As Gardner's idea began to take hold, teachers embraced the idea that they could use it to identify the particular kind of intelligence each child had. Teachers felt that Gardner's scheme helped them fine-tune their curriculum to fit the particular kind of intelligence each child possessed. As the theory got diluted within schools, some teachers simply used the idea of multiple intelligences as a way to help each child feel smart, even when he or she didn't excel at traditional school tasks. This egalitarian conception of intelligence has taken hold like wildfire, and in every town in America, you can hear teachers talk about the specific kinds of intelligence their students possess. "He might not be good at math, but he sure is smart when he's on the basketball court." "She struggles with English class, but she's so artistic; her visual intelligence is outstanding." Or the most common form: "He might be book-smart, but he's street-dumb." The theory of multiple intelligences makes people happy. But does it explain the data?
The Smartness Thermometer
Ever since psychologists began formally measuring things, they've been trying to measure intelligence. Until Howard Gardner introduced multiple intelligences into the common lexicon, people tended to use the term IQ as a stand-in for intelligence. The intelligence quotient is a mathematical expression, devised in France at the turn of the twentieth century by Alfred Binet. He developed the test to help the French school system identify children who were retarded or significantly slower than others in their age group. His goal was to make sure that "slower" children were not punished for their inability to learn. The original IQ (as well as almost all subsequent forms of it) was based on a very simple concept in psychological assessment: Ask children of a certain age a series of questions. The number of questions they get correct is then divided by their chronological age. Thus, although a given ten-year-old might answer more questions than a seven-year-old, the younger child might well have a higher intelligence quotient. Using the test to compare children depends on a much-trusted practice among researchers: norming. This means that in order to evaluate a given child's IQ, you have to compare it to the average (mean) score of children that age. As a result, each child of a given age is being compared with what is considered normal or typical of all the other children in that age group.