A Computational Theory of the Mental Imagery Medium
Steven A. Pinker · Oxford University Press eBooks · 2013
Does it make sense to think of a mental image as a picture in the head? And if it does, how are three-dimensional objects and scenes represented in mental “pictures”? This chapter suggests that these questions can be answered using the theoretical principles of computational cognitive science, analyzing the imagery system in terms of the data structures and processes that manipulate information during imaginal thinking. In particular, it outlines a theory in which images are patterns of activation in a 3D array of cells, accessed via two overlaid coordinate systems: a fixed viewer-centered spherical coordinate system, and a movable object-centered or world-centered coordinate system. By inserting information into the array using one system, and accessing it using the other, a variety of 3D spatial information processes, such as generating, inspecting, and transforming images, recognizing shapes, and at tending to locations, can be handled within a single framework.