The Intelligence Quotient

Alfred Binet designed his test in 1905 to identify children in French schools who were struggling with the curriculum and might benefit from additional support.

The test was practical in its intent and bounded in its application—it was a tool for a specific purpose, calibrated for a specific population, designed to produce information useful in a specific institutional context.

Binet himself was cautious about what the test measured and what it did not. He did not believe it captured a fixed or general capacity. He believed it captured current performance on certain tasks, which was relevant to educational placement and not much else.

The test left Binet’s hands and became something he did not design it to be.

The Intelligence Quotient—the ratio of mental age to chronological age, multiplied by one hundred—is a formula that is mathematically coherent for children and mathematically meaningless for adults. Mental age, as a concept, describes how a child’s performance on the test compares to the average performance of children at different ages. A ten-year-old performing like a twelve-year-old has a mental age of twelve. The formula produces a number. The number is legible. For children whose cognitive development is genuinely the variable being assessed, the number carries some information.

For adults, the brain has reached its developmental endpoint, at least in the sense that the developmental trajectory the formula was tracking has concluded. There is no adult mental age in the original sense. The modern solution to this is deviation IQ—comparing the adult’s score not to a mental age ratio but to the distribution of scores in their own age cohort. The adult is measured against other adults of similar age and placed within the distribution. This produces a number that describes relative position within a population.

It does not measure intelligence. It measures relative position on a set of tasks within a population. These are not the same thing, and the conflation of them is where the construct begins to generate the confusion it has generated for over a century.

 

The tasks that IQ tests measure are a selection from the larger space of human cognitive capacity, chosen partly because they produce reliable and consistent scores, partly because they correlate with outcomes the test designers valued, and partly because they were tractable to the paper-and-pencil format the tests required. Logical-mathematical reasoning. Pattern recognition. Verbal comprehension. Spatial reasoning. Working memory. These are real cognitive capacities. They are not the totality of cognitive capacity, and the fact that they are measurable with reasonable reliability does not mean that the unmeasured capacities are less real or less significant.

Emotional intelligence—the capacity to read social situations, to understand what others are experiencing, to navigate the complexity of human relationships with accuracy and flexibility—is not on the test. It is also not trivially distributed across the population. The person who is poor at logical sequencing but exceptional at reading people and managing the social dynamics of a group is performing a cognitively demanding task that the IQ test does not measure. The test records them as having a lower score than the person who is excellent at logical sequencing and poor at reading people. The test has not measured their intelligence. It has measured the subset of cognitive capacity that the test was designed to measure.

Creativity is not on the test. The IQ test typically has one correct answer to each question. Creativity, in the sense of generating novel solutions to problems that do not have predetermined correct answers, is a different cognitive operation from the one the test is measuring. The research on the relationship between IQ and creativity shows a modest correlation at lower ranges and a very weak relationship at higher ranges—meaning that above a certain threshold of logical reasoning ability, IQ tells you almost nothing about creative capacity. The high-IQ person and the person scoring somewhat lower may be equally creative or very differently creative, and the test will not distinguish between them on this dimension.

Practical intelligence—the capacity to navigate the actual problems of actual life in actual contexts—is not on the test. Robert Sternberg spent decades documenting the gap between academic intelligence, which IQ tests approximate, and practical intelligence, which predicts performance in real-world contexts with a different pattern of correlations. The person who solves abstract logical puzzles efficiently is not necessarily the person who manages complex interpersonal situations, or who navigates institutional systems effectively, or who makes good decisions under conditions of genuine uncertainty. The test measures the puzzle performance. Life is not primarily composed of puzzles with one correct answer.

 

The cultural bias embedded in standard IQ tests is the dimension that has received the most sustained critical attention, partly because its consequences are the most socially visible. Tests developed in Western academic contexts, calibrated against populations with access to formal education in the Western tradition, measure performance on tasks that are more familiar and more practised by people who have been educated in that tradition. The person who has spent years in schools that teach the specific cognitive habits the test rewards—the linear sequencing, the abstract categorisation, the decontextualised problem-solving—will perform differently from the person whose cognitive development occurred in a different educational or cultural context, regardless of the underlying capacity the test claims to measure.

This is not a peripheral criticism. It is a fundamental one. If the test is measuring exposure to specific cognitive training rather than underlying cognitive capacity, then the scores it produces are scores of educational experience, not intelligence. The two are related but not equivalent. The relationship between them is further complicated by the fact that the cognitive training itself alters the cognitive capacity, because the brain is not a fixed processor but a system that reorganises in response to the demands placed on it. The person who has practised abstract reasoning extensively will perform abstract reasoning differently from the person who has not. Whether the performance difference reflects intelligence or practice is not a question the test can answer, because the test does not control for practice.

 

The relationship between IQ and the neurodivergent framework is one of the places where the circular logic the framework depends on becomes most visible. If IQ is a valid measure of cognitive capacity, then giftedness—the upper tail of the IQ distribution—is a meaningful category that identifies people with genuinely exceptional cognitive capacity. If IQ is not a valid measure, or is a measure of something narrower than cognitive capacity, then giftedness is a category defined by performance on a specific set of tasks, which may or may not correspond to the kind of exceptionality the term implies.

The gifted child who scores in the top two percent on an IQ test has performed in the top two percent on the specific tasks the IQ test measures. This may mean they have exceptional logical-mathematical and verbal reasoning capacity. It does not necessarily mean they have exceptional capacity across the full range of cognitive functions. It may mean they have had exceptional exposure to the cognitive training that the test rewards. It may mean a combination of both. The test cannot distinguish between these possibilities, and the giftedness label that follows from the score does not acknowledge the distinction.

The person who thinks faster or more deeply than the people around them—who makes connections that others do not make, who processes complexity at a speed or depth that creates a gap between their experience and the experience of the people they interact with—may or may not score in the gifted range on an IQ test, depending on which dimensions of their thinking the test happens to measure. The experience of the gap is real regardless of the score. The social consequences of the gap—the difficulty of finding interlocutors who engage at the same pace or depth, the experience of conversations that feel truncated, the isolation that comes from operating at a frequency that the immediate social environment does not share—are real regardless of whether the underlying cause is captured by a percentile.

 

The neurotypical standard against which neurodivergence is measured depends, in part, on IQ and similar psychometric instruments for its definition. The typical range of cognitive development is established partly by the distribution of scores on tests like IQ. If the tests are measuring something narrower than intelligence, and if the typical range they establish reflects performance on that narrow measure rather than the full range of cognitive variation, then the typical range is a narrower target than the concept implies, and the divergence from it is a divergence from the narrow target rather than from some broader standard of typical cognitive function.

This does not dissolve the experience of being different. The person who processes information differently from the people around them, whose cognitive style creates friction with the environments designed for a different cognitive style, whose experience of social interaction is shaped by a different relationship to the cues and conventions that the environment takes for granted—this person’s experience is real and its consequences are real. The question of whether the experience is captured by the available labels and measurements is separate from the question of whether the experience exists.

The labels and measurements were designed for Norm and Norma’s children—the children whose cognitive development follows the average trajectory at the average pace in the average environment. The children who do not are being assessed against instruments calibrated for children who do. The instruments produce scores. The scores produce categories. The categories produce labels. The labels become identities.

What the instruments were measuring in the first place remains, after more than a century of use, genuinely contested.

The quotient was a ratio of one number to another.

The numbers were averages.

The averages were populations.

The populations were not individuals.

The individual carries the label.

The label carries a number.

The number was someone else’s average.