by Morton Hunt
In 1891 Cattell moved to Columbia University as professor of psychology and head of the department. He expanded his battery of tests and each year gave them to fifty volunteers from among the entering freshmen. His admirable aim was to prove that the tests measured intelligence by showing a relationship between the test results and the students’ grades; toward that end, he collected test data and student grades for close to a decade. Meanwhile the same method of testing intelligence was demonstrated at the Chicago World’s Fair of 1893, where Joseph Jastrow, a leader of the American Psychological Association, created a virtual replica of Galton’s Anthropometric Laboratory. Visiting psychologists undoubtedly found it interesting and impressive; during the 1890s such testing was begun in a number of laboratories in America and Europe.
By 1901 Cattell had collected enough data for a definitive study, and Clark Wissler, one of his students, performed a Galton-Pearson correlation analysis of them. His findings astonished and dismayed Cattell: there were no significant correlations between the students’ grades and any of the anthropometric tests. If grades and academic standing were indications of intellectual ability, the anthropometric tests were not.22 Furthermore, the tests were so little correlated with one another that it seemed plain they were not measuring a common attribute, as intelligence was presumed to be. Thus by yet another paradox, it was Galton’s discovery, correlation analysis, that invalidated his own method of intelligence testing.
But that was not the end of the story for Cattell or for mental testing. Cattell, undaunted, developed a number of other tests, particularly in the field of value judgments, edited two science magazines, founded the Psychological Corporation to do applied psychology as a business, and became the prime exemplar of the hustling, practical, commercial side of psychology.
Although Galton’s anthropometric approach to mental testing died out rapidly, intelligence testing of a different sort shortly took its place and soon made the study of individual differences the leading area of American psychology. By 1917, well over half of all research reported at meetings of the American Psychological Association dealt with individual differences.23 Galton’s assessment of the value of mental testing came to dominate American psychology, and intelligence testing became the means by which his hereditarian views influenced the schooling offered to students, the assignments given men in the military service, and the immigration policies of the nation.
A final paradox is that none of these results was the intent of the man who developed the intelligence tests that supplanted Galton’s. Alfred Binet’s tests won out over Galton’s; Galton’s views won out over Binet’s.
The Mental Age Approach: Alfred Binet
Alfred Binet, whose name every undergraduate learns in Introductory Psychology, was not a great psychologist; he formulated no important theory, made no brilliant discoveries, and was not a charismatic teacher. But he had one original and relatively simple idea, on the basis of which he and his collaborator, Théodore Simon, fashioned a mental test that has profoundly affected the lives of millions of people.
Binet was born in Nice, France, in 1857; his father was a physician and his mother a woman of some artistic talent.24 His parents separated when he was young and he was raised by his mother; whether due to that circumstance, unusual in his time, or to his being an only child or to a constitutional bent, he grew up an introverted man who made few friends and was most comfortable working and studying alone.
Seeking to find his proper métier, Binet made many false starts. In his student years, he earned a law degree, but then decided that science was more interesting and began the study of medicine. However, having an independent income and not faced with the need to earn a living, he dropped out of medical school to study psychology, to which he had felt drawn for years. He unwisely chose not to pursue psychological training in a formal way but immersed himself in solitary reading in the library (where, among other works, he studied Galton’s Hereditary Genius).
His self-education might have led nowhere, but in 1883 an old classmate, Joseph Babinski (who would later discover the infant reflex that bears his name), introduced him to Charles Féré, a staff member at the Salpêtrière, who in turn introduced him to Jean Martin Charcot, its director. Though Binet had no degree in medicine or psychology, Char-cot was impressed by his intelligence, knowledge, and interest in hypnosis, and offered him a position at the Salpêtrière clinic of neurology and hypnosis.
After some productive years there, Binet took another wrong turn. He and Féré conducted some poorly controlled experiments in hypnosis, imagined that they had discovered a previously unknown phenomenon in hysterical patients, and made public their findings. By the use of magnets, they said, they had been able to shift any action the patient performed under hypnosis, such as lifting an arm, from one side of the body to the other. Even more remarkable, they had been able, also by means of magnets, to transform any of the patient’s emotions or perceptions into the opposite—the fear of snakes, for instance, into a fondness for snakes.
Their report of this hocus-pocus, which would have looked suspect even in Mesmer’s time, brought immediate criticism. Auguste Liébeault and his followers, the Nancy School of hypnotism, said that the effects were produced by suggestion; they proved it by eliciting the very same responses in nonhysterical persons by suggestion alone, without any use of magnets. Binet, who had staked his reputation on the results of the work, had to admit publicly that the results had been brought about by inadvertent experimenter suggestion and were worthless. (Afterward he would often say, “Tell me what you are looking for and I will tell you what you will find,” a succinct statement of what came to be known among psychologists as “experimenter expectancy effects.”)
