At the beginning of the 20th century the British psychologist Charles Spearman “discovered” the idea of general intelligence. Spearman observed that students’ grades in different subjects, and their scores on various tests, were all positively correlated. He then showed that this pattern could be explained mathematically by assuming that people vary in special abilities for the different tests as well as a single general ability—or “g”—that is used for all of them.
John Duncan, one of the world’s leading cognitive neuroscientists, explains Spearman’s work early in “How Intelligence Happens,” before moving on to his own attempts to locate the source of Spearman’s “g” in the brain. To get us grounded, Mr. Duncan also provides a wonderfully compact summary of brain architecture and function. Throughout the book, he makes it clear that his fascination with intelligent behavior has to do with how the brain brings it about—he leaves it to others to ponder things like the economic import of intelligence and how it is influenced by genes, upbringing, and education.
He also doesn’t waste time dilating on the question of what, precisely, we mean by “intelligence.” Defining terms is not the expertise of scientists, but their attempts can be thought-provoking. Two decades ago, the cognitive science and artificial-intelligence pioneer Allen Newell proposed that an entity should be considered intelligent to the extent that it uses all the information it has when making decisions. But according to that definition, a device as simple as a thermostat would have perfect intelligence—not terribly helpful when trying to understand human differences.
I have been doing research on intelligence for more than a decade, and I have to confess that I do not know of a perfect definition. But most psychologists consider intelligence a general ability to perform well on a wide variety of mental tasks and challenges. In everyday speech, it sometimes means roughly the same thing: We call someone “intelligent” if we believe that their mental abilities are generally high—not if they are skilled in just one narrow field.
Mr. Duncan’s early work on intelligence and the brain resolved an old paradox. Before imaging technologies like MRI were invented, neuropsychologists used IQ tests to determine what parts of the brain were damaged in patients suffering from strokes and other closed-head injuries. If the patient had trouble with the verbal parts of the test, the damage was probably in the left hemisphere; if the trouble was in the visual parts, the damage was probably in the back of the brain; and so on. But oddly, damage to the frontal lobes seemed to have very little effect on IQ—despite the frontal lobes’ constituting nearly 40% of the cerebral cortex.
Mr. Duncan found that patients with frontal-lobe damage were impaired on tests of “fluid intelligence” that, until recently, were not part of standard IQ tests. These tests measure the ability to solve abstract nonverbal problems in which prior knowledge of language or facts is of no help. For example, a “matrix reasoning” problem presents a grid of complex shapes with one empty space that the test-taker must fill by choosing the correct option from a set of up to eight alternatives. Such tests seem to reveal a raw ability to make optimal use of the information contained within a problem or situation.
Later, Mr. Duncan used PET scanning to measure the brain activity of people without brain damage as they solved problems that varied in difficulty. Regardless of content, as the tests got harder, the subjects made more use of areas in their frontal lobes, as well as in their parietal lobes, which are farther toward the back of the brain.
Mr. Duncan makes a convincing case that these brain areas constitute a special circuit that is crucial for both Spearman’s “g” and for intelligent behavior more generally. But his book elides the question of whether this circuit is also the source of IQ differences. That is, do people who score high on IQ tests use the frontal and parietal areas of their brains differently from people who score lower? The answer, discovered by other researchers, turns out to be yes.
There are other properties of the brain that contribute to “g,” including the speed of basic information-processing (measured by how fast people can press buttons in response to flashing lights) and even the total size of the brain (larger is better). One of the next steps in understanding “g” is to figure out how all these factors interact and combine to produce the wide range of differences we see in human intelligence. Mr. Duncan no doubt will be a key player in this effort, frontal and parietal lobes firing way.
Mr. Chabris is a psychology professor at Union College and the co-author, with Daniel Simons, of “The Invisible Gorilla: And Other Ways Our Intuitions Deceive Us” (Crown).
Full article and photo: http://online.wsj.com/article/SB10001424052702304410504575560153250663606.html