Why Johnny and Joanie Can't Read

With 20-20 hindsight, the key to unlocking the puzzle of dyslexia was as plain as an exasperated teacher's order to "sound it out!" But for decades neuroscientists searching for clues to why 5 to 10 percent of the population has such trouble learning to read have studied how the eyes see and how the brain processes language. They haven't looked at how people hear. Now they'll have to. Last week a paper in The Proceedings of the National Academy of Sciences reported the first evidence that dyslexia arises from abnormalities in a part of the brain that processes sounds. Dyslexic children, it seems, cannot sound out many words for the simple reason that they have never clearly heard what sounds certain letters make. "In dyslexia, the brain is miswired for language and thus reading," says Glenn Rosen of Boston's Beth Israel Hospital. "With this discovery, we are on the cusp of understanding this miswiring."

Dyslexia means difficulty reading for any reason, despite normal intelligence and education, and not just because you reverse letters -- seeing "b" as "d" and "saw" as "was." The problem seems to lie in a part of the brain that acts as an auditory relay station (diagram). This medial geniculate nucleus takes incoming sound signals from the ear, encodes them in some unknown way and sends them to the cortex, which makes sense of them. Rosen, and Albert Galaburda and Matthew Menard of Harvard Medical School, found that in dyslexics, the language-processing left side of this relay station had fewer of the neurons that process fast, staccato sounds -- such as ba, da, ka and ta -- than did the brains of normal readers. These so-called stop consonants last a mere .04 second, rather than the .1 second of a vowel such as "aaaahh."

If a child cannot properly hear stop consonants, he cannot construct a mental dictionary that keeps track of what letters sound like. "Each letter has to be connected up to this auditory template in the brain," says neuroscientist Paula Tallal of Rutgers University in Newark, N.J. In the early 1970s, Tallal discovered that children who later had trouble learning to read first had problems distinguishing sounds that last a short period of time, as stop consonants do. Although proficient readers recognize words by sight, beginners must sound them out. "Reading is a real problem if you can't hear the differences between sounds and set up the template," says Tallal, who estimates that 80 percent of dyslexics have this anomaly. (Children with chronic middle-ear infections also have trouble with language and reading.) But until now, there was no neurological evidence to support her auditory theory of dyslexia.

The discovery that dyslexics have fewer of the neurons that process fast sounds is based on the brains of only five dyslexics and seven control subjects. But if it holds up, it offers clear lessons to educators. "By understanding the biological facets underlying dyslexia," says Rosen, "perhaps we can bypass the standard ways of instructing the brain to read." One alternative is to trace letters on a child's skin, rather than pronounce them. Tallal is taking a different approach. With a new, $2.3 million grant from the Charles A. Dana Foundation, she is heading a consortium trying to develop computers that pronounce the stop consonants slowly, enabling dyslexic children to hear them clearly and thus form the mental templates needed to sound out words. Neuroscientists have a more basic goal: figuring out what causes the brain abnormalities of dyslexia in the first place.