The shattering experience led to Binet’s resignation from the clinic and his withdrawal from contact with other psychologists. In virtual isolation for about two years, he wrote and produced several plays with themes of terror, murder, and mental illness. Happily, he also spent much time observing the thought processes of his two children, Madeleine and Alice, who were then four and a half and two and a half. To study the nature of thinking at their ages, he devised a number of simple tests: In one he asked them to name the uses of certain everyday objects; in another he asked them to judge which of two piles of coins or beans contained more items; in a third he removed a group of objects from view and then put them back one by one, asking whether any remained unreturned. When the girls were older, he gave them little problems to solve in order to study the growth of reasoning processes. These studies, which he described in three papers, foreshadowed the achievements of Jean Piaget, the developmental psychologist, and were the first step toward the work that would make Binet famous.
Another step in that direction was his return, at thirty-five, to professional life. In 1892, on a train platform, he happened to meet Henri Beaunis, director of the Laboratory of Physiological Psychology at the Sorbonne, and fell into a friendly argument with him about hypnosis. The upshot was that Beaunis invited Binet to become his assistant, and two years later, on Beaunis’s retirement, Binet succeeded him as director. At the laboratory he conducted his own research studies, directed those of many students, and at thirty-seven belatedly earned a doctorate. The degree was in natural science, not psychology, but by this time, thanks to his position and publications, he was a recognized figure in French psychology; and, what with his twirled, pointed mustache, pince-nez, and hair swirled artfully across his forehead in the mode of the god Pan, he looked the part. But his dearest wish, to become a professor of psychology, never came true; to members of the establishment, his notorious work on hypnotism, his unorthodox education, and his wrong kind of doctorate stood against him.
Besides, there was his latest bizarre enthusiasm: an effort to prove that intelligence was directly linked to brain size and could be gauged through “craniometry” (skull measurement). He had read and been convinced by Paul Broca’s (and possibly also by Galton’s) views to this effect. Binet revi
ewed previous craniometric studies, made a number of skull measurements on his own, and between 1898 and 1901 published nine papers on the subject in L’Année Psychologique, a journal that he had founded and of which he was the editor.
Once again he had taken a wrong trail. Early in the series he had said it was “incontestable” that head size was correlated with intelligence,25but later he measured the skulls of a number of schoolchildren identified by their teachers as the most intelligent in their classes and others as the least intelligent, and found that the differences in head size were insignificant. After much remeasuring and reconsideration of his data, he concluded that there were indeed regular but quite small differences in head size, but only between the five brightest and five dullest students in each group. He abandoned craniometry as an approach to the measurement of intelligence.
One could hardly have guessed, at this point, that Binet, in middle age, would shortly produce a work of solid scholarship that would have a considerable effect on the world.
Still interested in the measurement of intelligence, he went back to the method he had used to study his daughters’ thinking. Conceiving of intelligence not as Galton had, in terms of sensory and motor abilities, but as a combination of cognitive abilities, Binet and a co-worker at the laboratory, Victor Henri, began trying out on Parisian schoolchildren a number of tests of those abilities—memory tests (for words, musical notes, colors, and digits), word-association tests, sentence-completion tests, and so on. Their findings suggested that a battery of such tests might measure intelligence if one knew how to weigh the data.
A propitious turn of events spurred Binet to develop this promising lead. Mandatory universal education of children had been instituted in France in 1881, and in 1899 the Free Society for the Psychological Study of the Child, a professional group of which Binet was a member, began urging the Ministry of Public Instruction to do something about retarded children who had to go to school but were unable to cope with standard classroom work. In 1904, the ministry appointed a commission, of which Binet was a member, to study the problem. The commission unanimously recommended that children who had been identified by an examination as retarded should be placed in special classes or schools where they could get education suitable to their condition, but it said nothing about what the examination should consist of.
Binet and his former colleague in craniometry, Théodore Simon, took it on themselves to create such an examination. First they assembled a number of tests, some drawn from studies made earlier at the Salpêtrière, others from the work of Binet and Henri at the Sorbonne laboratory, and still others that they formulated. They then visited some primary schools and tried their tests on children ranging from three to twelve who were considered normal by their teachers, and on others who were considered subnormal. They also tested a number of children institutionalized at the Salpêtrière who were classified as idiots, imbeciles, and débiles. *
After laboriously administering the examination to hundreds of children and omitting or modifying those tests which proved unfeasible, Binet and Simon fashioned what they called a “measuring scale of intelligence.” They described it in L’Année Psychologique in 1905 as “a series of tests of increasing difficulty, starting from the lowest intellectual level that can be observed, and ending with that of normal intelligence. Each group [of tests] in the series corresponds to a different mental level.”26
It was not yet an intelligence test, since it provided no method of scoring the results; it was only a first effort to suggest how one could be constructed. The first of the thirty tests in the battery was extremely simple. The experimenter moved a lighted match back and forth before the eyes of the subject to see whether there existed the coordination of head and eyes associated with vision. The later tests were increasingly difficult, involving such tasks as the ability to judge which of two lines was longer, to repeat three numbers, to repeat a sentence fifteen words long, to draw from memory a design that had been displayed, to say how a folded and refolded paper, out of which a small piece was cut, would look when unfolded, and finally and most difficult, to define abstract terms (“What difference is there between esteem and affection? What difference is there between weariness and sadness?”).27 At each age, normal children could answer questions and accomplish tasks satisfactorily up to a point; the older they were, the farther they could go through the series. The scale was indeed a measuring device of sorts.
