The Grand Locus / Life for statistical sciences



Genetics and racism (2)

In the first post of this series on genetics and racism, I explained how Richard Lewontin concluded from his work on human diversity that human races are of no value for taxonomy (the classification of living begins). This view was later criticized and even termed Lewontin's fallacy by A. W. F. Edwards. Yet, nobody ever doubted that Lewontin was honest in his approach. But more recently came another case that gives the shivers. The great Stephen Jay Gould, the author of the acclaimed Mismeasure of Man was accused of data manipulation.

The mismeasure of Gould

Stephen Jay Gould was this kind of scientist who pops up everywhere. I discovered him in a comment about the opinion of the Vatican on Evolution, others knew him for his statistical analyses of baseball records, while he was actually a paleontologist, author of the theory of punctuated equlibrium. But his most famous work is undoubtedly The Mismeasure of Man.

Like the author, the book is a strange chimera, somewhere in between scientific research and history, with a touch of lyricism. The Mismeasure of Man is a journey through the differences between people, or more precisely through the scientific discourse over this difference. From racism through IQ testing, he re-examines the evidence and shows how personal opinions or interests of the scientists of the time have systematically biased the debate to put white men at the top of a natural hierarchy. But under the light of The Mismeasure of Man, none of the facts speaks for a natural hierarchy between us.

And this is how the story was told until an article published in PLoS biology by Jason E. Lewis and his collaborators challenged the unbiasedness of Gould. The debate goes over the work of Samuel George Morton, the godfather of craniometry (the measurement of skulls). Throughout his life (1799-1851) Morton gathered hundreds of human skulls from different origins and analyzed their cranial capacity. His conclusions were that human races consist of four distinct acts of creation, Caucasians being at the summit with the largest cranial capacity.

The skulls are stored at the university of Pennsylvania, which allowed Gould and others to reanalyze Morton's work. Gould claims that Morton did not include all the skulls in his analyses and that his measurements were biased towards his prejudice. The evidence for differences in cranial capacities are simply inconclusive. Here is how he summarizes it in his own terms.

In short, and to put it bluntly, Morton's summaries are a patchwork of fudging and finagling in the clear interest of controlling a priori convictions.

Ironically, Jason E. Lewis and his collaborators claim that Gould did not include all the skulls in his analyses (while Morton did) and that his measurements were biased towards his prejudice (while Morton's were not). The clash between the allegations is discomforting. Who should we believe? And what are the arguments that matter when conclusions differ that much?

Lies, damn lies and statistics

I am not an expert of craniometry, and I do not have access to the Morton collection, so I cannot tell who is right about the measurements. But beyond measurements, there are statistical analyses and summaries, which often speak for themselves. Here is a passage of the book that drew my attention.

Morton's 144 skulls belong to many different groups of [American] Indians; these groups differ significantly among themselves in cranial capacity. Each group should be weighted equally, lest the final average be biased by unequal size of subsamples.

My own note, written in 2008 next to this passage reads “Here Gould's argument is stupid, if Americans consist of several populations with different means, it makes no sense to compute the population mean”. Gould has to admit that cranial capacity differs between groups in order to fight this very argument. Whether he made the measurements right or not, this sounds like bias.

It is such an irony that Gould repeatedly claims throughout The Mismeasure of Man that science is tainted with prejudice. It seems that he proved the point the best way he could have done. For his defense, he never claimed to be unbiased himself.

My original reasons for writing The Mismeasure of Man mixed the personal with the professional. I confess, first of all, to strong feelings on this particular issue.

It is sad to think that his passion led him to go after Morton's integrity. After all, nobody claims that cranial capacity has anything to do with intelligence any more. And even if Morton's measurements were correct, his interpretation did not dominate very long, so why not let sleeping do lie? It is all too easy to judge our ancestors by the sins of their time, while we have enough of our own.

Additional information:
Here is an account of the dispute by the university of Pennsylvania.

« | »

blog comments powered by Disqus

the Blog
Best of

the Lab
The team
Research lines

Blog roll
Simply Stats
Ivory Idyll
Bits of DNA