What about a Savant? They don't understand themselves how they get the "right" answer.
I think that's why they study savants so much--it's sort of like the two-slit experiment.
I presume one is referring to an 'autisitic savant' as a opposed to a learned person, well versed in literature or science as well as the fine arts, often with an exceptional skill in a specialized field of learning.
Autisitic (or perjorativley, idiot) savant - 1 : a person affected with a mental disability (as autism or mental retardation) who exhibits exceptional skill or brilliance in some limited field (as mathematics or music) —called also savant.
2 : a person who is highly knowledgeable about one subject but knows little about anything else.
The second definition indicates the limitations of those with autistic savant syndrome. They may know in a factual sense but have limited understanding, and quite often know or understand little outside of a limited knowledge base. I believe that in general, savants have poor social skills.
As rewebster indicated, autisitic savants are studied to better understand the brain (or cerebral cortical) structure to understand how and why they function.
My statement was intended to emphasize the learning of how and why as opposed to simply substituting (plugging in) numbers for variables and getting an answer. I was reflecting on some comments and apparent expectations in the Homework forums.
These savants can get the "right answer," but they need to constantly be told what the "questions" are.
Understanding how you get the right answer allows you to take the next step and solve more difficult problems that require variations in the solution method; and then to synthesize several methods of solutions into unique applications.
Otherwise, you are just an extension of your calculator.
Give the person a fish...blah blah blah.
In fact, while studying science or engineering, the answer itself often counts for very little compared to the work needed to get the answer.
I don't know what it's like in college today, but my professors would often design tests so that a calculator either wouldn't be of use, or in a few rare cases a calculator could even prevent you from getting the right answer. I remember acing out the rest of the class at the end of my lower div physics courses this way. Only by working the problem by hand did one see that approximations could be used several times to arrive at the correct solution. Those using pre-programmed algorithms trailed off into never-never land and got stuck.
Separate names with a comma.