May 09, 2010 11:46
I wonder what would happen if you developed an expressively-complete language for real numbers using color-shape pairs, probably at least partially by working with people who have a synesthesic ability to multiply large numbers in essentially constant time in their heads already, and tried teaching it to young children alongside the normal symbolic representations we normally use.
Would the children internalize the representation you taught them, and thereby be able to do truly hilarious feats of math using the massively parallel hardware we have baked in? Or would they fail to internalize it, and just do it mechanically like so many other tasks.