A granny summary of an article by Dror Dotan and Naama Friedmann
Separate mechanisms for number reading and word reading: Evidence from selective impairments
The word “dyslexia” is one of the more confusing terms in the field of learning disorders. In everyday speech, we often use the word dyslexia to describe a learning disorder of any kind – difficulties in reading, in writing, in calculation, attention disorders, and more. Some might even go the extra mile with something like “I have a dog-care dyslexia”.
So let’s be more precise. Dyslexia is a specific type of a learning disorder that causes reading difficulties. Dyslexia does not refer to difficulties in writing (dysgraphia), in calculation (dyscalculia), or with dogs (catophilia?). Distinguishing between different types of disorders is important because each learning disorder requires different diagnostic methods and different treatment types. In fact, even the term “dyslexia” is not absolutely precise: there are many types of reading disorders (dyslexias), and similarly many subtypes of dysgraphia, dyscalculia, and attention disorder.
Learning disorders in math are often described by the general term “dyscalculia”, another confusing term. The problem is that there are several different types of dyscalculia: one may find it difficult to remember the multiplication table (sometimes not because of a specific problem with numbers, but due to a verbal memory deficit); may get confused in multi-step calculations; may make errors reading or writing numbers; or may get anxious when having to calculate. Each of these conditions requires a different diagnosis and treatment.
A deficit in number reading is an interesting case: it concerns math (because it’s numbers), but it’s also about reading and language. In the study described here, we examined the relation between the reading aloud of numbers and words. The basic question is this: do you use the same cognitive mechanisms when you read a word (“dogs”) and a number (“492”)? Put in terms of learning disorders: will someone with a word-reading learning disorder (dyslexia) always have a difficulty in reading numbers, and vice versa?
The short answer: no.
The longer answer: big no. There are people with a word-reading disorder (dyslexia) who can read numbers easily, and there are people with the opposite pattern. Moreover, a detailed analysis shows that completely different processes are involved in word reading and in number reading – see the diagram on the right. Even if we meticulously examine each of the sub-processes involved in reading words or numbers (the boxes in the diagram, and even finger-granularity processes), we discover that it specifically serves either words or numbers, but not both (almost: this still needs to be examined for few of the sub-processes. This will probably be done soon). In the present study we examined this way two previously-unexamined processes of number reading, and we showed that they are specific to number reading and do not serve word reading.
Why is this question – whether reading words and numbers involves separate mechanisms – interesting at all? We already saw the clinical reason: understanding this separation helps improving the assessment and treatment of children with learning disorders. Because dysnumeria (a number-reading disorder) and dyslexia are two separate learning disorders, they should be diagnosed and treated with different methods.
But the question is also interesting because it may help understanding how our brain works. If reading words and reading numbers are separate mechanisms, located in different brain areas (they indeed are), it begs the question – why is it so? Why would the brain “waste” two separate areas for two tasks so similar, the processing of letter strings and of digit strings? The reason does not seem to be the visual difference between letters and numbers: the visual properties of letters and digits are quite similar. A cool study proved this point in the following way: the participants saw the Roman digits I, V, or X, and had to press one of 3 buttons to indicate whether the number presented was 1, 5, or 10. The researchers observed (on fMRI) brain activity in the area responsible for processing visual numbers. Then, the same participants saw the same characters (I, V or X), but now they had to decide whether they saw the English letter i, v, or x. This second task caused activity in a different brain area – the one that processes visual words. Thus, the participant saw the exact same thing in both tasks (Roman digit identification and letter identification), i.e. there was no difference in terms of visual properties, and still their brain used different areas in the two tasks. The researchers proposed the following explanation. They discovered that the letter-processing brain area (the one activated in the letters task) has high neural connectivity to the brain areas of language processing. In contrast, the digit-processing brain area has high neural connectivity to the brain areas that process quantities. They proposed that to processed a certain system of visual symbols, the brain will allocate an area that is well-connected to the destination of the processed visual information. Letters represent words, so they require connectivity to the brain’s language areas. Numbers represent quantities, so they require connectivity to quantity-processing areas.
Following our research, we proposed another explanation. For words and numbers alike, the reading mechanisms can be classified into two types of processes: one set of processes handle basic elements – a letter or a digit in the visual system, a sound or a combination of sounds in the verbal-phonological system. A separate set of processes handle the word’s or number’s structure: for numbers, specific structure-detection processes identify, for example, how many digits the number has, and where the digit 0 is located (these two parameters determine how many words the verbal number contains). For letters too, specific processes identify the word’s structure: they detect letters typical to common word templates, e.g., the plural suffix “s”. Such letters are quite common in English, and are even more common in languages with deep morphology, e.g., Arabic and Hebrew. Crucially, although letters and digits may be similar to each other, the structure of letter strings is very different from the structure of digit strings, so different cognitive processes would be required for extract the structures of words and numbers. This may be why our brain prefers assigning different brain areas to deal with words and with numbers.
Interested in more details? The full article is here.