Here's a little more information on this subject. I have a relative who is Professor Emeritus of Education at the University
of Oregon, and he does research in is "spare" time on how the brain learns. Anyway, here's his take.
---
This past week I've received at least a half dozen similar versions of the Cambridge University language research discovery (above), so it's obviously scooting around the Internet. Since I think I understand the neurobiological base of the research, and am in the initial process of developing one of my future monthly Brain Connection columns on it (
http://www.brainconnection.com/sylwester), I thought I would provide a little background information on the discovery to friends who might have similarly received the Internet version and are curious about it.
The next time you're in a public library, you might want to look at the July 5 edition of the British journal, New Scientist, which does a fine job of explaining scientific developments in non-technical language. Read Stanislas Dehaene's "Natural Born Readers" (pages 30-33). You can also probably read it online even as a non-subscriber via
http://archive.newscientist.com Stan Dehaene is a French neuroscientist best known for his work on figuring out the underlying neurobiology of mathematics. His 1997 book The Number Sense is the classic in the field.
The point of this line of research is that proficient readers read most words (except very long ones) principally through the first and last letters and a rapid recognition of the general shape and content of the entire word. That's why we typically read misspelled words with no trouble.
We can also easily read words in different fonts (including fonts we hadn't seen before), words with mixed capital and small letters, manuscript and cursive, etc. We also easily distinguish between words like eight and sight that have only one subtle spelling difference but are pronounced very differently.
Written language is processed in about a dozen different brain areas (and principally in the left hemisphere in most people). Written language doesn't have the innate elements that spoken language has, and so is learned with more difficulty. The brain structures we use to recognize words also exist in the visual systems of primates, where they are used to recognize important objects -- dangers and opportunities.
For example, a monkey must be able to quickly recognize a lion, but although each lion is slightly different from other lions, all have a characteristic lion face, and it is that configuration (a sort of cartoon lion with only the essential details) that the monkey's object recognition system tunes into. Further, the lion might be looking up/down/sideways/etc, so the object recognition system has to be able to recognize the conceptual lion in any possible positional variation.
Dehaene argues with considerable research support that we simply recycled this existing system when written language entered into human culture. For example, draw a capital A with the horizontal line extending a bit beyond the diagonal lines, and place a couple dots above the horizontal line within the triangle. Invert the drawing and it's a cartoon bull. Our letter A derives from the Greek alpha, which emerged out of the ancient Semitic word for bull (alf) -- and our brain has no difficulty in recognizing A whether it's right side up, upside down, or sideways. The rest of our alphabet has similar ancient beginnings.
Our alphabet is made up of combinations of vertical, horizontal, diagonal, and curved lines, which our brain's visual system easily recognizes -- alone and in combination. Rapid object recognition depends more on the outer shape of something than on the inner (often quite variant) details -- and so what works for an object also works nicely for a word with two outer and several inner letters.
By connecting innately meaningless sounds to innately meaningless lines within our incredibly inventive brain, we developed a marvelously meaningful language-driven civilization.
But alas, we didn't follow up on this brain property and invent a computer that could similarly recognize the external essence of an email address and ignore incorrect internal letter sequences.
Bob Sylwester