By combining brain imaging with behavioral tests, he studies the neural basis of learning in children, including those with dyslexia or who are at risk for reading difficulties. Dr. Gabrieli is also interested in the development of cognitive skills in school-age children, and in identifying ways that neuroscience could help improve educational outcomes.
Dr. Gabrieli also has appointments in the Harvard Graduate School of Education and the Department of Psychiatry at Massachusetts General Hospital.
To learn more, visit the Gabrieli Laboratory at MIT.
How can neuroscience help improve educational outcomes for children with dyslexia?
Dyslexia is defined as a specific learning disability in reading that is neurological in origin. By neurological we mean it’s a brain difference in a child that makes learning to read difficult. So neuroscience is the way we have to understand brain differences, and for humans we primarily do that through brain imaging.
And aligned with the idea that it’s a brain difference — neuroimaging has shown in both children and adults many differences in how children and adults process print and language who struggle with reading. So the evidence is very powerful and compelling that there is a brain difference that underlies dyslexia.
So in the long run if we’re going to do something remarkable to help children with dyslexia, understanding the cause of it is an essential step. And that’s going to come through neuroscience if it’s a brain difference, which it is. So that’s that long run. In the short run, parents send their kids to school, and teachers teach reading, and the action occurs in the classroom for most children in terms of reading. And we’re not going to have brain scanners in classrooms or neuroscientists roaming around hallways in elementary schools.
So we — in the short run, the contributions I think are twofold. One is that neuroscience can add to the basic science understanding of what goes on in dyslexia in a way that touches on policies and practices that matter in schools and for children and families. For example, for a long time people said dyslexia only occurs in children who have a high IQ and a low reading score, and that those are the children who deserve or merit or would be helped by specific and intensive instruction.
Behavioral evidence, intervention evidence, and now recently neuroscience evidence has shown compellingly that dyslexia looks the same in the brain and in the mind regardless of IQ. You can have a low IQ, you can have a high IQ — that brain difference looks just the same. And that really gives us evidence that all of these children should be getting the kind of help that helps children with reading difficulties like dyslexia learn to minimize the consequences of that.
So neuroscience can help us have a firm basis for making public policy and practices in the schoolroom that really helps children in a data-driven or research-validated way. The last frontier is that sometimes understanding the brain of a child who struggles to read gives us insights that are beyond what we now understand from behavioral or educational measures.
For example, my colleague Fumiko Hoeft and I did a study where we looked at brain images and many behavioral and educational measures in children and then tracked them over the next two-and-a-half years. And we asked, what predicts which child will make progress in reading — relatively — and which ones won’t? And the brain measures were much more accurate predictors than any current educational or behavioral measure.
So this is a big insight into understanding why some children make some progress and some don’t. And the reason that’s important is those children who don’t make progress with current practices deserve a different kind of support, and rather than give them the same kind of educational inputs that we now do — and wait for them to fail — if we knew up front which child is going to not respond to current educational practices, we could try something different.
And so the brain measures at the moment predict which child is in that group who will struggle better than current educational and testing measures. That could change in the future – people are constantly developing new measures — but at the moment those brain measures are more predictive of long-term outcomes, and then give us a better hint as to which child needs a different kind of help.
How can brain measures help predict long-term outcomes for struggling readers?
Our study found that the brain measures were better predictors in children about age ten or twelve over the next two-and-a-half years. And we don’t know why. We do know that that’s really important, because right now the way that interventions work — or special education works — is that children are put into some kind of program, and some children make substantial progress, which is great.
And some children don’t. And we only know that at the end. We only say, why, this didn’t work for this child and we’re really frustrated and disappointed and I’m sorry that it didn’t work. If we knew up front individual differences from child to child, and know that one kind of program is likely to help a child — well, we want to give that program to that child. But if we know another child won’t benefit so much from the program, then we really need to develop new kinds of programs targeted at those children, instead of just repeatedly sending them into programs of education that will not work.
So over time it could be that behavioral measures or testing measures or school measures will predict this differential or individual response to education, but at the moment the brain measures are more compelling and send us in a direction of thinking about forms of education that are personalized or individualized based on the brain and mind of a particular child.
How are the brains of people with dyslexia different?
The brain differences are really evident in both children and adults on average in terms of people with dyslexia or people without dyslexia. The most common differences are that areas of the reading network that respond to print show less response in several brain areas that are typically engaged in typically developing readers.
