The Chaotic Word Blogging with a Hammer


In this case, Steven Pinker is right

Dr. Steven Pinker, author of How the Mind Works, among other books, and Malcolm Gladwell, author of Blink, are currently engaged in a debate that, like human insanity and unnecessary misery, never goes away. Gladwell is accusing Pinker of inhabiting the "lonely ice floe of IQ fundamentalism." Many intellectuals, particularly those of the unrepentantly optimistic kind, do not like IQ research, and it's no wonder.  If liberals (and conservatives) took the science behind IQ research seriously, they would have to radically change their worldviews. But should it be taken seriously?

To answer this question, you might well first ask, what is IQ? It's actually rather simple to understand. Imagine that you had a very large library of questions and puzzles, tens of thousands of items. Then imagine that you randomly chose sets of 50 questions from that library. Then imagine that you gave those 50 questions to a hundred people, scored them right or wrong, and obtained a sum of correct answers for each person. Finally, imagine that you rank-ordered their performance. Voila! Those people are now ranked by IQ. It turns out that no matter which 50 questions you used, the rank order would be extremely similar.

Despite the rather simple and obvious manner in which IQ can be derived (and explained), there is probably more nonsense written about it than about any other scientific concept. The most egregious purveyors of such nonsense include Robert Sternberg, Howard Gardner (the twin darlings of educational theorists devoted to wishful thinking) and, most unforgiveably, the Harvard paleontologist, Stephen J. Gould who, among his many other sins, felt it necessary to torture Edmund O. Wilson for his biologically influenced views of human behavior.  Gould regarded IQ as a reification, and its method of derivation as the inappropriate "abstraction of intelligence as a single entity."

It is almost impossible to overstate how intellectually reprehensible Gould's stance really was. IQ is perfectly well measured as a simple average. Its derivation, as explained above, does not have to be mathematically or conceptually complex. Gould complained about the complex "factor analytic" techniques often used to produce estimates of intelligence, but did not or refused to realize that the simple average is a form of factor analysis. So Gould basically argued that the average ("reified" and "abstracted") does not exist, when he knew perfectly well that mathematical abstractions are extremely real, by any reasonable standard,  and that the average is a useful and unavoidable scientific measure.

This simple average, which can be measured reliably and validly (the twin markers for a well designed test), appears to index the human abilities to abstract, and to manipulate abstractions. It turns out that in a complex society, those abilities are strongly linked to success, academic and practical. When I say strongly, I mean strongly: First, IQ predicts such things better than any other measure, with the possible exception of conscientiousness (which is the proclivity to truth, hard work, and orderliness, and is a scientific mystery); second, the strength of this relationship exceeds the strength of 95% of the relationships identified by social scientists, according to James Hemphill, a psychologist at Simon Fraser.

The association between IQ and success is not surprising, because most jobs that are not merely rote repetition of the same act require abstraction, including the use of language. There is a positively lovely paper by Linda Gottfredson, who also happens to be a lovely person, explaining this in clear language. IQ also predicts other strange things, like quality of human semen. It is important to note that it does not explain everything, however (and this is also the mark of a good measure, and is called divergent validity). Conscientiousness, a trait which includes orderliness, persistence and integrity, is also a very strong predictor of success, broadly defined, as is emotional stability (relative freedom from anger, fear and emotional pain). These traits happen to completely and somewhat surprisingly uncorrelated with IQ, and conscientiousness is not understood at all.

Here's another reason it is hard to simply dismiss the reality and the importance of IQ (unless you want to dismiss the entire corpus of work in social science, which is of course your prerogative). The psychologists who laid the groundwork for IQ research, Spearman and Pearson, were also instrumental in establishing the statistics that psychologists (and, to an extent only limited by their expertise, other social scientists) use almost exclusively to test their claims to truth. What that means is that if you believe anything a social scientist has ever claimed to demonstrate scientifically, rather than merely opining about, you are pretty much stuck believing in IQ.

