Dyslexia
I'm the proud owner of a bookish child. Since he learned to read it's been the main activity of his days. Francis Spufford in The Child That Books Built describes how his parents could tell when he was reading when a special silence descended: "a reading silence... that somehow travelled through walls and ceilings to announce that my seven-year-old self had become about as absent as a present person could be." I know how they felt. The reading silence is unmistakable, and it can go on for hours.
But there are only so many hours in the day. The boy does not spend much time kicking a ball or climbing trees or generally running about. From the complicated nexus of interest, aptitude and accumulated practice, kids emerge with very different levels of ability in different areas. My boy has more ability than most of his contemporaries at reading, and less than most in gross motor skills. At school, the kids who are worse in gross motor skills get taken out occasionally for extra practice - throwing a bean bag, jumping on a mini trampoline, that sort of thing. And of course the kids who are struggling with reading get taken for extra literacy help.
In Ireland at least, primary schools are quite free to allocate their teaching resources however they see fit - a child doesn't specifically need a diagnosis of any disability to get some extra help. But if they do have such a diagnosis, it can change things for them. They can receive an exemption from studying Irish (it's mandatory in Ireland, but considered difficult by most learners). They can get assistive technology, like smart pens that can scan printed text and read it aloud. Later on they can get extra time in exams, or a scribe to help with writing, depending on the extent and type of their disability. One of the most common diagnoses is dyslexia - a delay in learning to read, not better explained by something else like a visual or learning disability (Gialluisi et al 2021). A slightly different definition comes from the Discrepancy Model - if a learner has normal results in a non-reading intelligence test, but much lower scores in reading, they have dyslexia (Siegel 2006).
If we think of a disability in a broad sense of having a problem doing something important, dyslexia clearly fits the bill. And if someone has a problem doing something important, we as a society generally help them. If a wheelchair user needs to access a public building, we build them a ramp or lift. Everyone needs to able to at least read a medicine bottle, and should get the help they need to do so. The accommodations provided don't really hinge on the aetiology of the disability. The ramp helps the wheelchair user whether they have arthritis or cerebral palsy, for example. But dyslexia is usually seen as something more than just a functional disability - it's a specific medical condition, which needs to be diagnosed. And whether or not you have the diagnosis of dyslexia affects the kind of help you get. But how much should we let the diagnosis determine the accommodation, given that it's a diagnosis of exclusion? If the learner who's struggling to read doesn't have a genetic or chromosomal disorder, or a a brain injury, or a visual problem, or a hearing problem, or anything else, then by definition the condition they have is developmental dyslexia. Is it really worth diagnosing anything in this case, when in way what we've done is confirm the learner doesn't have a disease?
The more practical-minded among you are probably thinking "It doesn't matter what we diagnose them with, as long as diagnose them with something we can write on a form to get government money to provide help. If they had a generalized learning disorder we'd just put that on the form and get the same money." This is definitely true! And I'm not concerned about diagnoses of convenience to get much-needed resources. What I am concerned about is using a diagnosis to gain an advantage over children with a similar (or greater) level of need. This brilliant, infuriating Guardian article shows how some middle-class parents in England are paying privately for a dyslexia assessment. Having gotten the diagnosis, they then sue their local authorities to force them to place their children in extremely expensive private specialist schools. Of course, this means the local authorities have less money to help the kids left behind in the state schools, some of whom have much more complex needs and disadvantaged backgrounds. They are certainly would benefit more from the tens of thousands of pounds spent on private schooling. No-one ever made a policy decision to instead spend that money on kids from privileged backgrounds who just have trouble reading. But the model of disability provision forced their hand.
And of course, even the extra accommodations given in a mainstream school environment can be manipulated. A dyslexia diagnosis can get you extra time to sit an exam, but every student would benefit from extra time. Aria Babu notes on Twitter that in the UK, 26% of students now get extra time, and are more likely to do so at a private school than a state school. Are there really more disabled students at private schools? I'd expect it to be the reverse. I'm suspicious that this is more to do with competitive university entry than an explosion in intellectual disabilities (there are 14% more children with special arrangements than last year).
Imagine a pre-literate society for whom spear throwing is the most important thing, and kids spend hours a day learning it. If a child failed to learn how, they might coin a term like dysiactia (if they were fond of Latin roots) and diagnose him with that. A visitor from our society might think "Eh, I'm sure not being able to throw a spear is bad, but is it really a medical condition? I mean, the kid could have a shoulder injury, or cerebral palsy or something, but that would just be two different conditions that cause one particular symptom." Dysiactia wouldn't exist in our society, and dyslexia wouldn't exist in theirs.
