THE MISMEASURE OF MANall developed expertise in two areas, then most unfamiliar (if not
anathema) to practicing paleontologists: statistics and computers.
I was therefore trained in the statistical analysis of genetically
based variation within and between populations—again, the key
subject of The Mismeasure of Man (for Homo sapiens is a variable bio-
logical species, no different in this regard from all the other organ-
isms I had studied). I think, in other words, that I approached the
mismeasure of man with requisite and unconventional expertise
from an appropriate profession that has not often enough pro-
moted its special say about a subject so close to its center.
In writing numerous essays on the lives of scientists, I have
found that books on general topics or full systems usually originate
in tiny puzzles or little troubling issues, not usually from an abstract
or overarching desire to know the nature of totality. Thus, the sev-
enteenth-century scriptural geologist Thomas Burnet built up to a
general theory of the earth because he wanted to know the source
of water for Noah's flood. The eighteenth-century geologist James
Hutton developed an equally comprehensive system from an initial
niggling paradox: if God made soil for human agriculture, but soil
derives from erosion of rocks; and if the erosion of rocks will ulti-
mately destroy the land and put the entire earth under water, then
how could God choose a means of our eventual destruction as a
method for making the soil that sustains us? (Hutton answered by
inferring the existence of internal forces that raise mountains from
the deep, thus developing a cyclical theory of erosion and repair—
an ancient world with no vestige of a beginning, or prospect of
an end.)
The Mismeasure of Man also began with a tiny insight that stunned
me with a frisson of recognition. Our young Turk generation of
paleontologists linked statistics and computers by learning the tech-
nique of multivariate analysis—that is, the simultaneous statistical
consideration of relationships among many measured properties of
organisms (length of bones, perhaps, for fossil species, performance
on numerous mental tests for humans in the mismeasure of man).
These techniques are not all conceptually difficult; many had been
partly developed or envisioned earlier in the century. But practical
utility requires immensely long computations that only became pos-
sible with the development of computers.
I was trained primarily in the granddaddy of multivariate tech-