Down in the basement of Haskins Laboratories, psychological researchers are starting to unravel a mystery that has long puzzled educators: What happens in the human brain as it wrestles with words?
Crowded around computer screens, scrolling through images that show the brain as it reads, the researchers are gaining insight into how we perform this crucial, yet complex, task. It's a scene being repeated in labs everywhere, part of an explosion of imaging research on learning over the past decade.
While researchers at sites such as New Haven-based Haskins investigate reading, psychologists in Paris watch what happens in people's brains as they tackle math problems, and neuroscientists at Stanford University in California puzzle over unusual brain patterns in people with attention-deficit hyperactivity disorder (ADHD).
Imaging research is pinpointing what the brain does as people read, calculate and estimate. It's also showing what goes wrong when people have difficulty with those tasks.
On the forefront of this work are the researchers at Haskins, whose labs are affiliated with Yale University and the University of Connecticut. They find that the brain of someone with dyslexia functions differently from a typical brain as it processes phonemes--the "c," "a" and "t" that come together to form "cat." Kenneth Pugh, PhD, is the experimental psychologist who heads the Haskins side of the research collaboration with Sally and Bennett Shaywitz, both MDs, of the nearby Yale University School of Medicine.
Back at the Haskins headquarters, Pugh grows animated as he clicks through slides of brain images imprinted with computerized red dots. The images show different blood flow patterns in the brains of fluent readers and dyslexics--evidence, says Pugh, that this research is beginning to reveal dyslexia's neurobiological basis, or signature in the brain.
"We're bringing together imaging with sophisticated cognitive-behavioral work to better understand how reading failure occurs and, from this, better techniques to correct it," says Pugh. "That's good use of red dots."
Imaging during specific learning tasks has potentially important implications for education, along with more basic imaging research being doneon the role of memory, attention, emotion and motivation in learning. Not only does it help with diagnosing and treating learning disabilities, but itmay even change the way we teach all children, say cognitive experts, helping those slower to grasp reading or math to tap other, more efficient brain circuits.
"Imaging offers the chance to document a child's learning disability, as well as to catch the disability early and change it with remediation," says Duane Alexander, MD, director of the National Institute of Child Health and Human Development (NICHD). "It also has the potential to help researchers develop more successful learning approaches in general." Under Alexander's leadership, NICHD has been a major supporter of imaging research on learning. Other supporters include the National Institute of Mental Health (NIMH) and the National Science Foundation.
Imaging research spurs controversy, though. Some cognitive researchers aren't yet ready to venture into the basement, believing there's still basic research to be done upstairs. Still others think the brain is too complex to crack with imaging, or that imaging is the wrong tool to get at the brain's inner workings.
But for the most part, researchers and educators agree on this: If imaging research on learning is to be at all meaningful--well designed and interpreted--more psychologists should be involved.
Certainly that's the view of NIMH director Steven Hyman, MD.
"There are too many psychologists who still are brain phobic, and that is a tragedy because we have a shortage of really good cognitive neuroscientists in the imaging arena," says Hyman. "This technology is only going to live up to its enormous potential if it's harnessed to the best possible cognitive and affective neuroscience."
Casting light on the brain
Hyman wants to see more cognitive scientists and educators marrying their behavioral understanding of learning with physiology--bringing together an "outside" view of the brain with an "inside" one. It will take, he says, stepped-up collaboration between cognitive scientists and physicists, computer scientists, physicians and school teachers.
Still, imaging researchers have already made inroads into reading and dyslexia, and are beginning to make strides in math and ADHD. Some psychologists even believe that imaging can one day help people shift the way they use their brains to boost their learning performance.
First, though, researchers must establish typical systems of learning in the brain, something neuropsychologist Stanislas Dehaene, PhD, is doing with math in his lab outside Paris at the Service Hospitalier Frederic Joliot. He is using functional magnetic resonance imaging (fMRI)--which traces blood flow--to support his theory that memorized math facts inhabit the same brain area as language, but that "real number sense" lives elsewhere.
In studies on college students, Dehaene finds that people tap brain networks in the left frontal lobe--areas associated with such verbal memory tasks as stringing together a sentence--to do multiplication tables. But they tap networks in the left and right parietal lobes--areas in the rear brain associated with visual and spatial tasks--to perform estimations. The reason, he theorizes, is that estimation is more "intuitive" math, an innate sense of quantity developed through human evolution.
Based on his findings, he says, children who are slow to read can still efficiently learn math.