A second striking finding is that some of the white matter pathways — those are the pathways that connect different parts of the brain into big networks, and we need that for reading — those look very different, especially in the left hemisphere — the reading hemisphere — of children and adults with dyslexia. We showed in collaboration with Nadine Gaab some years ago that you see this difference even before a child has his or her first day in school.
So it’s not a consequence of something the teachers do or the school does, it’s something different in that child’s brain, who will find reading more difficult. And Nadine spectacularly has shown this even in infants born into families at high risk for dyslexia. So we can see this path of a developing brain even before print becomes part of their lives — even before learning to read becomes part of their school experience — that’s already showing a difference in brain structure and function that will make reading extremely difficult for some children.
What are the causes of reading difficulties? And should interventions change based on the cause in a particular child?
So children come to have reading difficulties in a variety of ways. There’s evidence that it runs in families, and that suggests a strong genetic component for some children who seem to be, from the moment they’re born, at high risk for reading difficulty. Other children, we think have come to reading difficulty because of environmental factors.
The best understood one is children who are in environments where there’s minimal exposure to language — where they hear fewer words, have fewer conversations with their parents. And those children are less school-ready and fall behind in reading, as well. In terms of education, the evidence is that the same kind of interventions help children who seem to have a genetic, neurological pathway towards reading difficulty and those children who have reading difficulty because of environmental factors.
All those children need to learn how to crack the code of print, and at the beginning grades it seems like the same kind of interventions that have been developed with dyslexia in mind also helps children who we believe are more likely simply to not be as ready for school due to environmental factors. Over the years those things change, because the children need all kinds of support in terms of background knowledge for reading and other kinds of skills.
But those first few years of elementary school, where we want to get all children to that beginning moments of reading — to be able to read a single word accurately, to be able to sound out words they’ve never seen before in text — those early steps of reading seem to benefit from explicit instruction in the early grades regardless of the cause of a child’s reading difficulty.
In later grades it becomes somewhat different, and we understand less of it scientifically, but we understand that children who have environments where they’re not highly supported — where they don’t do a lot of reading, and so on — that there’s a lot of other things that schools have to do to promote their knowledge of history, their knowledge of the world, things that help make reading easier for somebody as they encounter a new book or new text.
So it’s by no means the case that the early instruction in grades is all that you need. Education keeps on going. But we do know that those first steps of learning to read single words accurately children seem to benefit regardless of the cause of their reading difficulty.
What has neuroscience taught us about dyslexia?
The lessons from neuroscience — one of them is — one that almost feels like historically we should not be worried about anymore, but just to be really clear about it — many adults who have dyslexia had the experience that people ascribed their difficulty – lack of motivation or lack of parental seriousness or some sort of character flaw. And the brain imaging data just simply shows that it’s a brain difference from birth — before school and continuing for a child’s or an adult’s entire life.
So that completely has eliminated this other kind of misunderstanding of why a child struggles to read. So that’s been a huge step forward with clear scientific evidence. The other thing is that the differences are present early. And we think that, so often for many reasons, schools wait too late to help children, and that these differences are present the moment these children arrive for the first day of kindergarten or pre-kindergarten, and that action should be taken through early screening — through early intervention — because the brain difference is there already, and just waiting for that difference to hold a child back and then attempting too late to help a child — all the neuroscience says the problem is present at the first moment. All the help should be provided at the first moment.
Why is waiting to intervene so disastrous?
The obstacles to early intervention are primarily that people don’t know in school systems how to identify these children in an easy way. A huge challenge is that the best measures we have now — which are not routinely employed anyway — but the best measures we have now overestimate the number of children who will struggle to read. So school systems feel a pressure to say, well, we don’t know for sure, and you’re giving us too many children and we don’t have enough resources. So let’s wait awhile to be really sure.
But that waiting is disastrous. It’s disastrous by everything we understood, including the spirit of the child. Reading is a child’s first huge experience in education, and children are exquisitely aware that they’re behind their peers. We did a study where we asked first-graders, “How good are you compared to your peers in reading?” And the spontaneous simple answer was about as accurate as our three hours of testing.
These children know. Other children sometimes laugh at them when they’re do — reading aloud. They’re aware that they’re reading material that’s behind other children. They’re exquisitely aware, and that’s a huge burden to place on the small shoulders of a beginning learner. His or her sense of how effective they are, how successful they are, whether education and learning is for them – waiting is very destructive to that child’s spirit.
