This charming study turned up today and led me back here:
'ScienceDaily (2011-12-02) -- "Face recognition is an important social skill, but not all of us are equally good at it," says a cognitive psychologist. But what accounts for the difference? A new study provides the first experimental evidence that the inequality of abilities is rooted in the unique way in which the mind perceives faces.'
The brain uses a wholistic method for recognizing faces, whereas recognizing objects is more analytical -- meaning that the object is parsed out by the brain and assessed in chunks. Doesn't work well on faces, though.
One very pleasing, intelligent thing to note: these researchers concluded that the logical next step is to train people to use their brains more effectively -- not to look for yet another drug to fiddle with one part of a complex process and leave the patient more helpless than before.
My own facial recognition has plummeted with the evolution of CRPS, but it's a skill I acquired rather late in life. I'd always recognized people by the way they held themselves and moved -- a different region of the brain entirely. So I trained myself to recognize faces as such.
I'm not sure I could recapture that ability, but when a few other mental faculties are more predictable, I'm game to try.