Now imagine if our society got 100 kids to learn spear throwing for a month. After a month we plot a graph of how far they can throw. We'd probably end up with a bell curve - most kids somewhere near the middle, a few at the top end and a few at the bottom. We would probably get a similar curve for any human ability we cared to measure. Sure, we could set some arbitrary threshold for how far they should be able to throw, and call it a disorder if a child fails to reach it. But that doesn't automatically make it a real disorder.
I'd better pause here and clarify what I mean by real! Difficulties learning to read are very much real, very significant and are quite common. Almost every Irish primary school has dedicated special education teachers, and literacy is a big part of their job. And reading fluently is a requirement for many things we might want to do in life. The question is, is dyslexia a single condition, or a dustbin category of disorders united only by their symptoms?
There are plenty of studies that have tried to get behind the symptoms and see what's going in the brains of people with dyslexia. Reading, of course, isn't a native function of the brain. If there are modules in the mind for language, reading can't be one of them, as reading was not part of the environment where the human brain evolved. Many (Vandermosten et al 2012, Ozernov-Palchik and Gaab 2016) think that dyslexia is caused by a problem in phonological awareness. That is, dyslexics have problems breaking down speech sounds into meaningful components. This then leads to problems connecting written symbols to those phonological components. In this model, a "real" underlying problem in speech perception manifests as a problem in the specific culture-bound activity of reading.
It's a pretty well-supported theory, but I'll note that if there is an underlying speech perception problem in dyslexia, it must be subtle. I don't think I know a single adult (without a hearing or intellectual disability) who has any trouble following a conversation. But statistically I must know several dyslexics, as at least 5% of people have it. But subtle as these perception deficits are, they have been found in the lab (Siegel 2006).
Of course, there are competing theories, involving visual processing deficits and other problems. This GWAS from 2020 finds some interesting associations.
Side note: A GWAS is a Genome Wide Association Test, where you sequence the entire genome of every subject in a large study group, and see which genes are associated with which attributes of the study group (like disease risks or height for example). Specifically they search for variants of genes that differ by a single letter in their genetic code - Single Nucleotide Polymorphisms (SNPs, pronounced "snips"). (There are other kinds of genetic differences, like copy number variations).
We are living in a golden age of GWASs, as the falling cost of genomic sequencing has made it possible to do these kind of studies with big enough sample sizes and statistical power. GWASs have helped to find some of the so-called "missing heritability" in heritable traits. If you do a classic twin study on something like height, you might find it's 80% heritable. But if you search for individual genes for height, you might not find that the twins have enough variants of known genes in common to explain their similarity.
It turns out that many traits are affected by small influences from many rare variants scattered across the genome. A GWAS can sum all these tiny affects together into a Polygenic Score (PGS) for a trait.
The GWAS above found that lower PGSs for intelligence are significantly associated with dyslexia risk. And higher risk scores for ADHD, bipolar disorder and schizophrenia also explain a lot of the variance in dyslexia risk scores. This doesn't agree with the auditory processing model at all! Dyslexia is meant to be independent of intelligence or ability to concentrate, and I don't see how mental illness fits into the picture. Is the causality Have reading problems -> Have tough time at school -> Increased mental illness risk
?
But they do find a reasonably big chunk of dyslexia risk is caused by genes associated with lower word-reading ability, independent of intelligence. And intriguingly, they find that dyslexia risk is reduced by a higher PGS for differences in... the primary auditory cortex! Extract: (DD is Developmental Dyslexia)
Finally, the analysis of PGS influencing different brain cortical regions revealed a small, but robust and significant, protective effect against DD risk for a PGS increasing thickness of the transverse temporal gyrus. This region, also known as Heschl’s gyrus, is located within the primary auditory cortex — which is fundamental for auditory discrimination and speech perception — and has been previously implicated in dyslexia by neuroimaging evidence, although not always consistently across studies.
So some support for the auditory processing model in the genetics. Let's dig into some of those neuroimaging studies. Vandermosten et al looked at the density of neuron connections in different brain regions (using diffusion tensor imaging, a kind of directional MRI). And they do indeed find differences between dyslexics and non-dyslexics. Among other things, they found the left temporoparietal region pretty consistently has a lower connection density in dyslexics. This region contains Wernicke's Area, which we know is important for understanding language.
Both Vandermosten and others (Elnakib 2014, Casanova 2010) also find dyslexics have a larger corpus callosum. This is the dense bundle of nerves that connects the left and right hemispheres of the brain. So more inter-hemisphere connectivity equals... less reading ability? That seems surprising to me, but maybe that's just because I don't know much about neuroscience! Pugh et al (2001) think that dyslexics are more likely to use a visual processing part of the right hemisphere to compensate for problems in the language areas of the left hemisphere:
A ... compensatory shift, from posterior LH to posterior RH, likely reflects the development of an additional word recognition process that is essentially visual–perceptual; graphemic patterns in the printed word are associated directly with entries in the [dyslexic] reader’s mental lexicon. Thus, these visual patterns do not code the phonological or morphological information that the [non-impaired] reader perceives within the printed word, but instead represent the printed word as a nonlinguistic visuo-semantic icon.
