What Galton Wrought (4)

A Knight and a Fraud

Twin studies are considered the gold standard for measuring the genetic component of human traits, for those who mistake heritability for inheritability. But putting aside that elementary confusion, heritability estimates based on twin studies are plagued with misbegotten assumptions. And even if we play along in this game, there is a sordid history to consider, encompassing massaged data, cooked data, invented data, selective reporting of data, dubious experimental designs and sloppy research practices that don’t even begin to rise to the standards expected in any other domain of scientific research--including studies of heritability in other animals (and plants). Given this history, any reasonably dispassionate person should view even the most recent, state-of-the-art reports with deep skepticism. If, that is, she bothered to dig a little. Unfortunately, that kind of digging is not what the popular media is known for. But the lack of scrutiny goes beyond the popular media; scientists who should know better have fallen for this crap hook, line and sinker.

Broadly, the experimental designs most favored now are based on twins. There are two types of twins, monozygotic and dizygotic. Monozygotic (colloquially called identical) twins are genetically identical because they are the product of one fertilized egg, or zygote (monozygotic = one zygote). Very early in development, from days 4-13, the ball of embryonic stem cells divides in half, resulting in two balls of genetically identical embryonic stem cells.

Dizygotic (colloquially called fraternal) twins, on the other hand, occur when two eggs are independently fertilized and hence derive from two zygotes. (dizygotic = two zygotes.) As such, dizygotic twins are not more genetically similar than any other two siblings chosen at random. On average, their genetic similarity is 50%.  About half of dizygotic twins are expected to be same sex and half mixed sex.

From comparisons of these different twin types, measures of heritability have been derived for myriad human traits, including, height, neuroticism, religiosity, and most notoriously intelligence. This measure of heritability is the extent to which monozygotic twins are more similar than dizygotic twins. For this to be true, though, several assumptions must be made about genetics and environments, assumptions that are quite dubious.

Francis Galton initiated this line of research in the 1870s. He sent out questionnaire to 35 (self-reported) pairs of monozygotic twins and 20 pairs of (also self-reported) pairs of dizygotic twins, asking for anecdotes related to their similarities and dissimilarities, which he then compiled. There is no evidence that Galton ever made direct assessments of his subjects, or even visited them. He concluded from their anecdotes, nonetheless, that “nature prevails enormously over nurture”. As such he greatly feared for the genetic quality of future English citizens, given that the poorer—and hence dumber—portion of the population was far outbreeding the wealthy—and hence smarter—portion of the population. Thus, was eugenics born.

Galton’s First Acolyte

Cyril Burt met Francis Galton through his father, a physician who treated Galton’s brother at the estate where both Francis and his brother lived. Burt was young and impressionable, and Galton made a huge impression. One of Galton’s more lasting impressions on young Burt concerned causes of individual differences in intelligence. Burt inherited Galton’s hereditarian biases and managed to promulgate them intact. In carrying Galton’s torch, Burt had enormous influence on the English educational system in his various official and unofficial capacities. In essence his advice was to reduce investment in educating the poor; money spent there was pissing in the genetic wind. Better to devote scarce resources to those who could benefit most, the genetic elite, who also happened to be the economic elite.

But Burt recognized the shortcomings in Galton’s methodology and sought to put the hereditarian argument on firmer experimental grounds. His methodological breakthrough was to focus on monozygotic twins reared apart. Burt also directly tested the intelligence of his subjects but in ways that remain opaque. His estimtes for the heritability of intelligence were remarkably consistent. Too remarkably consistent, in fact, to have been empirically got. For example, in one study he reported a heritability of .771, that is, 77.1% of the intelligence in his study population had a genetic source. In a subsequent study, more than a decade later, in which he added a considerable number of new reared apart twins to his original population, he reported a heritability of, well, .771. The same number, down to the third decimal point. That just doesn’t happen in real science.

Fraud has always been a problem in science, increasingly so in the last two decades. But any contemporary fraudster would do a much better job in covering his/her tracks. Burt was no fool; there must have been a reason he thought he could get away with it. Arrogance, perhaps. He was knighted after all. (That’s Sir Cyril Burt, thank you very much.) Remarkably, he did almost get away with it. His fraud was not discovered in his lifetime. He died the most celebrated of English psychologists, praised for his influence on the English education system. Burt was largely responsible for instituting the 11-plus, the test—or its equivalent-- that countless English children take at that age to this day, which will largely determine their educational, and hence economic fate. There is not a more effective bastion for the English class-based social structure.

The Princeton psychologist, Leon Kamin ( ), did most of the heavy lifting in exposing Burt’s nefarious ways, which extended well beyond compiling data without the hassle of actual research. For example, he also seems to have invented two of his collaborators. Burt had a knack for making shit up.

Again, the question that screams for an answer: “why wasn’t he called out during his lifetime”? It goes beyond Knighthood. It goes to a prevailing hereditarian bias among Burt’s colleagues and those in political power, who found it agreeable to believe that their privilege was part of the natural order. No need to waste time and money ameliorating the condition of the great unwashed.

In the next post in this series I will turn to Burt’s American counterpart.