While Binet and Simon were testing some children who were identified as normal and others considered retarded, they had a brilliant insight: the retarded children’s intelligence was not of a different kind from that of the normal children; it was simply not as developed as it should have been by their age. By and large, they responded in the same way as normal children younger than themselves. Thus, intelligence could be measured by comparing the performance of a child with the average performance of normal children of the same age. As Binet and Simon put it:
We shall therefore be able to know…if one [child] rises above the average level of other individuals considered normal, or if he remains below. Understanding the normal progress of intellectual development among normals, we shall be able to determine how many years such an individual is advanced or retarded. In a word we shall be able to determine to what degrees of the scale idiocy, imbecility, and débilité correspond.28
Defining intelligence in terms of age and assembling a set of cognitive tasks that measured the mental age of a child replaced Galton’s anthropometric testing and became the foundation of the intelligence-testing movement.
After publishing their study, Binet and Simon took into account shortcomings they had discovered and criticisms offered by others and revised the scale extensively in 1908 and again in 1911. These revisions were supplied with scoring information—a set of standards as to the questions and tasks a child of any age should be able to answer and perform. (If 60 to 90 percent of the children at any given age could pass a particular test, Binet and Simon considered it normal for that age.) Here are some of the items in the 1911 scale:29
Three years:
Points to nose, eyes, and mouth.
Repeats two digits.
Enumerates objects in a picture.
Gives family name.
Repeats a sentence of six syllables.
Six years:
Distinguishes morning and evening.
Defines words by use (e.g., “A fork is to eat with”).
Copies a diamond shape.
Counts thirteen pennies.
Distinguishes drawings of ugly and pretty faces.
Nine years:
Gives change out of twenty sous.
Defines words in a form superior to use (e.g., “A fork is an instrument for eating”).
Recognizes the value of nine pieces of money.
Names the months in order.
Answers easy “comprehension questions” (e.g., Q: “When one has missed the train what must one do?” A: “Wait for another train”).
Twelve years:
Resists suggestion. (The child is shown four pairs of lines of different length and asked which is longer in each case; in the last case the lines are the same length.)
Composes a sentence using three given words.
Names sixty words in three minutes.
Defines three abstract words (charity, justice, goodness).
Makes sense out of a disarranged sentence.
The 1908 scale included tests for age thirteen and the 1911 scale for adults; as later researchers would show, the growth of intelligence continues to early adulthood and then ceases.
The 1908 and 1911 revisions were the first functional tests of intelligence validated against classroom performance and “normed” (provided with scores representing the normal level of response at every age). For the first time, psychologists could determine how far, in years, a child’s mental development was behind normal or ahead of it. Binet and Simon said that if the child’s mental age was two or more years below h
is or her chronological age, the child was likely to require special education. They also defined three levels of retardation in terms of mental ages. The idiot, they said, had a mental age of two or less; the imbecile, between two and seven; and the débile above seven but significantly lower than his or her chronological age.
The weakness in these ratings was that they were fixed mental ages, while nearly all retarded children continue to develop mentally, although at a slower pace than normal. A child of four with a mental age of two is an idiot, but at age eight or ten, though still an idiot, his mental age will probably have reached the four- or five-year-old level.
In 1912 a German psychologist, William Stern, solved the problem by suggesting that if the child’s mental age is divided by its chronological age, the result will be its “mental quotient”30 (soon renamed “intelligence quotient,” or IQ), a ratio that expressed the child’s relative degree of retardation or advancement. The IQ of a child of four with a mental age of two is 50 (the ratio is multiplied by a hundred to get rid of the decimal place), and at ten, with a mental age of five, is still 50. Similarly, a child of five with a mental age of eight, or of ten with a mental age of sixteen, has a genius-level IQ of 160. The IQ is thus a useful way of expressing test results and offers a basis for predicting a child’s potential development.
Although Binet and Simon, in their choice of test material, sought to measure “natural intelligence”—innate ability—rather than rote learning,31 Binet was not a rigid hereditarian like Galton. He explicitly stated that the scale said nothing about the child’s past or future but was an appraisal only of his present state.32 Binet warned that the test results, if interpreted rigidly, might label and condemn to an inferior life some children who, with special help and training, could raise their intelligence level, and in a late work he cited with pride the increases in intelligence that had been obtained in special classes for subnormal children in an experimental school that he had founded.33