Does IQ measure cognitive capacity?

Does IQ measure cognitive capacity?

Here is a very non-controversial way to frame the IQ Debate: at any given moment, some people have better cognitive functioning than others.  That much is obvious, isn’t it?  Some people can spell better, some people can do more complicated math problems, some people have more concrete information about the world.  It would take pretty rosy glasses to deny this.  If that is all the IQ Debate it is about, it is hard to see why people get so upset about it or even why everyone is so interested in it.  In this weak version of intelligence, IQ is a description of the fact that some people can think better than others; it is not an explanation of why this is so.

But it seems to me that at least some of the time, when we talk about intelligence we mean something more than that.  When we say someone has low intelligence, we mean not only that the person’s observable ability to get correct answers is low at the moment, but that they have a diminished capacity for cognitive functioning.  We mean not only that they are not doing so well now, but also that they don’t have the potential for doing better.  Same thing at the opposite end:  if we say that a great mathematician is very intelligent, we don’t just mean that she happens to be doing well right now, we mean that she has more capacity than most of us.  It’s not only that you or I can’t do that math ourselves right now, we can’t imagine being able to do it:  we aren’t intelligent enough.  When you think this way, the concept of intelligence is being used as an explanation, not simply as a description.

I want to think about whether the weak or strong interpretation of intelligence is correct, or when we are prone to adopt one over the other.  I should say at the outset that I don’t see easy answers here, and in particular that I am not trying to score cheap points against the concept of intelligence.  Sometimes, my intuitions suggest a capacity interpretation of intelligence is correct.  If someone with a frank brain injury has a low IQ score, I take it as a description not only of what they actually do, but what they can do.  Even more, like everyone else I look around me and see people who seem more or less intelligent, and to some extent some of the time I treat that intelligence as a capacity, not just as a description. Nevertheless, figuring out what those intuitions mean is far from simple.  If you look through definitions of intelligence in books, you find the word “capacity” all the time: Intelligence represents a difference in people’s cognitive capacity.  But I have never seen anyone worry about exactly what the word capacity means in that sentence.  This post is basically a research proposal for thinking about the problem.

As an example of what I mean, consider the concept of health. Suppose you got a population of people and obtained a bunch of health measures: their blood pressure, their cholesterol level, lung function, you get the idea. You take the average of all of these (never mind for the moment how correlated they would be with each other) and call it their HQ. Someone has probably done something very much like this. I bet HQ would be a good predictor of all sorts of things, mortality and medical expenditures and the like. There is no doubt (see below) that it would be heritable. But I don’t think we would be prone to think about HQ the same way we (sometimes) think about IQ, as some sort of internal quality that explains why some people are healthier than others. We might observe that one group of people has poorer HQ on average than another, but I think it would be difficult for even the most racially oriented person to think that this was somehow an inherent biological quality of the group. People wouldn’t be digging around in the brain for health centers, or in the genome for health genes. HQ would just be a description, a useful one, of the current contingent state of affairs.

Here are some of the issues that might come up in an investigation of IQ as a capacity:

 

  • Maybe it doesn’t mean anything at all. Maybe people just use the idea of cognitive capacity as another way of saying cognitive functioning, without meaning anything in particular about it. If so, fine, but I think that leaves the whole intelligence field way less interesting and important.

 

  • One intutition that seems to be held by a lot of intelligence people is that it is the positive correlations among mental tests that seals the deal on intelligence as a crucial construct. I have always been a g-skeptic, because a) Positive correlations among tests aren’t the same thing as tests being unidimensional, and in fact they are not; b) Tests that are highly but less than perfectly correlated can be represented as g plus special abilities, or as a bunch of correlated abilities without g, and the data don’t give strong reasons to choose one over the other; c) Even if you accept g more or less at face value, environmental as well as biological processes could produce it; and d) even if tests were completely independent, you could still compute an average score, and people would differ on it. In some ways, all the positive correlation among the tests does is to change the variance of their average. But anyway the implications of g would have to be worked out.

