* At least in his early work. Later, as we have seen, he abandoned the word intelligence as a result of its maddening ambiguity in common usage. But he did not cease to regard g as the single cognitive essence that should be called intelligence, had not vernacular (and technical) confusion made such a mockery of the term.
* Richard Herrnstein and Charles Murray emphasize the same arguments to obviate a charge of racism against The Bell Curve (1994)—see first two essays at end of book.
* Of Burt’s belief in the innateness of intelligence, Hearnshaw writes (1979, p. 49): “It was for him almost an article of faith, which he was prepared to defend against all opposition, rather than a tentative hypothesis to be refuted, if possible, by empirical tests. It is hard not to feel that almost from the first Burt showed an excessive assurance in the finality and correctness of his conclusions.”
* Sometimes, Burt descended even further into circular illogic and claimed that tests must measure innate intelligence because the testers constructed them to do so: “Indeed from Binet onwards practically all the investigators who have attempted to construct ‘intelligence tests’ have been primarily searching for some measure of inborn capacity, as distinct from acquired knowledge or skill. With such an interpretation it obviously becomes foolish to inquire how far ‘intelligence’ is due to environment and how far it is due to innate constitution: the very definition begs and setdes the question” (1943, p. 88).
* Hearnshaw (1979) suspects that this paper marks Burt’s first use of fraudulent data.
* This accidental variance, representing peculiarities of particular testing situations, forms part of what statisticians call “measurement error.” It is important to quantify, for it may form a basic level of comparison for the identification of causes in a family of techniques called the “analysis of variance.” But it represents the peculiarity of an occasion, not a quality either of a test or a testee.
* “’Other scholars often complained of Burt’s tendency to obfuscate, temporize, and argue both sides as his own when treating difficult and controversial issues. D. F. Vincent wrote of his correspondence with Burt about the history of factor analysis (in Hearnshaw, 1979, pp. 177–178): “I should not get a simple answer to a simple question. I should get half a dozen foolscap sheets of typescript, all very polite and very cordial, raising half a dozen subsidiary issues in which I was not particularly interested, and to which out of politeness I should have to reply … I should then get more foolscap pages of typescript raising more extraneous issues.… After the first letter my problem has been how to terminate the correspondence without being discourteous.”
* One might resolve this apparent contradiction by arguing that Burt refused to reify on the basis of mathematical evidence alone (in 1940), but did so later when independent neurological information confirmed the existence of structures in the brain that could be identified with factors. It is true that Burt advanced some neurological arguments (1961, p. 57, for example) in comparing the brains of normal individuals and “low grade defectives.” But these arguments are sporadic, perfunctory, and peripheral. Burt repeated them virtually verbatim, in publication after publication, without citing sources or providing any specific reason for allying mathematical factors with cortical properties.
* Hearnshaw (1979) reports that Burt had greatest influence over the 1938 Spens report, which recommended sorting at 11 plus and explicitly rejected comprehensive schooling under a single roof thereafter. Burt was piqued at the Norwood report because it downgraded psychological evidence; but, as Hearnshaw notes, this annoyance “masked a basic agreement with the recommendations, which in principle did not differ so much from those of the Spens committee, which he had earlier approved.”
* The recycling reached full and lengthy fruition when Herrnstein and Charles Murray used the same claim as the opening gambit and general basis for The Bell Curve (1994).
* Thurstone reified his factors, calling them “primary abilities,” or “vectors of mind.” All these terms represent the same mathematical object in Thurstone’s system—factor axes placed near clusters of test vectors.
* Readers who have done factor analysis for a course on statistics or methodology in the biological or social sciences will remember something about rotating axes to varimax positions. Like me, they are probably taught this procedure as if it were a mathematical deduction based on the inadequacy of principal components in finding dusters. In fact, it arose historically with reference to a definite theory of intelligence (Thurstone’s belief in independent primary mental abilities) and in opposition to another (general intelligence and hierarchy of lesser factors) buttressed by principal components.
* Thurstone, like Burt, submitted many other sets of data to factor analysis. Burt, chained to his hierarchical model, always found a dominant general factor and subsidiary bipolars, whether he studied anatomical, parapsychological, or aesthetic data. Thurstone, wedded to his model, always discovered independent primary factors. In 1950, for example, he submitted tests of temperament to factor analysis and found primary factors, again seven in number. He named them activity, impulsiveness, emotional stability, sociability, athletic interest, ascendance, and reflectiveness.
* Tuddenham (1962, p. 516) writes: “Test constructors will continue to employ factorial procedures, provided they pay off in improving the efficiency and predictive value of our test batteries, but the hope that factor analysis can supply a short inventory of’basic abilities’ is already waning. The continuous difficulties with factor analysis over the last half century suggest that there may be something fundamentally wrong with models which conceptualize intelligence in terms of a finite number of linear dimensions. To the statistician’s dictum that whatever exists can be measured, the factorist has added the assumption that whatever can be ‘measured’ must exist. But the relation may not be reversible, and the assumption may be false.”
* The brouhaha over sociobiology during the past few years was engendered by this hard version of the argument—genetic proposals (based on an inference of adaptation) for specific human behaviors. Other evolutionists call themselves “sociobiologists,” but reject this style of guesswork about specifics. If a sociobiologist is anyone who believes that biological evolution is not irrelevant to human behavior, then I suppose that everybody (creationists excluded) is a sociobiologist. At this point, however, the term loses its meaning. Human sociobiology entered the literature (professional and popular) as a definite theory about the adaptive and genetic basis of specific traits of human behavior.
* Lest homosexuality seem an unlikely candidate for adaptation since exclusive homosexuals have no children, I report the following story, advocated by E. O. Wilson (1975, 1978). Ancestral human society was organized as a large number of competing family units. Some units were exclusively heterosexual; the gene pool of other units included factors for homosexuality. Homosexuals functioned as helpers to raise the offspring of their heterosexual kin. This behavior aided their genes since the large number of kin they helped to raise held more copies of their genes than their own offspring (had they been heterosexual) might have carried. Groups with homosexual helpers raised more offspring, since they could more than balance, by extra care and higher rates of survival, the potential loss by nonfecundity of their homosexual members. Thus, groups with homosexual members ultimately prevailed over exclusively heterosexual groups, and genes for homosexuality have survived.
Copyright © 1996, 1981 by Stephen Jay Gould. All rights reserved.
For information about permission to reproduce selections from this book, write to Permissions, W. W. Norton & Company, Inc., 500 Fifth Avenue, New York, NY 10110. Printed in the United States of America.
The text of this book is composed in Baskerville with the display set in Caslon 540. Composition and manufacturing by the Maple-Vail Book Manufacturing Group. Book design by Marjorie J. Flock
Library of Congress Cataloging-in-Publication Data
Gould, Stephen Jay.
The mismeasure of man/by StephenJay Gould.-Rev. and expanded.
p. cm.
Includes bibliographical references and index.
ISBN:978-0-393-34040-2 (e-book)
l.Intelligence tests-History. 2. Ability-Testing-History. 3. Personality tests-History. 4. Craniometry-History. I. Title.
BF431.G68 1996
153.9’3—09-dc20 95-44442
W. W. Norton & Company, Inc.
500 Fifth Avenue, New York, N.Y. 10110
www.wwnorton.com
W. W. Norton & Company Ltd.
Castle House, 75/76 Wells Street, London WIT 3QT
The Mismeasure of Man Page 51