Artificial eye dominance may speed reading for people with dyslexia
I have to admit that my only experience of dyslexia is via family members. My youngest daughter has just started high school, and she's struggling with all the reading associated with three languages, an issue that bleeds over into all the other subjects. In testing, she scores high on reading comprehension but really low on reading speed. My oldest son reads and reads and reads… but cannot write worth a damn. Both have trouble internalizing spelling rules and multiplication tables.
These all standard symptoms of dyslexia. Dyslexia is a spectrum disorder, one that covers many aspects of reading and writing, so when people start touting single causes, my skepticism goes into overdrive. But it turns out that new research on its causes is reasonably solid, and it raises some interesting questions.
Your brain in the mirror
When the brain creates an image, it's faced with a problem. The two eyes report two images that are extremely similar, but shifted with respect to each other. The displacement is awesome, because it provides us with better depth perception. However, in the absence of a large amount of alcohol, the brain still has to decide on a single coherent image so it has something to present to our consciousness. To do that, the two images are melded into one, which is fine for displacement. But for mirror images, the brain must choose a single image.
One idea for how this is handled is that, during the first eight years of development, one eye becomes dominant and determines the handedness of images.
This has led to the idea that the lack of dominance might cause perceptual problems during reading, which seems like an obvious line of inquiry to pursue. The problem, according to the paper, is that the idea of dominance is not fully accepted, and tests for dominance are not as reliable as we would like. The underlying problem is that the tests rely on spatial perception, so the reported dominance can change depending on how the test is set up.
This is where the researchers' main contribution comes in. In their test, you stare at a bright light in a dark room for a bit, then close your eyes and cover them with your hands. The image of the light seems like it is burnt in, and the after-image persists for quite a while (six or so seconds). The researchers then get the subjects to remove one hand or the other. This allows light to leak through the eyelid, which stimulates the neurons to keep reporting the image.
If one eye is dominant, and light leaks through the eyelid of the dominant eye, the after image will stick around for a lot longer compared to if light only leaks through the eyelid of the non-dominant eye. The test doesn't rely on spatial perception at all, and it's still relatively simple to perform.
In tests on subjects who did not have dyslexia, the standard tests and the new test for dominance were pretty much the same. That is, all the subjects seemed to have a dominant eye, and all tests agreed on which eye was dominant. There were two exceptions. In both, the standard tests could not identify a dominant eye, while the new test found one.
Unfortunately, all this tells us is that the tests agree; they could still be consistently wrong. However, tests on dyslexic subjects showed that the new test really was reporting something different. For these subjects, the standard dominance test was pretty close to random in choosing a dominant eye or failing to find one. The after image test showed that almost all the dyslexic subjects had no dominant eye. That is a pretty exciting correlation.
What's inside your eye
What came next was even cooler. The researchers imaged the retinas of the subjects. Humans have three color sensors in the eye: red, green, and blue. But the blue sensors are not located in the densely populated center spot of the retina. The researchers made use of the lack of blue light sensors in the center to map out the shape of the center spot.
It turns out that it is not really just one spot. For the dominant eye, the spot is nearly circular; in the non-dominant eye it is highly elliptical. This provides a link between eye-dominance and the physical layout of the eye, suggesting that the brain uses this physical difference as a signal for image processing.
If the brain really is experiencing a perceptual conflict during reading, then this should be possible to test. The researchers performed a reading test with their subjects. They used the fact that, although the brain presents us with one image, there is a delay between the the first eye forming an image and the second doing so. So, if you are sitting in a dark room and a page of letters is illuminated by a flashing light—the light flashes are too fast to perceive—then only one eye at a time can form an image. In this case, the brain should not be presented with conflicting images.
The researchers state that there was no change in reading performance for control subjects and, more importantly, the reading performance of dyslexic subjects was on a par with the controls. That is a pretty amazing result.
That said, I'm still a bit dubious. That last part of the experiment was reported almost as an afterthought. There was no data on the actual reading tests, and it was not clear that the experiment was blinded at all (meaning that neither the researchers nor the subject should know if the light is flashing or not).
It is also unclear to me why this would work, because it's not like people can't read with one eye closed. If I put an eye-patch on my daughter she doesn't find reading any easier, yet this research suggests that she should. What am I missing?
My final comment is a reminder that dyslexia is a spectrum of symptoms. What excites me about this research is the physical link between eye physiology, eye dominance, and a traditional dyslexia diagnosis. However, it is unclear why the differences in eye physiology should result in problems with writing and, even more puzzling, in automating certain processes. While this is interesting, it doesn't seem to get at the root of dyslexia.
Proceedings of the Royal Society B, 2017, DOI: 10.1098/rspb.2017.1380