In "Women's Brains," how does Gould prove that Broca's data was not scientifically accurate?
Women's Brains is an essay reprinted in Stephen Jay Gould's 1980 book The Panda's Thumb. It deals with the common misconception of the day that gender plays any great role in intelligence.
Gould deals with the defining work of Paul Broca, a French scientist who used the pseudoscience of craniometrics to determine that women and non-whites had smaller brain capacity than males and whites, and so were innately inferior. While this method of measuring intellect and intelligence potential was debunked in the 20th century, the general opinion remained a steady part of public opinion and prejudice until new scientific evidence came to light.
To disprove this theory, Gould first writes:
I have the greatest respect for Broca's meticulous procedure. His numbers are sound. But science is an inferential exercise, not a catalog of facts. Numbers, by themselves, specify nothing. All depends upon what you do with them.
The point here is that Broca was not intentionally cooking the data, but simply used the data to "prove" the theory which he already believed. Broca also arrived at his conclusion after examining a very small data group: seven male skulls and six female skulls, all of which were from different age groups, making correlation impossible. Basically, the conclusion depended not on a controlled study of similar skulls from a wide range of ages and backgrounds, but on a tiny data-set that could not, by itself, provide enough evidence to prove or disprove the initial hypothesis. Gould concludes that biological or gender differences are largely irrelevant to intellect and potential.