Computational modeling of high-level cognition and brain function

This article describes a computational modeling architecture, 4CAPS, which is consistent with key properties of cortical function and makes good conta ct with functional neuroimaging results. Like earlier cognitive models such as SOAR, ACT-R, 3CAPS, and EPIC, the proposed cognitive model is implement ed in a computer simulation that predicts observable variables such as huma n response times and error patterns. In addition, the proposed 4CAPS model accounts for the functional decomposition of the cognitive system and predi cts fMRI activation levels and their localization within specific cortical regions, by incorporating key properties of cortical function into the desi gn of the modeling system. (C) 1999 Wiley-Liss, Inc.