A mostly interesting history of the development of intelligence tests that studies in depth what is really being measured by them and whether intelligence is a 'thing' that can be given a simple score - the author's conclusion being that it cannot.

The earlier chapters describe the various fallacies of the attempts by scholars and scientists from about the eighteenth century onwards to classify intelligence, and reveal the racist, sexist and class based assumptions which relegated everyone but white men of the prosperous classes into a sliding scale of decreasing intelligence with black people at the bottom. It discusses the development by Binet of tests which were meant to identify children who were struggling at school and the areas they were having difficulties with, intended by him to help the children concerned but how these tests and ratings were taken up as a "thing" that could encapsulate a person's intelligence in a single score, treated as set in stone for their whole lifetime. The tests were misapplied and used to subsequently label most children as failures through schemes such as the 11 exam in the UK and the IQ tests in the USA based on the misuse of the results of testing on the USA army during WWI.

The author analyses in detail the long report on army testing which showed how the testing was fatally flawed for various reasons including, among other things, lack of buy-in from officers, verbal browbeating instructions of the picture-based tests given to men who were illiterate and often couldn't hear the instructor due to the rooms being overcrowded or who couldn't understand because they spoke little English, illiterate men being sent to do the written test because there were too many queued up to do the picture based tests, the picture based tests needing marks to be made on paper often by men who had never picked up a pen/pencil before, and too many tests crammed into the time allowed. All this information was in the massive report but no one read it and just took the 'doctored' figures in the summary which cut out a lot of the negativity.

The army tests were then used uncritically for decades as the basis for IQ tests. Gould gave a batch of the tests including the picture based ones to his students as an experiment and despite their fully comprehending what the tests were - another problem to the original people who had to be familiar with American culture which a lot of the immigrant recruits weren't - they couldn't finish a lot of them in the time given because it wasn't physically possible to mark the paper in time on long batteries of repetitive number based tests.

The final chapter deals with factor analysis which is a mathematical technique applied to test performance. Most of this went rather over my head, but I did grasp the fact that one of its major champions had been recently, at the time the book was published, revealed in detail to have falsified most of his conclusions. That part of the book was the least engaging for me and I had to break off from reading to refresh myself with some fiction in order to eventually get through it. So overall I would rate the book at 3 stars. ( )

Earnest, accurate and always interesting to read. ( )

> Nuit blanche, No 10 (automne 1983), pp. 35–36 : https://id.erudit.org/iderudit/21331ac
> Babelio : https://www.babelio.com/livres/Gould-La-mal-mesure-de-lhomme-Lintelligence-sous-...
> BAnQ (Québec science, 1983, Décembre) : https://collections.banq.qc.ca/ark:/52327/2874052

> LA MAL-MESURE DE L’HOMME, par Stephen Jay Gould, éditions Ramsay, 398 pages, traduit de l’américain, bibliographie, index des principaux noms, $22. — Le sous-titre : « L’intelligence sous la toise des savants » donne le sens du volume, histoire des tentatives scientifiques de mesurer l’homme, ses inégalités, etc. Le résultat est des plus discutable et a certainement favorisé un racisme dit « scientifique » sans bases réelles. À lire.
—Le devoir, 9 juil. 1983

The first half of this book is taken up with the science of phrenology; the next quarter or so with IQ tests getting their sea legs. The author did not convince me of his thesis which is that intelligence is a nebulous thing that tells us noting about a person. ( )

This book takes a hard look at early 20th century attempts by several psychologists and scientists to prove that human intelligence has genetic components defined and separated by factors such as race and ethnicity. Gould takes several of these scientists to task for egregiously biased methodology in testing and their conclusions.

While most honest, ethical scientists today would dismiss such claims of ethnic superiority, significant damage has been done by lingering refusal to accept the fallacy of such claims. Gould carefully exposes the errors and biases of these early pioneers in human intelligence. The book is somewhat long and tedious as a carefully written and documented academic account would naturally be, but it is a classic from an era that must be understood if we are to move beyond our understanding of human inequality. ( )