Neuroimaging is opening a window onto how we learn to read. Acquiring this complex, demanding skill, researchers find, is a richly orchestrated process that recruits and connects diverse brain regions.
Ultimately, they hope, what’s learned in the laboratory will guide more powerful teaching methods adapted to the quirks and variations of individual children’s brains.
Dyslexia has drawn the most attention. Earlier research identified defective phonological processing—a relative inability to break words into their component sounds—as the core problem, while other studies focused on visual areas that recognize the shapes of words (the left temporal-occipital cortex, or “visual word-form area”). Researchers are now looking more closely at how parts of the brain that process letters and language sounds work together.
"When looking at letters and words, skilled readers activate a specialized part of the visual system in a way that dyslexic readers typically don't. One thing that is driving this effect may be its coordination with the phonological system,” says Bruce McCandliss, chair of psychology and human development at Vanderbilt University. As the reader seeks phonological information from print, “it puts pressure on the visual system to reorganize and deliver that information in more and more effective ways.”
In an fMRI study reported in the March 2010 Cerebral Cortex, McCandliss and his colleagues recorded brain activity in the visual systems of literate adults who were asked to judge whether spoken words rhymed. “We saw massive top-down, very specific activation of the visual word-form area,” he said. The fact that a part of the visual network was engaged by attending to phonology suggests that “these two systems really come to interact” in fluent readers.
Another study, which appeared in the March 2010 Brain, identified differences between dyslexics and normal readers in brain functions linked to sight-sound integration. Researchers led by Vera Blau, then at Maastricht University in the Netherlands, examined activation patterns when school-age children simultaneously looked at letters and heard sounds.
Among fluent readers, several areas (including the interface between the visual and auditory cortex) were more strongly activated by “congruent” combinations—the sounds and letters matched—than by incongruent ones. No such difference was seen in dyslexic children.
“To become a fluent reader you need to develop an automated integrated representation of how the specific letter on the page corresponds to a particular speech sound,” she says.
Differences between dyslexics
Other researchers are exploring the neurobiology behind one of the mysteries of dyslexia—its varying course. About one-fifth of dyslexic children eventually develop adequate reading skills, and standard testing has limited success in identifying those most likely to progress, says Fumiko Hoeft, of Stanford University, lead author of a paper in the Jan. 4 Proceedings of the National Academy of Sciences.
Hoeft and her colleagues conducted fMRI studies to measure brain activity as 25 young adolescents with dyslexia performed a reading task, and diffusion tensor imaging (DTI) studies of white matter fibers that linked various parts of the brain. Two and half years later, they retested the adolescents’ reading abilities.
Analysis of initial scans showed that differences in activation of a right frontal area, and anatomical differences in nerve tracts linking front and rear parts of right hemisphere could distinguish with 72% accuracy those dyslexics whose reading improvement would be better than average when retested. A more sophisticated analysis that compared activation patterns across the entire brain predicted reading success with over 90% accuracy.
Hoeft noted that corresponding frontal areas of the left hemisphere play an apparently important role when fluent readers read. The successful readers in her study “might be using the right hemisphere to compensate,” she said.
By the same token, “the equivalent white matter fibers in the left hemisphere are known to be important for language learning” in normal readers, she said, and right hemisphere tracts might have taken over their function in compensated dyslexics.
The fact that whole brain analysis provided the most accurate prediction of subsequent success confirmed that “many parts of the brain other than frontal areas are essential in reading, including some we might not have predicted,” Hoeft said.
Martha Denckla, professor of neurology at Johns Hopkins University, praised the research and observed that for the youths in the study, at least, “the neural substrate for improvement seemed to be in the right hemisphere system.” But she noted that recent, as yet unpublished research using different measures of brain function found that left, not right frontal activation predicted reading success.
“The problem is that we’re dealing with such a moving target,” said Denckla, also a member of the Dana Alliance for Brain Initiatives. “How to define dyslexia is always a question: do you include reading comprehension? Spelling? If you have 10 researchers, you have 11 opinions.” And age is a factor: A similar study with younger children might have had different results.
Denckla hopes to see the emergence of biomarkers for more reliable and objective definitions of dyslexia and reading readiness than behavioral tests can provide. Brain imaging studies like Hoeft’s and, ultimately, genetic studies could help identify children who might be “dyslexic” in the first grade but biologically ready for reading instruction several years later.
Neuroscience may perform an even more central service in helping educators teach more effectively. “Educational activities play a potentially key role in shaping the brain reorganization” underlying learning to read, “and some activities may be working better than others,” said McCandliss.
Ideally, research will help clarify “what aspects of the learning experience wind up being crucial for driving change in brain circuitry,” he says. “What is the teaching process? What makes an effective teacher so effective in transforming the mind of a learner?”
Researchers who had identified structural and functional differences between dyslexics and skilled readers, “are now asking whether reading intervention brings about anatomical and functional brain changes that we can evaluate with these same tools,” says Guinevere Eden, director of the Center for the Study of Learning at Georgetown University, and president of the International Dyslexia Association.
Her own research, reported in the Oct. 26, 2010, NeuroImage, linked increases in grey matter volume to reading improvement following intensive tutoring. But the question remains: “If I learn a particular brain area is involved [in learning to read], does that mean I would change the type of intervention or aspects of its administration?” she says.
Vera Blau, whose research identified deficits in sight-sound integration among dyslexics, is exploring this question as well. In a follow-up study, her team scanned dyslexic children before and after training that focused on phonological awareness and word-sound relationships. They are currently analyzing the data; if, as they expect, the second scan shows a pattern closer to what they see in normal readers, “it would make a case for offering audio-visual interventions early on,” Blau says.
Another goal of educational neuroscience is to clarify differences among learners, McCandliss says. Understanding how neural connections between language and visual systems differ from one child to another, for example, may enable teachers to tailor reading instruction more effectively.
For all the excitement engendered by neuroimaging studies, “we need to recognize that these are research tools right now,” says Eden, who as president of the IDA constantly hears questions from parents and educators about putting them to use. “It’s easy to overreact. We have a lot of work ahead of us.”
Martha Denckla thinks that neuroscience may serve education best simply as the voice of reason. Recent dyslexia studies suggest that “biology is destiny… you can’t neglect underlying brain circuity as a factor, and rosy optimism about wonderful interventions should be limited by neurobiology,” she says.
Realism about brain development could counter the social pressures that have driven school systems to teach reading earlier and earlier. “They are doing enormous harm by blithely disregarding neurological readiness to learn these skills,” Denckla says. Correcting their misguided enthusiasm “is a very important thing neuroscience has to do.”
If neuroscience can inform education, the relationship is bidirectional, McCandliss observes; unraveling the complexities of reading can open up insights into how the brain works.
He cited a recent fMRI study of people who learned to read as adults. As reported in the Dec. 3, 2010, Science, the researchers found that the fusiform gyrus, an area associated with shape recognition, was more responsive to words and letters in these individuals, compared to their peers who remained illiterate—an observation that had been made before. But this area was also less responsive to faces, in the new readers.
“Looking at an educational process added to our general understanding of functional reorganization in the adult brain,” McCandliss says. “That’s exciting.”