Because dyslexics frequently confuse "b" and "d," they originally were thought to have vision problems. Scientists now believe that dyslexics confuse the two letters because they sound alike.
Dyslexia has been the focus of the MRI studies at Yale and four other sites. The Shaywitzes, co-directors of the Yale Center for -- Learning and Attention, wanted to see what was - or wasn't - happening within the brains of dyslexic children.
A functional MRI machine, a type that requires no injection of dye and is safe and painless, takes thousands of pictures of the brain, at rest and as it processes information. When brain cells go to work, oxygen-rich blood rushes to fuel their activity, much as gasoline fills the engine of a running car. Because this oxygenated blood has different magnetic properties, MRIs can discern it.
The MRI pictures are then color-coded by a computer. In the Yale study, for example, resting brain cells are blue. Blood rushing glucose and oxygen to brain cells at work is yellow and red.
The brains of people who have difficulty reading show a more diffuse blood flow, Lyon says.
By asking people to perform different tasks - identify single letters, pick out rhyming words or signal whether two words belong to the same category - researchers have been able to pinpoint the discrete sections of the brain used in reading.
One part of the brain, the extrastriate cortex, identifies letters. Another part, the inferior frontal gyrus, identifies the sounds associated with those letters. And a third section, the superior temporal gyrus, reaches for meaning.
Yale researchers see this difference explicitly when they show research subjects made-up words, such as "joat" and "mote," and ask them to signal whether they rhyme. To do so, they have to be able to sound out the words.
People who can't sound out words appear to have lighter blood flow to the language regions of their brains, Lyon says.
"We apparently don't get as much flow and volume in those regions of the brain" in dyslexics, Lyon says. "But it's very important to point out we don't know exactly why. All we know is there are brain differences in people who do not read well if
In February 1995, the Shaywitzes announced a major finding. They had identified and mapped the sections of the brain that process language. In so doing, they had discovered that men generally use only half of their brains for these tasks while women use both lobes.
As headlines around the globe trumpeted the "Battle of the Brains," the importance of the Shaywitzes' discovery about the brain's language processing was all but obscured by the popular appeal of the gender differences.
Much quieter but perhaps more significant has been their research into the ways that the brains of dyslexics differ from those of normal readers. The Shaywitzes are also beginning a trial with Syracuse University researchers to determine whether intensive training can improve the way dyslexics process information.
The Shaywitzes are reluctant to discuss unpublished findings, and they caution that it is still far too early to draw conclusions. But they say they hope one day to be able to demonstrate physical differences in the brain of a dyslexic.
"One reason we are so excited is that we do think we will be able to produce concrete evidence of this person's disability," says Sally Shaywitz. "It would be like an X-ray of a person's broken arm, something you could actually see. Reading disabilities for so long have been a hidden disability."
Jack Pikulski, a University of Delaware reading professor and president of the International Reading Association based in Newark, Del. - an organization that has promoted the whole-language approach - says the NIH research offers valuable insights. But he cautions against drawing conclusions from it about the general population and reading.
"Be careful. I They concentrated on the 20 to 25 percent having difficulty," Pikulski says of the studies. "We shouldn't generalize about what is happening with the broader population."
Pikulski says he is working to bring balance to the reading debate. Whole-language proponents, he says, have been misled their success with children for whom reading comes easily, incorrectly deducing that learning to read is as natural as learning to speak.
On the other hand, he also fears that the NIH research with children having serious reading problems will lead to intensive phoneme training and phonics instruction for children who don't need it. "We have to be careful," Pikulski says, "that we don't say that everybody needs 20 minutes of phoneme-awareness training."
But Lyon disputes that the NIH studies' conclusions are restricted to poor readers: "Normal readers require phoneme awareness and phonics and whole language just as lousy readers do. ... We're finding that normal readers benefit as well."
What leaves the scientific laboratory as a scholarly debate becomes a bare-knuckles brawl on the schoolyard.