1. There is a double association between the skills that high-IQ people with nonverbal learning disabilities do well and the skills that low-IQ savants can do. On the one hand are explicit skills that are either fundamentally verbal, or symbolic and easily verbalizable (like the skills learned in an English class or explicit deductive reasoning, for instance). On the other hand are implicit musical, artistic, spatial, time perception, arithmetic calculation, calendar calculation, or size estimation skills.
Why this division? Why do some skills go together and not others? Why is it that among gifted students who took the SAT, almost all were strongly skewed in favor of either verbal or math, and very few were equally good at both? There has to be a brain basis for why some of these skills go together, but not others, and why these tendencies to math-spatial or verbal skills tend to appear at an early age.
2. The key has got to be in the parietal lobe, at least regarding the nonverbal functions. Spatial, attentional, time-related, magnitude, size estimation, mental rotation, and such skills have all been associated with the parietal lobe, often with great overlaps in the regions activated (Sound familiar? These are pretty similar skills to the ones where savants excel). In general, similar types of computation tend to be performed near each other, so these various nonverbal functions should overlap in space because of some overlap in function.
3. What is that functional overlap, though? What exactly are all these component processes in the parietal lobe that so much resemble savant skills/nonverbal learning disability deficits, and what computations do they all have in common? Why are they all centered in the parietal lobe? Do they tend to be wiped out together in cases of brain damage because they're all next to each other, because they're functionally related in a network, or both? (Presumably both, as functionally similar areas tend to be located near each other, but it's not a given).
4. What makes some people develop strengths in nonverbal parietal-centered abilities and others develop weaknesses? (One would probably have to do a longitudinal study where people with nonverbal learning disability are matched with typical people on verbal IQ, and savants are matched with typical people on nonverbal IQ, and all three are compared both behaviorally and with neuroimaging).
5. It's pretty clear that to understand nonverbal strengths and weaknesses, we need to look to the parietal lobe. But what about verbal abilities (the language network seems to be distributed all over the brain, with short-range connections diminishing and long-range connections increasing over development)? There may be parietal involvement here too, as a parietal region is involved with integrating phonology and orthography, and the parietal lobe is considered part of a network relevant for (explicit) intelligence of the sort measured on IQ tests (the P-FIT model). My guess is verbal abilities are probably more distributed in location than nonverbal ones, and explicit IQ even more so, but why should that be? And again, what would that say about the relationship between verbal and nonverbal ability, and why they so often trade off?
Not an easy set of questions to answer, as the parietal lobe is full of areas that activate for many different kinds of tasks, with very subtle differences between coordinates and lots of room for differences of interpretation. One really needs to compare across experiments to understand what's going on, but that requires comparing not only hard to visualize coordinates (in different numbering systems), but also procedures and statistical thresholds...not to mention, taking into account the overall task-related network in which each activated region appears...in short, information overload. So, in spite of finding, and starting to work my way through, a bunch of fascinating papers on the parietal lobe and its functions, I haven't made much progress on thinking through these questions.
Also, these questions (and the reading I would need to do to address them) have only the most tangential relationship to my homework, which I should really be doing instead...