To speculate wildly for a moment, maybe as the dyslexic brain recruits the right hemisphere to help with language processing happening in the left hemisphere, the corpus callosum grows to accommodate the extra inter-hemisphere traffic.
A few caveats with all the imaging studies - sample sizes are often in the dozens, and they sometimes come up with contradictory results, like a pattern of activation being reversed in adults versus children, or females versus males. fMRI studies are not straightforward to analyze, and if you don't carefully correct for multiple tests, you can end up finding that a dead salmon's brain activates differently in different social situations.
But that being said, we can be fairly confident that dyslexics and non-impaired readers tend to differ in their brain structure. Which in a sense isn't surprising - they were hardly going to differ in their spleen structure! Ozernov-Palchik and Gaab mention another problem though:
It remains debated which brain characteristics of dyslexia are a result of reduced reading practice (e.g., due to the daily struggle to read) and which predate the onset of reading instruction.
That is, maybe the brain differences haven't caused the reading problem. Maybe the brains of fluent readers change as a result of learning to read, in a way that non-readers' brains don't. It's analogous to the famous study of London cab drivers' brains as they learn "The Knowledge". Drivers who successfully memorized the layout of London's streets had an enlarged hippocampus compared to controls, or trainees who failed The Knowledge. And it wasn't the case that they had a larger hippocampus to start with, which gave them an advantage. They all had roughly the same size hippocampus to start off.
But the paper above concludes that at least some of the differences in dyslexic brains existed beforehand:
Neuroimaging studies in pre-reading children and beginning readers have provided ample evidence for the presence of brain alterations early in development and prior to formal reading instruction.
So after considering all the evidence, I'm tentatively accepting that there is something specific and consistent behind at least some cases of dyslexia. It's partially genetic, and involves a slight problem in breaking down heard speech into phonological components. When it comes to learning to speak and understand speech, it can be overcome. But it becomes a problem when the child has to connect those phonological components to written symbols. If you don't have a solid grasp of the sounds, the symbols are meaningless.
Maybe it's like me trying to learn the Pinyin system for writing Chinese sounds (I failed). Chinese distinguishes 4 different "ch" sounds, that to me all sound identical. In Pinyin they are represented by j
, q
, zh
and ch
. But I can't tell the difference between the sounds, so I don't have a hope of producing the right sound when I'm reading.
But if I'm somewhat convinced that there really does exist a "classic" aetiology for dyslexia, I'm even more convinced that the bin labelled dyslexia has loads of people with a completely different problem, like attention deficits or visual problems or mental illness or a chaotic home life, which has left them with reading problems. And it contains kids who are merely below average readers but are cursed with ambitious parents who won't accept "below average".
And while the neurological aspects of dyslexia are certainly interesting, for an individual kid I don't think there's much value in "diagnosing" them - just check the reading scores. And if they need help, give it to them! I suspect putting a label on a child's struggles won't much change the kind of help they get from their teachers. The "treatment" for dyslexia is more reading practice. I mentioned above how dyslexics may use more visual decoding of words compared to non-impaired readers. This might suggest using an approach based on visually recognising common words (like the Dolch Sight Words List) in addition to phonetics. But a teacher doesn't need a dyslexia diagnosis to use the Dolch List. They can just print it out and teach any child with it who they think might benefit. My sense is that by and large teachers know what a child needs because they know them as an individual, not because of the specific diagnosis.
One reason I'm slow to dispense with the label of dyslexia completely is the reassurance it gives to sufferers. If you're a child who is watching all your classmates learn to read while you struggle, it's natural to wonder "what's wrong with me?". It might be easier to hear "you have dyslexia" than "you just happen to fall near the bottom of the bell curve". Most kids won't care that that's effectively the definition of dyslexia! And sadly, other kids can be cruel. Kids who struggle with reading may be mocked by their peers. If the label of dyslexia can invoke a taboo against mocking the disabled that's worth something. Ideally kids wouldn't be mocked for struggling in any activity, disability or no disability. I do think kids in this generation are a little kinder and more accepting of difference. But schoolyard taunting isn't going to disappear any time soon.
In the end, dyslexia is just a label. That doesn't mean it isn't useful. Many useful terms don't map cleanly onto a real thing in the world. They divide the world into categories which may be more or less useful depending on what we're trying to achieve (because The Categories Were Made For Man, Not Man For The Categories). If we want to divide readers into the disabled and the non-impaired to decide who needs more help, fine. But don't cling too tightly to labels for categories that are affixed only for convenience.