Research in mathematical cognition has primarily focused on how basic numerical information (like number magnitude) is represented in the brain and how this information is used in simple tasks (like number comparison and mental arithmetic). How the brain subserves more complex forms of mathematics remains largely unknown. To address this issue, we go back to what mathematics essentially is about: a symbol system that allows the description of patterns  in an abstract manner. We will use fMRI to uncover the neural systems involved in the processing of patterns. We will present visual stimuli that consist of regular patterns. We will manipulate the level of abstraction of the regularity, ranging from the visual feature that defines regularity (e.g. color) to symbolic descriptions of a regular pattern. We will investigate which brain regions are sensitive to within-feature regularities, which regions are sensitive to abstract regularities. A posterior to anterior gradient is predicted.