And so that’s – that alone would be a reason, but we also think that because these differences are present, letting that child fall further and further behind in reading just puts them way behind their chance to express their full potential.
Reading is essential for all of education and when a child feels they’re behind on that — that they’re not doing well on that — then they have a sense of themselves as a child who will not be a good learner. And that’s the last thing that we want for any child is a sense of pessimism about their ability to learn. And so we need to deal with that. We need to identify so early that they don’t even think of it as a problem.
So I think of it when I got my glasses as a second-grader. I was just, I need glasses to see. It wasn’t a big deal in some sense — reading is a bigger deal, but I’m hoping that we’re not too far from a day where the identification occurs just like an eye test, and it’s no more dramatic for a child to get extra support for reading than it is for a child to get glasses so they see the blackboard.
What are you currently working on in your lab?
So we’re working on two things that we’re very excited about. So one of them is, again, improving early identification through a combination of behavioral measures and brain measures so that we can get more accurate. So right now, for every child who’s identified as being at risk who truly is at risk, another child is identified as being at risk who is not. So we need to get that better somehow, so that school systems can address the child who really needs the help.
The second thing we’re very interested in is why one child responds to an intervention or support and another does not. These individual differences are very striking, and we often don’t talk about that, as we don’t in lots of education and medicine — we say a program works or it doesn’t work. But when you look a little bit more closely, what you find strikingly is that for those programs that are “good” it works really well for some children and almost not at all for others.
What is it about the differences in these children’s minds and brains that makes a particular program really effective for some children and really ineffective for others? And if we could understand that, maybe we could find a set of interventions or treatments designed uniquely for children with different patterns of strength and weakness. And we really need to understand that, because even the best programs nowadays are not effective for a meaningful number of children.
The Read Project: Can we predict who will struggle if we screen before kindergarten?
The Read Project was supported by the National Institutes of Health, and Nadine Gaab and I recruited a number of schools — a very diverse set of schools, about 20 of them — that allowed us to go into the school and with parental permission give a very short screening measure to over about 2,000 kindergartners, or sometimes we caught them in the summer before kindergarten, so before they got reading instruction from the school.
And our goal was to understand if we gave these kind of short measures and then looked at some children more in-depth, could we understand who would go on to become a poor reader by the end of second grade, where it’s pretty clear if somebody is doing well at reading or struggling? And what we found is that standard measures that are well-known now in terms of testing – phonological awareness and rapid naming – were very good at telling you who would go on to be a good reader — nearly perfect — and were pretty good at telling you who would go on to be a poor reader, but not good enough.
So for all the children who are identified as being at high risk for poor reading, about half of them went on to become poor readers and about half of them went on to become satisfactory readers. So we now — and that aligns with everything else that we know — our prediction there is good but not great. And so we’re very excited about trying to improve that through better research — through better measurement — and to figure out ways to be more accurate in that identification.
And so we don’t know why those half of children who look like they are at risk become satisfactory readers. Some possibilities are that children who are four and five years old are fairly noisy research participants.
Sometimes they didn’t have a good nap — you know, they don’t like the examiner, and they just don’t do a good job on a test. And if they don’t a good job on a test on a particular morning they count as being at risk. So it’s just the human nature of working with four- and five-year-olds. Other possibilities are that maybe they got more help, and it’s hard for us to know that in any deep way.
And maybe better educational supports at home and at school made a difference. Another possibility — and there’s a little bit of evidence for this — is that some children have a form of resilience in their brain that can develop alternative reading pathways more easily. So although most children read primarily through the left hemisphere as they become skilled readers — that’s the language hemisphere, and reading is tied into language — some children who do the best at overcoming reading difficulties seem to develop a right-hemisphere pathway for reading.
Not all children do, just the ones who seem to make the most progress. So it might be that there’s in some children a sort of backup, alternative pathway that they can build on — that they can be resilient with. They don’t become necessarily the best readers, but they can become quite good readers. And if they’re good enough it allows the rest of their potential to be expressed, you know, even in literature, but also in science and math and history and everything else where text is part of learning.
So we don’t know whether it’s one or the other of these possibilities, but we just do know that half of children make more progress than we would’ve guessed from screening measures.
How has our understanding of dyslexia changed?
The biggest changes, I think, in the understanding of dyslexia is, first, compared to over the last 20 years we have a clear understanding of the brain difference. Twenty-five years ago, 30 years ago, when people talked about a brain difference, that was a guess, and now it’s a scientific fact. We also know that interventions that are given earlier are more effective than later. So that’s really pushed up the need to understand what happens early in a child’s development, both behaviorally and in the brain, and to get them rapidly to effective interventions.
