By Lindsay Clare Matsumura and Cheryl Sandora

Institute for Learning

Big Ideas from the IFL Interview with Dr. Julie Fiez

  • Many different areas of the brain work together as a functional system when people read.
  • Struggling readers often have difficulty engaging the phonological pathways of the brain while reading, and instead rely more on other areas of the brain to comprehend text.
  • Understanding more about the brain systems involved in reading for typical and struggling readers, and readers from countries that use different orthographies, could lead to more effective reading interventions in the future.
How do you study reading processes in the brain?
The primary method has been to use functional magnetic resonance imaging or FMRI. Basically, you use the MRI machine to measure changes in neural activity by indirectly measuring changes in blood flow. The idea is that as a part of your brain starts to work a little bit harder, you increase the blood flow and hence the oxygen that is flowing to that part of the brain. The typical approach to study reading processes then is to put children or adults in a scanner and ask them to perform reading tasks, for example, to read words or do a rhyme judgment.
What happens in the brain when people read?

In the brain, there isn’t one place that holds all of the knowledge you need to read. Instead, research suggests there are areas that hold information about what a word looks like [the orthographic system], areas that hold information about how a word sounds [the phonological system], and one that holds information about the meaning of a word [the semantic system]. So when you look at the word “cat”, you need the visual areas, the sound areas and the meaning areas of the brain to be talking to each other at the same time. And when they’re all active together, then you know that you are looking at the word ‘cat’ and it looks like this, and it sounds like that, and it means this.

The field differs on this point, but I believe that there is a part of the brain that in some way stores knowledge about the printed forms of words that you have experienced and have gained familiarity with. It’s like somewhere in your brain it knows what ‘cat’ looks like, and it knows the line segments and position of the letters. That’s what allows you to look at that visual stimulus and automatically know what it means. I say that loosely because what the word ‘cat’ means could include your instructed knowledge about cats, your visual knowledge of what a cat looks like, as well as your personal experience or memories of cats.

Are there differences between individuals with dyslexia and typical readers in how these different areas of the brain interact?

The bulk of the evidence suggests that the children who are poor readers or have some kind of diagnosed reading disability are not as effective in engaging in the phonological pathways of reading. Instead, they rely more upon brain regions that support semantic processing with semantic representation.

They also often show less activation in the areas of the brain that code visual word form information. What’s not clear is whether that region itself is not typical or if it’s because the phonological processing is poor and so the individuals are less skilled at acquiring reading proficiency, and therefore, develop less robust orthographic processing. Their visual word form area might be delayed in its development because the reader is working harder to build that visual word recognition skill.

Do you think understanding what’s happening in the brain for dyslexic readers versus typical readers could help us develop more effective reading interventions?

I think in all honesty, that’s where I would say maybe the field has still come up short.

Imaging gives you a different window on reading processes, but it’s not necessarily one that is suddenly going to solve reading problems.

Based on some of the research I am involved with now though, I personally am beginning to think that the focus on developing decoding skills makes sense to me for what you would do for a typical reader. Some of our research suggests to me that if there really are some of these neurological limitations, it might be that there are other routes to developing effective visual word recognition that might be worth considering and evaluating for children with dyslexia.

Some evidence for this comes from some of the work that we have done looking at Chinese and English bilingual readers, and also some work that we’ve done with artificial orthographies. What we have found is that even if you don’t have great decoding skills, there are ways in which you might develop good proficiency at more holistically processing and representing the entire visual word form and mapping that onto its pronunciation and its meaning. I don’t think that’s the ideal way for typical readers to learn how to read the alphabetic writing system, because they would be less equipped to be an independent reader. You stumble across a word that you don’t know, and you wouldn’t really sound it out very well. But if you can’t do that very well anyway, I think this might be another interesting direction to think about.

So what I hear you saying is that for people with dyslexia, who have a tougher time mapping to the phonological areas of brain, it might be helpful to develop reading interventions that leverage other systems?

Yes, or maybe systems that use similar pathways, but operate on a different grain size. There’s a lot of work that suggests that the phonemic level [the individual sounds in words] can be especially challenging for people with reading difficulties. For example, the word ‘baby’ has four different graphs, B-A-B-EE. If you mapped the orthography to the phonology at a larger grain size it might make the phonological demands easier. For example, the word ‘baby’ would be represented by two graphs, one for ‘bay’ and one for ‘bee.’ Instead of needing to blend across four graphs, you would only need to blend across two.

Appreciating the visual similarity between words and their mapping to phonology I think could be really important. For example, A-T. You would see it in words like ‘sat’ and ‘mat,’ but then you would also see it in ‘bat’, and then ‘battery’. This could help people with dyslexia get up to multisyllabic words more quickly because they wouldn’t have to decode the letter to sound stuff to put a word together.

Are you looking at this in your research?

We are beginning to. We are conducting a very small pilot study now where we are asking adults, typical readers and a person with dyslexia, to learn an artificial orthography. We are interested in looking at what happens when they’re trying to sound out a new unfamiliar word. And then we’re looking a week later to look at the orthographic learning, that is, to what degree can we show that they have a memory for having seen this printed word before and having decoded it. What we are finding is that the individual with the history of dyslexia is two or three times slower than the typical readers to figure out what this word is. The decoding is definitely more effortful. But this individual was able to succeed in most cases, so their accuracy was pretty high, and their orthographic learning looks to be the same.

This finding is consistent with some theories that suggest the problem is doing the decoding to get to that moment of word recognition. If you can do it, then you’re going to learn just as much as someone who doesn’t have dyslexia. If we find this is true in a larger study, this could be a very informative outcome because it would suggest that the problem is not with the brain’s ability to do the orthographic learning; it’s about getting to that point of word recognition. So if you support that moment of word recognition, then you’re going to get the same amount of learning from that encounter with the unfamiliar word.

How do you think we could use this knowledge? How might this be beneficial for teachers?

Well if you think about what happens normally, a lot of reading happens independently. So if you have a child who’s mastered the alphabetic principle, they’re spending time reading in school but they might also be spending time reading by themselves out of school. And if they’re successful, every moment of their reading is translating into building their orthographic knowledge. Whereas the child who’s struggling to decode is likely not achieving as much success in the classroom or in their independent reading. The same amount of time spent reading is not necessarily translating into the same gains in their orthographic knowledge.

By understanding the mechanisms, we might have a little bit better understanding of where in the process from seeing an unfamiliar word to sounding it out, to learning from that success, is the problem? And knowing where in that pathway the problem lies, and what specific brain systems are involved, we might come up with alternative pathways.

So if I were to answer your question, “How would this work be beneficial to teachers,” in my dream world, maybe a few years from now, this basic work would lead to a reading intervention for a subset of readers who have difficulty mapping to the phonological system that would leverage other systems and pathways.