Alternative TL;DR: The experimental methods you use can color the results you get.

E.g., lesioning different parts of a microprocessor may reveal "clusters of function" much as such studies do in the brain.