And as the evidence on that has been compelling. So we wait too long now, and I think we want to be in a very active mode from the first moment when children are born to be ready to identify children who need that extra help and to give them that boost, not as a consequence of reading failure but as something that prevents that reading difficulty to become a major factor in their lives.
How can neuroscience help teachers?
I understand that the schools have a very difficult mission, and a complicated one. And I’m both sympathetic — you know, a part of me wishes that we knew how to be more helpful to schools as scientists than we are now — so that I think schools have huge challenges in terms of how teachers are prepared, in terms of how teachers are resourced and school systems are resourced. So it’s partly frustrating as it is for many people to make changes in our educational system. It’s very hard.
But it’s essential. So it’s one of these long-term efforts with a shared mission that we all have, to help these kids develop, and if there’s one thing that’s common in my experience with teachers and educators, it’s that they’re so passionate to help children develop. We just have to figure out — together, you know — how to make that happen really effectively.
One misconception about the role of neuroscience in education is an over-optimism about what neuroscience can do. I think people were so excited — I was too — to see these brain differences being visualized, being measured so clearly, that people sometimes said, well, the solution is at hand, and we still have quite a ways to go to really help children. So just knowing that we can measure and visualize this biological basis is not a solution to helping kids in the classroom. So I think there was too much optimism placed on that.
What are you most optimistic about with regard to dyslexia?
I’m very optimistic that in the early grades we can do fantastic things for children in the coming years. I think there’s a partnership among parents, amongst educators, among scientists — in the political realm as well — that wants to do something terrific for kids in the earliest phases of reading.
And that’s where we understand the most about what we can do. That we still have to worry about reading past third or fourth grade – I mean, children have to keep reading, students have to keep growing into adolescence and higher education, but we have so much scientific evidence about forms of education that really help children in those earliest grades that it feels to me like if we could make a major progress in almost any area of education that’s the one of getting so many more children to be reading at the grade level that we want them to read.
And once they read well enough they’re sort of empowered in the whole realm of education. Reading remains the portal to almost all knowledge in education — almost skill acquisition in education.
And so I think that early thing — to get everybody to the starting line so that as they move into middle school and into high school they’re able to access knowledge and skills through reading — is something that is in our reach with enough cooperation amongst all the parties that have to work together.
What’s your advice to parents of struggling readers?
Parents have a great challenge in navigating the school systems on behalf of their children, and the best thing that they can do is to get some form of evaluation, which can be a huge challenge. They’re often expensive. They’re not promoted by the school. But parents who advocate for their children often are effective, and without that, unfortunately, the school systems sometimes wait.
Understandably, they’re waiting to make it really clear who’s a struggling reader. But we know — from everything we know from education and psychology and neuroscience — waiting is not a good strategy. Waiting for failure and beginning with that is waiting too long. So I think that the parents just to be as active as they can. It’s often a great struggle, depending on the teacher and the school they deal with.
There’s an incredible variation in how understanding and supportive schools are. And I’ve talked with many parents who are very frustrated about what seemed to be a slow or confused response from their school. One of the things that’s confusing is that many schools will say that we cannot tell you your child has dyslexia. And this is a kind of a arcane technical point about what dyslexia means as a label, and whether it’s a clinical label or an educational label.
But for many parents it’s much more confusing, because there’s a commonsense use of the term — is what parents go by. So this is very confusing that the use of that terminology alone has been a source of great frustration for families trying to understand why their children struggle.
What can we do to combat the stigma of dyslexia?
Stigma is a great problem in dyslexia. It’s even sometimes been a worry — if we could screen perfectly, would we set low expectations for a child? If we could perfectly identify a child at risk in kindergarten — which is our dream in one sense — to intervene early and never let that child fail — I think there is also a sort of social and moral issue that we have to be careful about, which is not to diminish the expectation of that child’s future, either from the viewpoint of the school, but most important from that child.
But again if we could make it less mysterious — if we could just say: if you, you know, break your wrist you get a cast, if you don’t see so well at a distance you get glasses, and if reading is a bit of a thing for you, you get a little bit of an educational support for that, as well. If we can make it very objective and scientific, then I’m hoping we can drive away the sense of stigma and just say it’s another element of human diversity — and one happily that we can support, so that the rest of a child’s promise fully flowers in that child.