"Often people say you can't teach children math until they get language," he says. "I say you can still teach numbers to children in a nonverbal way. One way is to use black and white rods to do addition, subtraction and other concrete operations that map nicely to the nonverbal system without using language at all."
Dehaene's findings also shed light on the poorly understood area of math disability. Building on his work, school psychologist Lisa Rowe, EdS, at the University of Florida and the Nemours Children's Clinic in Jacksonville, Fla., is testing whether children with math problems have difficulty using the "intuitive" nonverbal math system that Dehaene finds associated with the rear brain. Based on her findings, Rowe hopes to eventually develop interventions that tackle developmental math disability.
That level of sophistication has already been reached in reading research with the help of a large injection of funding from the NICHD. The agency is supporting more than 40 reading research sites nationwide, including the Yale site. There, researchers believe they've zeroed in on the brain's phonological processing system--an interplay involving areas in the back of the brain that decipher words. The Yale/Haskins group's fMRI scans suggest that these regions are underactive in many dyslexics, a pattern also seen by researchers at the National Institutes of Health and elsewhere. Instead, most dyslexics show overactivity in the frontal regions where speech is produced, a finding consistent with research by educational psychologist Virginia Berninger, PhD, her colleague David Corina, PhD, and others at the University of Washington. They find overactivation in dyslexics' left frontal regions using different reading tasks and a different form of imaging.
Research from another NICHD-funded lab, that of Georgetown University's Guinevere Eden, PhD, also points to trouble in phonological processing but finds additional deficits in visual processing. "These results remind us of the complexity of the cognitive processing associated with reading," says Eden.
To get at correcting the phonological deficit, the Haskins/Yale team is using fMRI to evaluate an intervention program, developed by psychologist Benita Blachman, PhD, of Syracuse University. The program focuses on building phonological awareness by helping children match letters with sounds. The researchers theorize that after dyslexic children complete the program, their blood flow patterns will more closely match those of normal readers.
So far, scanning the brain before and after an intervention program is the most advanced use of imaging. But Douglas Chute, PhD, a neuropsychologist at Philadelphia's Drexel University, hopes that imaging can one day fix learning problems more directly.
"The point is to move from, 'Gee whiz, we're turning your brain on,' to knowing what to do with these images, how to use them in an intervention," he says.
With colleagues Paul Moberg, PhD, Phil Shatz, PhD, Britton Chance, PhD, and students Andy Zabel and Dennis Loreman, Chute is using near-infrared spectroscopy (nIRS) to image the frontal and temporal lobes. He hopes ultimately to use it in ADHD research--still in its infancy. But some imaging studies are beginning to localize ADHD problems in the brain, which could help with diagnosis. For example, fMRI research by neuropsychologist John Gabrieli, PhD, and others at Stanford University, finds that children with ADHD show disruptions in the frontal brain and striatal region that lies below it in the inner brain, an area involved with motor control.
Based on such findings, Chute hopes that eventually his research team can use nIRS--a portable imaging form involving a sensor clamped to the head--to find out if conditioned changes in blood flow tovarious brain regions can improve learning performance in ADHD children. In real time, children would watch on computer screens how their blood flow changes as they approach learning tasks differently. Chute believes that the technology could also help students improve in reading and math.
"I don't want to hold out false promise," says Chute. "But we're not going to know how the neuroimaging explosion translates to education unless we do this kind of work."
Too much hype about red dots?
But not every scientist is convinced that imaging is the right way to look at the brain. And despite a national push to fund imaging research, some heads of funding agencies are calling for more scientifically sophisticated imaging studies and cautioning against over-interpreting imaging findings.
For example, Hyman of NIMH believes that imaging is merely a tool for understanding learning, not the Holy Grail. Using it without a basis in cognitive research is akin to what he calls "neo-phrenology"--the only difference being that imaging traces colored dots instead of bumps. Phrenology was an attempt to match bumps on the skull with people's behavioral traits.
"What does it mean to just take a picture of the brain?" says Hyman. "It means nothing without an appropriate task or output." In other words, brain pictures should be a means of testing hypotheses about mental processes that come from cognitive psychology.
Other critics go a step further, claiming that cognitive psychology says much more about how people learn, and what works in teaching, than imaging ever can. Taking this view is John Bruer, PhD, president of the James S. McDonnell Foundation, a sponsor of biomedical, behavioral sciences and educational research.
"There's a wealth of cognitive psychology research that's more useful for improving teaching than simplistic brain images," he says.
