Center for Cognitive Brain Imaging

at Carnegie Mellon University


4CAPS Cognitive Neuroarchitecture

4CAPS is a cognitive architecture whose models can account for both traditional behavioral data and, more interestingly, the results of neuroimaging studies; in this sense it is a neuroarchitecture (Just & Varma, 2007). Cognitively speaking, it is a hybrid architecture that combines symbolic and connectionist mechanisms in a resource-constrained environment. Cortically speaking, it moves beyond the localism of most neuroscience accounts, proposing that thinking is a network phenomenon. 4CAPS is particularly well-suited for specifying models of high-level forms of cognition.

This research has been supported by the Office of Naval Research Grant N00014-02-1-0037 and the Multidisciplinary Research Program of the University Research Initiative (MURI) Grant N00014-01-1-0677.

Operating Principles of 4CAPS

4CAPS embodies six operating principles that specify the nature of cognitive and cortical information processing. An initial principle is intended to capture the current consensus of the field.

  1. Thinking is the product of the concurrent activity of multiple brain areas that collaborate in a large-scale cortical network.
  2. Each cortical area can perform multiple cognitive functions, and conversely, many cognitive functions can be performed by more than one area.
  3. Each cortical area has a limited capacity of computational resources, constraining its activity.
  4. The topology of a large-scale cortical network changes dynamically during cognition, adapting itself to the resource limitations of different cortical areas and to the functional demands of the task at hand.
  5. The communications infrastructure that supports collaborative processing is also subject to resource constraints, construed here as bandwidth limitations.

  6. Finally, we propose a measurement assumption that enables our theoretical constructs to make contact with neuroimaging data.

  7. The activation of a cortical area as measured by imaging techniques such as fMRI and PET varies as a function of its cognitive workload. The reader interested in the technical details of how these principles are realized within a hybrid symbolic-connectionist architecture is directed to Just and Varma (2007).

Just, M. A., & Varma, S. (2007). The organization of thinking: What functional brain imaging reveals about the neuroarchitecture of complex cognition. Cognitive, Affective, & Behavioral Neuroscience, 7, 153-191.

Varma, S. (2017). The CAPS family of cognitive architectures. In S. E. F. Chipman (Ed.), The Oxford Handbook of Cognitive Science. New York: Oxford University Press. doi: 10.1093/oxfordhb/9780199842193.013.002

Click the links below for more 4CAPS information:
Source Code & Documentation