Gould is the author of The Mismeasure of Man, a study of the history of psychometrics and intelligence testing as a form of scientific racism; the most recent edition includes an attempted refutation of the arguments of The Bell Curve.
Here it is the book:
I've read this book twice. It's an annotated history of the incorrigible human need to feel superior to others on group rather than individual grounds, especially heritable intelligence. It boggles the mind to learn how much brainpower, education, patience, hard work, and ingenuity have been turned to the question of "proving" that social distinctions are scientifically valid and immutable. One of the most salutary themes of the book is a reminder that almost every group has been "proven" intellectually or morally inferior at some point: Italians, Slavs, Irish, English, Germans, Hispanics, Blacks, Jews, Asians, Indians, Arabs, Native Americans, women, and even the white American male himself! Americans tend to be obsessed with race, while Europeans have devoted more time to issues of class and religion -- but everyone wants to believe they're smarter than the next guy.
Gould plays fair by choosing to take on the best theories of group intelligence, the fruit of celebrated minds such as Louis Agassiz, Paul Broca, Cesare Lombroso, Alfred Binet, and R.M. Yerkes. He also shoulders the unbelievably tedious task of replicating these scientists' results and/or checking their research methods. It's quite difficult to argue with his factual conclusions without revealing oneself as a poseur; and his critiques of their methodologies are equally difficult to deny. Gould also points out that many of the scientific racists/classists themselves came to recant their own ideas as they got older and presumably wiser.
The book gains relevance as it goes on. Much of the first half is devoted to "scientific" theories that no one could now take seriously, based on experimental methods like filling skulls with mustard seed, weighing the brains of famous men, measuring the spaces between the toes of criminals, and even simply claiming that beauty equals intelligence QED. The best part of the book is its examination of mass intelligence testing, and of the misuse of factor analysis. Gould's simple explanation of factor analysis goes a long way toward dispelling the urge to be cowed by statistical conjuring. Finally, in the 1996 edition Gould takes on the newest incarnation of mismeasurement, The Bell Curve.
Gould mostly refrains from speculating on the motives of his subjects, chalking it all up to the social construction of science or even allowing for benevolent intentions. He also avoids mention of the worst effects of scientific racism, such as the Nazi genocides. While I can understand why he took this low-key, professional approach, I wonder if his arguments are often too subtle for their intended audience. My experience has been that these types of theories are most appealing to precisely the demographic group least likely to be swayed by implicit historical arguments of the "Since these ideas are ridiculous, perhaps more recent but similar ideas are also ridiculous" type. This group seems to me to consist of people with high intelligence and strong faith in the purity of scientific truth, but low life experience, depth of knowledge, and critical acuity -- in other words, privileged undergraduates. Luckily, most people eventually realize that theories of biological determinism, no matter how emptily flattering to themselves, are an intellectual dead end and no substitute for individual achievement.