From the New York Times:
The effect goes well beyond letter recognition. In a study that followed children in grades two through five, Virginia Berninger, a psychologist at the University of Washington, demonstrated that printing, cursive writing, and typing on a keyboard are all associated with distinct and separate brain patterns — and each results in a distinct end product. When the children composed text by hand, they not only consistently produced more words more quickly than they did on a keyboard, but expressed more ideas. And brain imaging in the oldest subjects suggested that the connection between writing and idea generation went even further. When these children were asked to come up with ideas for a composition, the ones with better handwriting exhibited greater neural activation in areas associated with working memory — and increased overall activation in the reading and writing networks.My own thought is that writing changes the brain's computational architecture, establishing a matrix for new calculations. See papers David Hays and I have written: Principles and Development of Natural Intelligence (1987, on the brain's computational architecture), and The Evolution of Cognition (1990, on speech, writing, calculation, and computing).
It now appears that there may even be a difference between printing and cursive writing — a distinction of particular importance as the teaching of cursive disappears in curriculum after curriculum. In dysgraphia, a condition where the ability to write is impaired, usually after brain injury, the deficit can take on a curious form: In some people, cursive writing remains relatively unimpaired, while in others, printing does.