1.2 History as a Beaker of Mustard Seed 2 – Of Skulls & Glass Slippers

Posted on September 16, 2010


From Featherfolk, a work in progress.

How could sentimentalists and egalitarians stand against the dictates of nature?  Morton had provided clean, objective data based on the largest collection of skulls in the world.”  –Stephen Jay Gould

Samuel George Morton had put together a chart which appeared to prove that persons of his own ethnicity had larger brains.  It was impressive.  And very very scientific.

But there were problems.

In 1977, Gould carefully reassessed the published raw data upon which Morton’s conclusions were based, double-checked the math and reviewed the research methodology.  Morton had left a full and detailed record of his researches, complete enough for Gould to discover how the eminent polygenicist and craniometrician had arrived at his figures.

In the end, based on Morton’s own published materials, Gould concluded:  “My correction of Morton’s conventional ranking reveals no significant differences among races for Morton’s own data.”

Gould’s results—based as I said on the identical figures and measurements that Morton himself used—were the opposite of Morton’s results.  How could this be?  The operations involved in measuring and tallying skull capacities seem to be obvious, uncontroversial and simple, capable one would imagine of being duplicated in any junior high school science lab (given that the local school board had gotten in a good supply of human skulls beforehand.)  How had the eminent Dr. Morton gone wrong?

It turns out that there are pitfalls in measuring skulls.

Now a skull is a little like the glass slipper in the Cinderella story.  You may or may not be aware that the presence of the glass slipper in that story is an accident of translation.  Somebody mistook the word for fur as being the word for glass, and the fur slipper in the earlier version of the story became the glass slipper in the version most of us know today.

I learned about this curious error and thought, well, that may be so, but the glass slipper works better as a plot device.  If you’ve strayed too far from childhood to remember the Cinderella story—or you’re just stopping off for a rest and a read before continuing on your way through the solar system—I’ll remind you that in the story the prince conducts a search for Cinderella using the slipper as a way to identify her.  If the slipper fits, then he’s found his girl, and everybody lives happily ever after, etc. etc.

Glass footwear, however, serves better than fur in advancing the plot.  Fur is not like glass.  If you have every debutant and tavern wench in the kingdom stuffing her foot furiously into a fur slipper in hopes of bagging the prince, eventually the slipper is going to stretch.  By the time Cinderella dips her foot into it, the slipper could be so sad and out of shape that it would no longer fit her.  Not so, glass.

Why am I talking about fairy stories?  As I said, skulls are a little like that glass slipper in the sense that they don’t stretch or shrink.  You can reasonably expect them to stay the same size, to retain their dimensions inside and out over the course of time.  Anybody ought to be able to measure them and get the same answers as Dr. Morton.  That fact that skulls don’t shrink or stretch like fur slippers is probably one of the reasons why his researches were so compelling and convincing at the time they came out.  There is something certain about a skull.

But if that is what we think, then we are forgetting about feet.

As every fashion martyr knows, feet are squeezable.  I think Cinderella has to count herself lucky that some damsel in line before her didn’t have feet just a sliver larger than hers.  Feet of not quite the same size could be squeezed into glass slippers that maybe didn’t quite fit, and made to look like they belonged there anyway.

And before Prince Charming was any the wiser, he’d already be on his honeymoon with totally the wrong princess, the unhappy victim of flawed methodology.

Mustard seeds, like feet, are compressible.  You can pack them tighter in one skull than in another, and therefore fit more seed into one than in the other, even if the skulls being investigated are actually the same size.  If you reverse your policy in the second stage, when the seeds are poured from skull into beaker jar for measurement—that is, if you loose pack the seeds from the skull you favour, and tight pack the seeds from the skull you do not favour—you can edge the results even closer to what your instincts (and prejudices) tell you is right.  Fudge, and double fudge, and who’s to complain?

Except people who don’t like racism posing as science, and maybe scientists persnickety about accuracy.

A particular joy of cheating in this way, if you are Dr. Morton, is that you can do it without your really being aware that you’re doing it.  Not all cheating and fudging is done consciously, and it’s much more fun to get the results you want if you can also pretend that the results are genuine.  Fudge, double fudge, with a cherry on top.  We won, cries the eminent doctor, and we didn’t even lie.

Cross my heart and hope to die.

How imprecise is mustard seed?  Dr. Morton found that separate measurements of the same skull using mustard seed might differ by 5% or more—equivalent on the average to about four cubic inches, give or take a seed or two.[1] How did this variation affect the results?  Morton conveniently provided us with the means to find out.

Dr. Morton was enough of a scientist not to like a margin of error of 5% in his measurements.  When he decided to switch over to the more accurate BB shot method, he went back to remeasure some of the skulls he had measured using mustard seed.  This gave Stephen J. Gould a basis to compare the results from each method.  What Gould found was a consistent discrepancy, a discrepancy, that is, which followed a pattern.

In all cases, using mustard seed produced a lower number of cubic inches than using BB shot; statistically all the skulls ‘grew’ when remeasured using BBs.  Why this would be so is difficult to say for certain, except that perhaps it is easier to compress mustard seed in a beaker than it is in a skull, and that therefore the same amount of mustard seed would tend to take up less volume in the beaker than in the skull.  This would result in the volume of skulls being consistently underestimated.  BBs, on the other hand, resisted compression in either skull or beaker.

But while all skulls shrank when measured using mustard seed, how much a skull shrank depended on the “human species” it had been assigned to by Morton.  Caucasian skulls suffered the least shrinkage, an average of 1.8 cubic inches.   Indigenous American skulls were diminished by 2.2 cubic inches, African skulls by a whopping 5.4 cubic inches.  It is fairly clear which skulls were favoured by Morton and which were not, and the process was so consistent that we can even quantify the degree of favour.

Gould identified other problems.

While planning and organizing his research, Morton failed, for instance, to address the methodological pitfalls that flow from human variation.  Within each human group there are people with large skulls and people with small skulls.  Men tend to have larger skulls than women.  In Morton’s Indigenous American sample, Iroquois skulls were larger than Inca; and in his Caucasian sample, English and German skulls were larger than South Asian.  None of this variation would be relevant if you could measure the skulls of every member of every human group, because the numbers you produced then would be precisely the numbers you were looking for.  The experimental challenge is to somehow produce accurate and comparable numbers using a smaller but still representative sample.

Dr. Morton did not meet this challenge.

[1] I’m not sure how a 5% margin of error translates in terms of glass slipper sizes, but I think we can all agree that Prince Charming maybe ought to ask Cinderella’s number next time, rather than engaging in dodgy searches involving impractical footgear.


Posted in: Featherfolk