For several decades, unstandardized IQ scores in about 20 industrialized nations have been on the rise, presenting scientists with a serious riddle: the increase is far too fast to come from genetic changes and yet intelligence is presumed to be highly heritable.

The upward trend is known as the Flynn effect, for the scientist who first noticed it in the 1980s. James R. Flynn's finding gained added popularity in 1994, when it was widely used to counter ideas put forth in Richard J. Herrnstein and Charles Murray's controversial book, The Bell Curve. Now Flynn, a professor of political science at the University of Otago in New Zealand, and William T. Dickens of the Brookings Institution, claim to have cracked the paradox.

The duo developed a new mathematical model, described in this month's Psychological Review, that measures multiple influences on IQ and the interplay among them. Their conclusion, in short, is that environmental factors can have a significant impact on intelligence¿big enough, in fact, to explain the Flynn effect. "Heritability isn't a very useful number for thinking about the direct impact of genes on IQ," Dickens says. "Changes in environment can produce changes in IQ that are several times as large when both are measured relative to their variation in the population."

The model is consistent with a number of experimental findings¿such as the fact that adopted and nonadopted siblings have highly correlated IQs while they share an environment but move apart with age toward the intelligence of their biological parents. So, too, it helps account for the observation that childhood enrichment programs appear to improve IQ scores only while children are in them. To test the model, Dickens and Flynn propose testing IQ scores before and after radical environment changes, such as time in jail. Any volunteers