What is intelligence? Can we measure it? Do some have more of it than others?

I have just started to reread Stephen Gould’s “The Mismeasure of Man” If you have not read it—it is a must read, especially for anyone who calls themselves an educator.

He starts with two main points—or fallacies. One is the fallacy that intelligence is a thing at all. Rather, it is a construct, an idea. Intelligence is actually no more or less than we define it as.  The other fallacy he points out is ranking—as though there is some linear range, like height or weight on which to line up intelligence.

Our ideas of intelligence are socially and culturally created as well as historically situated, as Vyogtsky pointed out almost a century ago. Intelligence is only what we define it as. Our ideas of what it is are firmly entrenched in our belief systems, in our cultural paradigms. And also due to this any test of intelligence is to some degree a tautology. How do we prove someone is intelligent? Their score on the IQ test. How do we know that the IQ test is valid? We designed it so that those we “knew” were most intelligent got the highest scores and those we “knew” were less intelligence got the low scores. This is as true today of IQ tests as it was of the previous methods of measuring intelligence (craniology for instance). New versions of intelligence tests and even other forms of standardized testing are assessed on whether the same group that did well on the previous version do well on the new versions, and the same for those who did poorly—the curve needs to stay the same. If a different group does better on new test items (which are beta-tested first) those items are discarded as invalid (unless of course the test designers decide they want a different group to do better or worse).

The uni-dimensiality of intelligence has currently fallen into controversy, but it is as unprovable as intelligence itself—it is nether true nor untrue—since “intelligence” is what we define it as, we can choose to define it either way, and to categorize the different dimensions as we find useful.

The same is true of another assumption of intelligence—that it falls along a “normal curve.” This is a logical assumption based on other natural traits, such as height and weight. But we should not lose fact that it is another unprovable assumption, not a fact (and actually presupposes there is a thing called intelligence to measure and put on such a curve).

Because intelligence is a cultural construct, any test of it will be therefore biased toward those who share the knowledge, assumptions, world views and paradigms of the dominant culture. This again is unavoidable. A test has to have content, and any content exists in some context.

Because of these attributes of intelligence, I find the use of any measurement of intelligence highly suspect. When used to sort people in any official way, it is dangerous to a democratic society.