Preschool children often show confusion about letters of the alphabet. They write them backward or upside-down and don't seem to care about the differences we point out between their productions and the standard version. As they get older, some children continue to make the same mistakes, and their efforts to read are confused as they ignore the order of letters or even of words. After a few years of schooling, these children's continuing difficulties with reading are often classified as dyslexia. Dyslexic children may get behind in other subjects as well, because they are expected to gain knowledge by the very reading skills they are struggling to develop. It's no wonder that so many efforts have been made to understand dyslexia and to treat it effectively.
In the past, a common explanation offered for dyslexia was that these poor readers "saw backwards" and therefore reversed letters both when they read and when they wrote. It's hardly possible that people would "see backwards" only when looking at print, and if they "saw backwards" all the time it would certainly be apparent in their daily behavior. For example, if someone's vision reverses left and right, we'd expect that person to reach to the left for a doorknob that was actually on the right, or to find it impossible to look at and button a shirt with buttons on the left and holes on the right. But we don't see people with these problems.
In addition, logic tells us that if someone always saw left and right reversed, he or she could still learn to recognize letters. If such a hypothetical person wanted to write a letter, he or she would make it look correct to the rest of us, because only in that way could it appear in its normal or "reversed" position from the reverser's viewpoint. (This argument is similar to the logic that refutes the "El Greco fallacy" in the study of art history.) Consistent reversal of shapes would make no difference to reading or writing, and a person who experienced "flips" from one orientation to another would surely mention this or show it behaviorally.
All this suggests that dyslexia probably has little to do with vision itself. But learning to read involves another critical sense: hearing. A first step in reading is to associate the shape of a letter (grapheme) with the sound or sounds it represents (one or more phonemes). A child who has difficulty in processing speech sounds will also have trouble associating those sounds with shapes.
One of the difficulties in processing and recognizing speech sounds is that these are not simple patterns of sound waves--- even though they seem that way to a listener. We hear a speech sound as if it were a unit of stimulation, quickly begun, quickly finished, and of uniform characteristics from beginning to end. However, a speech sound involves complex patterns of sound waves, often varying a great deal from onset to finish. These sound wave changes happen extremely fast, and the auditory areas of the brain have to be very quick to respond to them. If the response is slow, the sound is all over before processing is well under way. A person who is abnormally slow in processing speech sounds could be expected to have trouble connecting the poorly-perceived sound with a specific letter of the alphabet. (Incidentally, such a person would have "normal hearing" in the sense of being able to detect very soft sounds; the problem would be not in detecting the sounds but in identifying them.)
In a recent article in Science (Gabrieli, J.D.E., "Dyslexia: A new synergy between education and cognitive neuroscience". Vol. 325, pp. 280-283), John Gabrieli described some of the difficulties dyslexic children have with "phonological awareness"-- the ability to detect and identify not only different sounds but the order they appear in. For example, they have trouble listening to words and deciding which ones start with the same sound (does "hat" start with the same sound as "hot" or as "bat"?). This shows that the problem with reading is not simply a matter of deciding what a letter looks like; effective listening needs to be achieved before reading can be done fluently. However, Gabrieli also pointed out some research showing subtle differences in dyslexics' visual abilities, although no problems that would appear on any ordinary test of vision.
Gabrieli and some co-workers have shown that a computer-game approach can help dyslexic children improve their ability for rapid auditory processing. However, it is not clear how directly this change contributes to improved reading. Specific instruction can be helpful, but it is expensive and should be directed to children who are most at risk for reading problems. An important task for the future is to develop methods of identifying these children early and intervening before they are far behind educationally.