Hoisting Steven J. Gould by his own petard

Steven Jay Gould’s essays in Natural History Magazine were as close to matters scientific my late father, an attorney, ever came. He loved them. People of that generation were cowed by science, thinking it beyond them.

Naturally, I’ve read a bit more, following the controversies about punctuated equilibria, the actual unit of natural selection etc. etc. as they appeared in the pages of Nature over the years. In particular, I found Gould’s book “The Mismeasure of Man” intriguing (and quite convincing)  It concerns one Samuel Morton, who measured cranial capacities of various ethnic groups in the early 1800s. He found that Europeans have bigger brains than everyone else. Gould accused Morton of (consciously or unconsciously) manipulating the data to come up with the conclusions he desired.

In particular, Gould accused him of grouping the data by arbitrarily amalgamating Native American populations, while breaking down the Europeans into subgroups.

Well, guess what? An anthropologist [ PLoS Biol. doi:10.1371/journal.pbio.1001071;2011 ] went back to Penn (where the skulls in question reside), and remeasured some 300 of them, blinding themselves to their ethnic origins as they did. Morton’s measurements were correct.  They also had the temerity to actually look at Morton’s papers.  They found that, contrary to Gould, Morton did report average cranial capacities for subgroups of both populations, sometimes on the same page or on pages near to figures that Gould quotes, and therefore must have seen.  Even worse (see Nature vol. 474 p. 419 ’11 ) they claim that “Gould misidentified the Native American samples, falsely inflating the average he calculated for that population”.  Gould had claimed that Morton’s averages were incorrect.

Certainly we all see what we want to see.  It happens all the time in medicine.  Patients want to get better.  Docs want them to get better.  It’s a plucky patient who will have the guts to say “No, I’m not better” when a doc asks them “You’re feeling better aren’t you”.  Of course, this is the wrong way to ask.  There is an art to taking a history and asking questions in such a way as to not get the answers you want (just the opposite of the cross examination skills of the lawyer).

Why would Gould do this?  Well, he hated racism, and wanted to discredit what he saw as a scientific basis for it. Coming of age after the horrors of World War II, he had particular reason to do so. After the war all sorts of distinguished biologists took an ad in the Times saying that there was no scientific basis for the concept of race.  Now with single nucleotide polymorphisms and linkage disequilibrium we’re not so sure.  All sorts of papers presently appear in reputable journals describing just how much European admixture is present in the genomes of various ethnic groups.

Moreover Gould was a man of the left. Morton was far from his only intellectual battle (which he, in effect, had won until now).   His and Lewontin’s treatment of E. O. Wilson and sociobiology was particularly appalling.  They regarded it as recrudescent social Darwinism.  The idea that humanity was not infinitely malleable and that they couldn’t be reshaped in thought and action by their environment, struck at their hopes that changing the system could end all social evil.  See http://en.wikipedia.org/wiki/New_Soviet_man.

I did try reading  Gould’s  “The Structure of Evolutionary Theory” and got through about 80 of its 1400+ pages or so before giving up.  It was rather Rabelaisian, frankly incomprehensible and in serious need of an editor.  Only after I saw him lecture in person at an affair my brother had set up at a fancy Washington Club (Cosmos) did I see why.  I’m not sure what he was talking about, but he appeared agitated and disorganized.  A neurologist friend and I sat together and spent the lecture trying to decide if he was manic and off his meds or actually psychotic.  Gradually the hall emptied as he continued to emote.

How much of his work will remain standing after the re-examinations likely to ensue after this, is anyone’s guess.

Post a comment or leave a trackback: Trackback URL.


  • Curious Wavefunction  On June 26, 2011 at 10:13 pm

    I thought it was precisely SNPs that told us that the variations among individuals within a “race” are more than those between “races”. Gould was a very smart and eloquent man (and for my money’s worth one of the best science writers of the twentieth century), but his treatment of E O Wilson- a gentle scholar- was indeed shoddy.

  • luysii  On June 27, 2011 at 8:05 am

    True enough for single SNPs, but combine them and you get something else. Here’s some background.

    A haplotype is a string of alleles at closely linked (adjacent) loci on a single chromosome. The alleles don’t have to be in genes, they can be single nucleotide polymorphisms (SNPs) close together in the sequence of nucleotides comprising the DNA. The point is that a particular combination of alleles (a particular haplotype) are inherited as a block because they are so close together that recombination between them during meiosis is extremely unlikely. The elements of a haplotype travel together — this makes haplotyping to find disease associated genes — much faster than looking at all 10,000,000 SNPs. First the haplotypes have to be found — this is the remit of the 100 million$ human haplotype map (HapMap).

    Linkage disequilibrium is the condition in which the haplotype frequencies in the population deviate from the values they would have if the allele frequencies of each gene in the stretch of DNA included in the haplotype were combined at random.

    [ Science vol. 324 p. 575 ’09 ] A study of 98 American blacks (from 4 states), 3194 blacks from 113 African populations and 952 people from all over showed that 71% of the American black genome is from western africa, 8% from other parts of Africa and 13% from Europeans (leaving them with a deficit of 8%. [ Science vol. 324 pp. 1035 – 1044 ’09 ] Another study which appears to reach the same conclusion, but with slightly different numbers (121 african population 60 non-African populations and 4 African American populations. They studied variation at 1327 nuclear microsatellite and insertion/deletion markers.

    Not quite the same thing as race, but close.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: