The Mismeasure of Man

by Stephen Jay Gould

  Burt’s identification of group factors may seem, superficially, to challenge Spearman’s theory, but in fact it provided an extension and improvement that Spearman eventually welcomed. The essence of Spearman’s claim is the primacy of g, and the subordination of all other determinants of intelligence to it. Burt’s identification of group factors preserved this notion of hierarchy, and extended it by adding another level between g ands. In fact, Burt’s treatment of group factors as a level in a hierarchy subordinate to g saved Spearman’s theory from the data that seemed to threaten it. Spearman originally denied group factors, but evidence for them continued to accumulate. Many factorists began to view this evidence as a denigration of g and as a wedge for toppling Spearman’s entire edifice. Burt strengthened the building, preserved the preeminent role of g, and extended Spearman’s theory by enumerating further levels subordinate to g. The factors, Burt wrote (1949, p. 199), are “organized on what may be called a hierarchical basis.… There is first a comprehensive general factor, covering all cognitive activities; next a comparatively small number of broad group factors, covering different abilities classified according to their form or content.… The whole series appears to be arranged on successive levels, the factors on the lowest level being the most specific and the most numerous of all.”

  Spearman had advocated a two-factor theory; Burt proclaimed a four-factor theory: the general factor or Spearman’s g, the particular or group factors that he had identified, the specific factors or Spearman’s s (attributes of a single trait measured on all occasions), and what Burt called accidental factors, or attributes of a single trait measured only on a single occasion.* Burt had synthesized all perspectives. In Spearman’s terms, his theory was monarchic in recognizing the domination of g, oligarchic in its identification of group factors, and anarchic in recognizing s-factors for each test. But Burt’s scheme was no compromise; it was Spearman’s hierarchical theory with yet another level subordinate to g.

  Moreover, Burt accepted and greatly elaborated Spearman’s views on the differential innateness of levels. Spearman had regarded g as inherited, s as a function of training. Burt agreed, but promoted the influence of education to his group factors as well. He retained the distinction between an inherited and ineluctable g, and a set of more specialized abilities amenable to improvement by education:

  Although defect in general intelligence inevitably places a definite limit to educational progress, defect in special intellectual abilities rarely does so (1937, p. 537).

  Burt also declared, with his usual intensity and persistence, that the primary importance of factor analysis lay in its capacity for identifying inherited, permanent qualities:

  From the very outset of my educational work it has seemed essential, not merely to show that a general factor underlies the cognitive group of mental activities, but also that this general factor (or some important component of it) is innate or permanent (1940, p. 57).

  The search for factors thus becomes, to a great extent, an attempt to discover inborn potentialities, such as will permanently aid or limit the individual’s behavior later on (1940, p. 230).

  Burt on the vesication of factors

  Burt’s view on reification, as Hearnshaw has noted with frustration (1979, p. 166), are inconsistent and even contradictory (sometimes within the same publication).* Often, Burt branded reification of factors as a temptation to be avoided:

  No doubt, this causal language, which we all to some extent favor, arises partly from the irrepressible disposition of the human mind to reify and even to personify whatever it can—to picture inferred reasons as realities and to endow those realities with an active force (1940, p. 66).

  He spoke with eloquence about this error of thought:

  The ordinary mind loves to reduce patterns to single atomlike existents—to treat memory as an elementary faculty lodged in a phrenological organ, to squeeze all consciousness into the pineal gland, to call a dozen different complaints rheumatic and regard them all as the effect of a specific germ, to declare that strength resides in the hair or in the blood, to treat beauty as. an elementary quality that can be laid on like so much varnish. But the whole trend of current science is to seek its unifying principles, not in simple unitary causes, but in the system or structural pattern as such (1940, p. 237).

  And he explicitly denied that factors were things in the head (1937, P. 459).’

  The “factors,” in short, are to be regarded as convenient mathematical abstractions, not as concrete mental “faculties,” lodged in separate “organs” of the brain.

  What could be more clearly stated?

  Yet in a biographical comment, Burt (1961, p. 53) centered his argument with Spearman not on the issue of whether or not factors should be reified, but rather how they should be reified: “Spearman himself identified the general factor with ‘cerebral energy.’ I identified it with the general structure of the brain.” In the same article, he provided more details of suspected physical locations for entities identified by mathematical factors. Group factors, he argues, are definite areas of the cerebral cortex (1961, p. 57), while the general factor represents the amount and complexity of cortical tissue: “It is this general character of the individual’s brain-tissue—viz., the general degree of systematic complexity in the neuronal architecture—that seems to me to represent the general factor, and account for the high positive correlations obtained between various cognitive tests” (1961, pp. 57–58; see also 1959, p. 106).*

  Lest one be tempted to regard these later statements as a shift in belief from the caution of a scholar in 1940 to the poor judgment of a man mired in the frauds of his later years, I note that Burt presented the same arguments for reification in 1940, right alongside the warnings against it:

  Now, although I do not identify the general factor g with any form of energy, I should be ready to grant it quite as much “real existence” as physical energy can justifiably claim (1940, p. 214). Intelligence I regard not indeed as designating a special form of energy, but rather as specifying certain individual differences in the structure of the central nervous system—differences whose concrete nature could be described in histological terms (1940, pp. 216–217).

  Burt even went so far as to suggest that the all-or-none character of neural discharge “supports the demand for an ultimate analysis into independent or ‘orthogonal’ factors” (1940, p. 222).

  But perhaps the best indication of Burt’s hope for reification lies in the very title he chose for his major book of 1940. He called it The Factors of the Mind.

  Burt followed Spearman in trying to find a physical location in the brain for mathematical factors extracted from the correlation matrix of mental tests. But Burt also went further, and established himself as a reifier in a domain that Spearman himself would never have dared to enter. Burt could not be satisfied with something so vulgar and material as a bit of neural tissue for the residence of factors. He had a wider vision that evoked the spirit of Plato himself. Material objects on earth are immediate and imperfect representations of higher essences in an ideal world beyond our ken.

  Burt subjected many kinds of data to factor analysis during his long career. His interpretations of factors display a Platonic belief in a higher reality, embodied imperfectly by material objects, but discernible in them through an idealization of their essential, underlying properties on principal component factors. He analyzed a suite of emotional traits (1940, pp. 406–408) and identified his first principal component as a factor of “general emotionality.” (He also found two bipolar factors for extrovert-introvert and euphoric-sorrowful.) He discovered “a general paranormal factor” in a study of ESP data (in Hearnshaw, 1979, p. 222). He analyzed human anatomy and interpreted the first principal component as an ideal type for humanity (1940, p. 113).

  One needn’t, from these examples, infer Burt’s belief in a literal, higher reality: perhaps he thought of these idealized general factors as mere principles of classification to aid human understan
ding. But, in a factor analysis of aesthetic judgment, Burt explicitly expressed his conviction that real standards of beauty exist, independent of the presence of human beings to appreciate them. Burt selected fifty postcards with illustrations ranging from the great masters down to “the crudest and most flashy birthday card that I could find at a paper shop in the slums.” He asked a group of subjects to rank the cards in order of beauty and performed a factor analysis of correlations among the ranks. Again, he discerned an underlying general factor on the first principal component, declared it to be a universal standard of beauty, and expressed a personal contempt for Victorian ceremonial statuary in identifying this higher reality:

  We see beauty because it is there to be seen.… I am tempted to contend that aesthetic relations, like logical relations, have an independent, objective existence: the Venus of Milo would remain more lovely than Queen Victoria’s statue in the Mall, the Taj Mahal than the Albert Memorial, though every man and woman in the world were killed by a passing comet’s gas.

  In analyses of intelligence, Burt often claimed (1939, 1940, 1949, for example) that each level of his hierarchical, four-factor theory corresponded with a recognized category in “the traditional logic of classes” (1939, p. 85)—the general factor to the genus, group factors to species, specific factors to the proprium, and accidental factors to the accidens. He seemed to regard these categories as more than conveniences for human ordering of the world’s complexity, but as necessary ways of parsing a hierarchically structured reality.

  Burt certainly believed in realms of existence beyond the material reality of everyday objects. He accepted much of the data of parapsychology and postulated an oversoul or psychon—“a kind of group mind formed by the subconscious telepathic interaction of the minds of certain persons now living, together perhaps with the psychic reservoir out of which the minds of individuals now deceased were formed, and into which they were reabsorbed on the death of their bodies” (Burt quoted in Hearnshaw, 1979, p. 225). In this higher realm of psychic reality, the “factors of the mind” may have real existence as modes of truly universal thought.

  Burt managed to espouse three contradictory views about the nature of factors: mathematical abstractions for human convenience; real entities lodged in physical properties of the brain; and real categories of thought in a higher, hierarchically organized realm of psychic reality. Spearman had not been very daring as a reifier; he never ventured beyond the Aristotelian urge for locating idealized abstractions within physical bodies themselves. Burt, at least in part, soared beyond into a Platonic realm above and beyond physical bodies. In this sense, Burt was the boldest, and literally most extensive, reifier of them all.

  Burt and the political uses of g

  Factor analysis is usually performed on the correlation matrix of tests. Burt pioneered an “inverted” form of factor analysis, mathematically equivalent to the usual style, but based on correlation between persons rather than tests. If each vector in the usual style (technically called R-mode analysis) represents the scores of several people on a single test, then each vector in Burt’s inverted style (called Q-mode analysis) reflects the results of several tests for a single person. In other words, each vector now represents a person rather than a test, and the correlation between vectors measures the degree of relationship between individuals.

  Why did Burt go to such lengths to develop a technique mathematically equivalent to the usual form, and generally more cumbersome and expensive to apply (since an experimental design almost always includes more people than tests)? The answer lies in Burt’s uncommon focus of interest. Spearman, and most other factorists, wished to learn about the nature of thought or the structure of mind by studying correlations between tests measuring different aspects of mental functioning. Cyril Burt, as official psychologist of the London County Council (1913–1932), was interested in ranking pupils. Burt wrote in an autobiographical statement (1961, p. 56): “[Sir Godfrey] Thomson was interested primarily in the description of the abilities tested and in the differences between those abilities; I was interested rather in the persons tested and in the differences between them” (Burt’s italics).

  Comparison, for Burt, was no abstract issue. He wished to assess pupils in his own characteristic way, based upon two guiding principles: first (the theme of this chapter) that general intelligence is a single, measurable entity (Spearman’s g); second (Burt’s own idée fixe) that a person’s general intelligence is almost entirely innate and unchangeable. Thus, Burt sought the relationship among persons in a unilinear ranking of inherited mental worth. He used factor analysis to validate this single scale and to plant people upon it. “The very object of the factor-analysis,” he wrote (1940, p. 136), “is to deduce from an empirical set of test measurements a single figure for each single individual.” Burt sought (1940, p. 176) “one ideal order, acting as a general factor, common to every examiner and to every examinee, predominating over, though no doubt disturbed by, other irrelevant influences.”

  Burt’s vision of a single ranking based on inherited ability fueled the major political triumph in Britain of hereditarian theories of mental testing. If the Immigration Restriction Act of 1924 signalled the chief victory of American hereditarians in psychology, then the so-called examination at 11 + awarded their British counterparts a triumph of equal impact. Under this system for streaming children into different secondary schools, pupils took an extensive examination at age ten or eleven. As a result of these tests, largely an attempt to assess Spearman’s g for each child, 20 percent were sent to “grammar” schools where they might prepare for entry to a university, while 80 percent were relegated to technical or “secondary modern” schools and regarded as unfit for higher education.

  Cyril Burt defended this separation as a wise step for “warding off the ultimate decline and fall that has overtaken each of the great civilizations of the past” (1959, p. 117):

  It is essential in the interests alike of the children themselves and of the nation as a whole, that those who possess the highest ability—the cleverest of the clever—should be identified as accurately as possible. Of the methods hitherto tried out the so-called 11+ exam has proved to be by far the most trustworthy (1959, p. 117).

  Burt’s only complaint (1959, p. 32) was that the test and subsequent selection came too late in a child’s life.

  The system of examination at 11 + and subsequent separation of schools arose in conjunction with a series of official reports issued by government committees during twenty years (the Hadow reports of 1926 and 1931, the Spens report of 1938, the Norwood report of 1943, and the Board of Education’s White Paper on Educational Reconstruction—all leading to the Butler Education Act of 1944, which set policy until the mid-1960s when the Labour party vowed to end selection at 11 plus). In the flak surrounding the initial revelation of Burt’s fraudulent work, he was often identified as the architect of the 11+ examination. This is not accurate; Burt was not even a member of the various reporting committees, though he did consult frequently with them and he did write extensively for their reports.* Yet it hardly matters whether or not Burt’s hand actually moved the pen. The reports embody a particular view of education, clearly identified with the British school of factor analysis, and evidently linked most closely with Cyril Burt’s version.

  The 11 + examination was an embodiment of Spearmen’s hierarchical theory of intelligence, with its innate general factor pervading all cognitive activity. One critic referred to the series of reports as “hymns of praise to the ‘g’ factor” (in Hearnshaw, 1979, p. 112). The first Hadow report defined intellectual capacity measured by tests in Burt’s favored terms as i.g.c. (innate, general, cognitive) ability: “During childhood, intellectual development progresses as if it were governed largely by a single, central factor, usually known as ‘general intelligence,’ which may be broadly defined as innate, all round, intellectual [my italics for i.g.c] ability, and appears to enter into everything the child attempts to think, say, or do:
this seems the most important factor in determining his work in the classroom.”

  The 11+ owed its general rationale to the British factorists; in addition, several of its details can also be traced to Burt’s school. Why, for example, testing and separation at age eleven? There were practical and historical reasons to be sure; eleven was about the traditional age for transition between primary and secondary schools. But the factorists supplied two important theoretical supports. First, studies on the growth of children showed that g varied widely in early life and first stabilized at about age eleven. Spearman wrote in 1927 (p. 367): “If once, then, a child of 11 years or so has had his relative amount of g measured in a really accurate manner, the hope of teachers and parents that he will ever rise to a much higher standing as a late-bloomer would seem to be illusory.” Second, Burt’s “group factors,” which (for purposes of separation by general mental worth) could only be viewed as disturbers of g, did not strongly affect a child until after age eleven. The 1931 Hadow report proclaimed that “special abilities rarely reveal themselves in any notable degree before the age of 11.”

  Burt often claimed that his primary goal in supporting 11+ was a “liberal” one—to provide access to higher education for disadvantaged children whose innate talents might otherwise not be recognized. I do not doubt that a few children of high ability were thus aided, though Burt himself did not believe that many people of high intelligence lay hidden in the lower classes. (He also believed that their numbers were rapidly decreasing as intelligent people moved up the social ladder leaving the lower classes more and more depleted of intellectual talent—1946, p. 15. R. Herrnstein [1971] caused quite a ruckus with the identical argument, recycled, a few years back.)*