Sternberg and Gardner claim, for their parts, that there are many "kinds" of intelligence. That is fine, but we already have a word for other kinds of "intelligence": talents. Confusing talents and intelligences is a mug's game, philosophically dishonest, and dangerously misleading. It is very important to be precise about terminology, to avoid confusing the issue. By his own admission, Gardner doesn't care about measurement (which eliminates him from any serious scientific game), while Sternberg plays games with his statistics and his methodology to ensure that the results of his own research never challenge his a priori beliefs. Neither of them has ever demonstrated that their “intelligences” can be measured. From a scientific perspective, this means that they simply do not exist.

The mere fact that IQ exists, however – at least in the pragmatic sense, if not in the ontological – poses a number of problems. One of these is that heritable factors strongly determine the differences in IQ between people. This becomes, paradoxically, more true as the environment equalizes. As more and more people have access to computers, books and even television, and their environmental opportunities for full cognitive development become more similar, the only factors left to determine variation in IQ are genetic. The relationship between environment and genes is not fixed.

This is a big problem for liberals, because there are real differences in human intelligence. This means that much of the inequality that exists does not exist for environmental reasons. This means, for example, that if education systems are improved, the smarter people will benefit more. This means that if there is more equality of opportunity, in a complex, abstracted society, the smarter people will do better. Thus, as we become increasingly meritocratic, biological differences increasingly account for the success or failure of people. This is the non-egalitarian and paradoxical consequence of equality of opportunity.

IQ differences are also a big problem for conservatives. Conservatives like to delude themselves into believing that there is a job for ever person, if only that person is sufficiently willing to work. As someone who once spent 35 hours training a dedicated and conscientious person with a nonverbal IQ of about 80 unsuccessfully to fold letters to place in envelopes, I can attest practically as well as intellectually to the fact that some people are just not smart enough to work productively. The US Army has learned this through hard experience. It is illegal in the US to draft anyone with an IQ of less than 82, if I remember properly. This is about 10% of the population. No one whose intelligence is below that can be trained to do anything productive and worthwhile for an organization chronically and desperately short of personnel, and designed in part as an agent of social mobility. What this means in part is that the problem of welfare dependency and unemployment and low income and social inequality IS a problem of the unequal distribution of biological resources. It would be good to have an intelligent discussion about the moral implications of that problem, instead of wishing it away, from the right and the left.

I should say, in general, that I am an admirer of Gladwell, although less so of Pinker, even though I believe that Pinker has the science on his side on this issue. Gladwell is a very good writer, and an intelligent conduit of information to the broad populace. However, he doesn't have the technical knowledge to hold his own in the IQ debate. As for Pinker, his comments on culture, art and religion (which basically boil down to the presumption that they are spandrels, byproducts of evolution) are both vapid, from the perspective of content, and treacherous, from the standpoint of fair argumentation.

They are vapid because Pinker believes that the role of art, culture and religion in human evolutionary and social development can be summarized in a single chapter in a single book: dismissed, essentially. He doesn’t take these phenomena seriously, because he thinks they are spandrels, and he thinks they are spandrels, because he doesn’t want to do the work that would be required to broaden his knowledge.

Pinker’s comments are treacherous because no one has been able to define precisely and exactly when a biological or cultural phenomenon is “genuine” (whatever that might mean, evolutionarily) or a “byproduct” of something genuine). This allows an evolutionary psychologist to define anything he or she doesn't want to take seriously as a spandrel, and to thereby dispense with it.

This has happened to me many times when debating the utility of religious belief (even its inevitability) with rationally-minded thinkers. I point out that religious experience and belief are human universals, that these universals have existed as far back in time as we have records of human activity, and that they may thereby (1) have evolved and (2) serve some useful biological function. They respond “spandrel,” despite the centrality of art, culture and religion to human behavior, and remove all necessity for further thought.

Filed under: IQ No Comments
Get Adobe Flash player