Neuroimaging has been used extensively to study cognitive, emotional, motivational, and social functions in humans. The various tasks target different levels of analysis: some focus on elementary cognitive operations, while others study higher-level processes (e.g., self-regulation, social behavior, and reasoning). Is there a hierarchical relationship among ‘basic’ and ‘complex’ processes? In this talk, I describe several fMRI experiments aimed at understanding two of the building-blocks of basic cognitive control, shifting attention and resolving interference. I examine brain-imaging evidence for how they may be both components of higher-level processes, and how they themselves may be decomposed into parts. I then introduce the concept of diagnostic value for brain imaging data, and use meta-analytic neuroimaging data to examine what brain imaging activations from these tasks can tell us about the structure of thought and feeling.