Ultimately, progress in cognitive and behavioral psychology will drive brain research, not vice versa, says Bruer.
What's more, some cognitive psychologists don't think that basic cognitive research can be joined with brain imaging. One of them, Guy Van Orden, PhD, of Arizona State University, notes, for example, that researchers still disagree about the fundamental causes of reading disability--whether it's mainly a phonological, auditory or visual problem.
"There's lack of agreement in basic theory, and because imaging research is parasitic on that, you've got problems," he says.
Also, says Van Orden, because imaging always implicates one or more brain regions in studies of learning, researchers often overinterpret findings.
Surprisingly, a number of Pugh's research partners at Haskins agree, at least somewhat. Says Haskins psychologist Michael Turvey, PhD, "I'm not a great fan of looking at the brain lighting up and using that as explanation. That's archaic. I'm not impressed."
But Turvey does believe that imaging has promise "as a good dependent measure, used on a par with other techniques to get at the reading process."
And even Bruer concedes that quality neuroimaging research can help researchers craft better educational programs.
Building better brain research
In fact, that's what Bruer hopes to see more of--cognitive psychologists translating brain research for teachers, helping them apply findings in schools and seeking teachers' input on the future direction of brain study. It's a weighty charge, but other educators agree that without more involvement from psychologists and other cognitive scientists, teachers won't be adequately informed about the brain.
"The reality is that we have to develop an educational system that understands brain biology better, that's more scientifically literate," says Robert Sylwester, EdD, an education professor at the University of Oregon.
Adds educational consultant and former teacher David Sousa, EdD, "Educators are trying to change the brain every day, so they ought to know as much as possible about it."
What's needed, say Sousa and Sylwester, is more neuroscientists providing instruction in schools of education and in-service workshops for teachers. Potentially, cognitive psychologists can serve as a bridge between the neuroscience and education communities, says Robert Bjork, PhD, a cognitive psychologist at the University of California, Los Angeles. Psychologists who translate research into educational practice and plug practice findings back into research gain a valuable perspective, Bjork believes.
"The neuroscientists uncovering the results that hold the most promise for understanding how humans learn and remember are those who are also well trained as experimental psychologists," says Bjork. "An understanding of the basic methods, paradigms and theories that characterize modern cognitive psychology can help neuroscientists use today's imaging tools far more productively."
But cognitive psychologists can't do imaging alone. To produce quality research, they need to draw on others' expertise, says NIMH's Hyman.
"This is interdisciplinary science, period," he says. "You need cognitive or affective neuroscientists. You need physicists and computer scientists. You need people who understand brain circuitry."
Certainly, the need for teamwork is understood by the Yale/Haskins team of linguists, psychologists, physicists and neuroscientists. Despite the old-fashioned accommodations down in their labs, the Haskins researchers have arranged their work space in a modern, open floorplan. No one here is squirreled away in an office. They're yelling questions at each other over cubicle walls and peeking over one another's shoulders as they clack on their computers.
Even the milk in the kitchen is communal, and everyone takes a turn making coffee. The setup comes from the Haskins philosophy that research on the brain has to mimic the brain itself, with many parts working together as a productive system. And from the perspective of Pugh, at least, "psychology is a glue holding the system together, mapping the direction in brain imaging, now, five years from now and into the future."
* Dehaene, S., Spelke, E., Stanescu, R., Pinel, P. & Tsivkin, S. (1999). Sources of mathematical thinking: Behavioral and brain-imaging evidence. Science, 284 (5416), 970974.
Pugh, K.R., et al. (2000). The angular gyrus in developmental dyslexia: Task-specific differences in functional connectivity within posterior cortex. Psychological Science, 11 (1), 5156.
Richards, T.L., Dager, S.R., Corina D., et al. (1999). Dyslexic children have abnormal brain lactate response to reading-related language tasks. American Journal of Neuroradiology, 20 (8), 13931398.
Shaywitz, S.E., et al. (1998). Functional disruption in the organization of the brain for reading in dyslexia. Proceedings of the National Academy of Sciences, USA, 95, 26362641.
Vaidya, C.J. (1998). Selective effects of methylphenidate in attention deficit hyperactivity disorder: A functional magnetic resonance study. Proceedings of the National Academy of Sciences, USA, 95, 1449414499.
Van Orden, G.C. & Paap, K.R. (1997). Functional neuroimages fail to discover pieces of mind in the parts of the brain. Philosophy of Science, 64, S85S94.