Re-Measuring Morton

“Unconscious manipulation of data may be a scientific norm.” – Stephen Jay Gould

“You’re just a woman with a small brain.  With a brain a third the size of us.  It’s science.” – Ron Burgundy, Anchorman

The subject this week is a recent paper by Lewis et al. titled “The Mismeasure of Science: Stephen Jay Gould versus Samuel George Morton on Skulls and Bias.”  The paper refers to the 1978 paper by Stephen Jay Gould, “Morton’s Ranking of Races by Cranial Capacity,” which is in turn a critique of the work by Samuel George Morton in the mid-19th century.

Background

Let’s start at the beginning.  Morton lived in a time when an important question was whether the various human populations were separately divinely created species (polygenesis), or a single divinely created species that subsequently diverged (monogenesis).  This was before Darwin, remember.  To address this question, Morton collected over 600 skulls from around the world, and measured their cranial capacity, with results showing variation in average capacity between groups, supporting his argument for polygenesis.

Gould argued instead that Morton skewed his data to support his a priori convictions, and re-analyzed Morton’s measurements, showing that “there are no differences to speak of among Morton’s races.”  Based in large part on Gould’s work, which is “widely read, frequently cited, and still commonly assigned in university courses,” Morton is now– or was, anyway– regarded as “a canonical example of scientific misconduct.”

Which brings us to this most recent study, in which a group of anthropologists actually re-measured the skulls in Morton’s collection (which Gould never did).  Their results contradict Gould, showing that Morton’s data and methodology were reliable “despite his clear bias.”  Indeed, “Gould’s own analysis of Morton is likely the stronger example of a bias influencing results.”

Thoughts

Why is this important?  I think that both Gould and Lewis discuss two related but very different issues, and I am really only interested in one of them here:

1. Whether Morton’s conclusions were wrong; and
2. Whether Morton’s conclusions were wrong due to a priori bias influencing methodology.

I am less interested in (1); although Morton’s reputation as an empiricist deserves vindication, I think the actual subject of cranial capacity variation and its support for monogenesis or polygenesis are no longer terribly relevant (unless you are still a creationist, in which case I am curious as to how this issue gets resolved).  I instead want to focus on (2).  The valid question raised by Gould is: is it even possible to conduct scientific study and obtain objective data and conclusions, when we are always influenced by our own preconceptions, biases, and beliefs?

(Aside: This same question has come up here before, particularly in the comments on this early post.  Interestingly, on re-reading that post, I see some similarities with another of Gould’s ideas, “non-overlapping magisteria” (NOMA).  But I think my conclusion differs from Gould’s, in that mine is not quite as friendly to actual practiced religions.)

Morton’s study is a vivid example of this question.  Reading his work, the racist views common in the 19th century seem outrageous today, and would seem to clearly indicate a priori bias.  And he is certainly not the only example of such bias; as Gould points out, Isaac Newton and Gregor Mendel also both “fudged” results to support some of their theories.

An interesting related question is: how much does it matter, at least in the end game?  As Gould puts it, “After all, Newton and Mendel were right.”  In other words, the world works a certain way no matter whether or how we observe or comment on it.  There are “right” answers out there, independent of our questionable motives, clumsy methodologies, or erroneous refutations.  As another example, consider one creationist argument that rejects the theory of evolution because Darwin was motivated by a desire to reject God.  On one hand, I think this is simply wrong, or at least a misrepresentation of the man.  But on the other more important hand, who cares?  Motivation doesn’t make a theory wrong, observation does.

In the end, based on their re-analysis of Morton’s work, Lewis et al. reject Gould’s suggestion that “unconscious or dimly perceived finagling is probably endemic in science, since scientists are human beings rooted in cultural contexts, not automatons directed toward external truth.”  I share their optimism and confidence in our ability to be objective despite our very human biases… but I also think that Gould’s warning is well heeded, and that our current greatest disease as a scientific community is an insupportable pressure to publish positive and unambiguous results, when the overwhelming majority of experimental results should be expected to be negative or ambiguous.  Many of us probably learned the scientific method in school as a simple short sequence of steps, when in reality, as I find I am very fond of saying, science is messy.

Stephen Jay Gould died in 2002.  It is extremely unfortunate that he is no longer with us.  I would be interested to hear his response to these latest developments.