There are several approaches to the study of cognitive science. These approaches may be classified broadly as symbolic, connectionist, and dynamic systems.

Symbolic systems - holds that cognition can be explained using operations on symbols, by means of explicit computational theories and models of mental (but not brain) processes analogous to the workings of a digital computer. Connectionist (subsymbolic) - holds that cognition can only be modeled and explained by using artificial neural network on the level of physical brain properties.

Hybrid systems - holds that cognition is best modeled using both connectionist and symbolic models, and possibly other computational techniques (see Sun and Bookman 1994).

Dynamical Systems - holds that cognition can be explained by means of a continuous dynamical system in which all the elements are interrelated, like the Watt Governor.