 

  • Essentialism and reification. The idea of g as a reification of intelligence goes back to Stephen J. Gould, the bête noire of the intelligence world, with whom I have some sympathy. Essentialism is a modern idea, and I think you could re-state what I am asking here by wondering about whether essentialist accounts of g add anything to the picture.

 

  • Reaction norms and ranges. Reaction norms are a concept introduced by Dobzhansky to illustrate the abstract relationship between genotype and environment in the genesis of behavior. A reaction norm is a bivariate surface in which phenotype can be predicted from genotype and environment. It helps you think about what it might mean for someone to have a biological potential that is realized in a particular environment, but it is a very complicated problem.

 

  • One way to speak about intelligence-as-capacity is that some people are innately more intelligent than others. Unlike some of the other topics here, there is a big philosophical literature on the concept of innateness.

 

  • Nature-nurture questions. Certainly one of the things that contributes to a capacity view of intelligence is that IQ is heritable. Children adopted away from their biological parents at birth have IQs correlated .3 with their biological parents. What could that be other than genetic transmission of capacity? But you have to remember that this would also be true for HQ, and in fact it is true for pretty much everything you can think of. This is the topic I have written the most about over the years.

 

  • One reason intelligence sometimes seems like a fixed capacity is that it is very hard to change. There is a long history of using Skodak and Skeels type adoption designs, randomized trials of education programs, and adoption studies from Scandinavian population registers to try and figure out how much IQ can be changed by the environment. It is no accident that Arthur Jensen’s signature work was titled, “How much could we boost IQ and educational achievement?”

 

  • Arthur Jensen. Whatever one thinks of his conclusions, and I disagree with most of them, there is no question that Arthur Jensen was the most important thinker about intelligence since Galton. Many of the themes of his work—malleability as above, the search for low-level cognitive explanations of IQ, genetics, group differences, brain differences—can be seen as an extended attempt to demonstrate that intelligence is a capacity, a biological capacity, that goes beyond a mere description of observable differences.

 

One final take, putting the intelligence question in a broader context: what is talent? If I say that Lena is a talented oboe player, does that have any surplus meaning over and above saying that she is right now a good oboe player? I might mean that I believe that if Lena worked long and hard at the oboe she would be an exceptional player, whereas someone else who is currently at her level would not; I might mean that Lena seems to learn the oboe more easily than other people; I might mean that it any imaginable environment, Lena would become better at oboe than other people with the same exposure; we might mean that there is something biological about Lena (long fingers, perfect pitch) that makes her good at the oboe. We talk about talent all the time, and it is remarkable to me that I can’t say what talent means. Does anyone know of a philosophical treatment of the concept of talent?

 

In general, I wrote this post to generate comments, ideas and things to read. Please make suggestions.

 

 

Eric Turkheimer
ent3c@virginia.edu

Eric Turkheimer is the Project Leader for the Genetics and Human Agency Project. Eric is a clinical psychologist and behavioral geneticist. For thirty years he has been involved in empirical and theoretical investigations of the implications of genetics for the genesis of complex human behavior. Current projects include understanding the interaction between socioeconomic status and the heritability of intelligence, and philosophical analysis of the ethical status of work that purports to demonstrate biologically based differences in behavior among racial groups.

1Comment
  • Elijah Armstrong
    Posted at 01:51h, 24 September Reply

    One interpretation of capacity is “rate of learning”. If so, IQ measures capacity in some respects (for instance, efficiency of job training) –– though there are probably some exceptions. I think the type of “learning” that is most intuitive in this context is the learning of highly important cognitive schemata and skills, like the ability to form analogies, to think deductively, to have assimilated a body of knowledge that can be reapplied, etc; here, too, there are exceptions. For instance, it may be that increases in fluid intelligence as a function of schooling, as measured on an absolute scale or using raw scores, are not greater for people with initially higher IQs. (Jim Flynn’s new book has some interesting data which touches on a similar question.)

Post A Comment