, Non-linear models can take an incalculable variety of forms—every form, in fact, other than the linear. The most famous equation in all of science, E=mc 2 is nonlinear. Einstein’s revolutionary equation is on the simple side of the non-linear spectrum. At the complex end are equations that describe phenomena that fall under the rubric of chaos. Many biological phenomena, including those related to brain development are close to chaotic.
In any scientific investigation it makes sense to start with simple models, that is linear equations. Problems arise though if the simple linear models are retained even in the face of overwhelming evidence of their deficiency. The desire for simplicity can be perverted into an irrational faith in the linear.