NPR logo To Read, But Not To See


To Read, But Not To See

Let's talk about reading.

You enter a room. Suppose there is a bit of graffiti on the wall. It is ugly and racist, let's say. You see it and your heart pounds, heat rises to your face. You are disgusted. You are insulted.

But not your companion. It is written in a language she cannot read. She may very well see the letters, the inscription on the wall. But she is immune to injury. Not because she's tougher than you. But because her inability to read blinds her to what is there. Her retinal image is no different from yours, but she is blind. Seeing depends on so much more than retinal images!

She can see the marks on the wall, but she can't read them. She can see them, but not their meaning.

Consider this text for example (from O'Regan and Noë, Behavioral and Brain Sciences, 24/5: 2001):

The illusion of

Of reading

If you are literate, you probably had no difficult whatsoever understanding this. But you are unlikely to have noticed the second occurrence of "of" on the second line. Your inability to notice this second "of" is a mark of your competence, not your incompetence. We aren't interested in typos or letters or marks when we read. We are interested in meaning. Your ability to see meaning requires you to stop paying attention to letters.

Your companion could see the marks of the graffiti, but not the meaning. You can perceive the meaning, but not the marks!

Scientists have demonstrated that the time it takes to read a word is not a function of the number of letters. We recognize word forms. As if words were faces. They jump out at us. It is strange to look at a face upside down. It loses its form and becomes, merely, a jumble of features. So with words. An upside-down word no longer has the form of a word. This shows that we don't read words, or faces, by paying attention to features.

Writing is a recent invention when looked at on the scale of human history. Humans have been talking at least for 50,000 years, and probably much longer. But we've been reading for no more than 10,000 years, and widespread literacy is much more recent than that. A few hundred years at most. And yet it is likely that there are neural structures (the so-called visual word form area, VFWA, located in the left fusiform gyrus) that are dedicated to reading. These areas light up when we read; damage to these areas corresponds to distinct deficits in our ability to read. (See McClandliss, Cohen and Dehaene 2003 for the evidence.)

Interestingly, a recent publication by Reich, Szwed, Cohen, and Amedi demonstrates that there is nothing really visual about our capacity to read. Blind readers of Braille also show neural activity in VFWA when they read.

But then there is nothing really visual about seeing either. We have known for a long time that blind readers of Braille show neural activity in the occipital lobe — in visual cortex — when they read Braille. Reading, like seeing, is about understanding, not marking visual features. This is paradoxical: to see well we need, in a way, to stop seeing.