Abstract
A new version of the dual-system hypothesis is described. Consistent with earlier models, the improvisational subsystem of the instrumental system, which includes the occipital cortex, inferior temporal cortex, and medial temporal cortex, especially the hippocampus, directs the construction of visual representations of the world and constructs ad-hoc responses to novel targets. The habit system, which includes the occipital cortex; parietal cortex; premotor, supplementary motor, and ventrolateral areas of frontal cortex; and the basal ganglia, especially the caudate nucleus, encodes sequences of actions and generates previously successful actions to familiar targets. However, unlike in previous dual-system models, human cognitive activity involved in task performance is not exclusively associated with one system or the other. Rather, the two systems make it possible for people to learn a variety of skills that draw on the competencies of both systems. The collective effects of these skills define human cognition. So, in contrast with earlier versions of the dual-system hypothesis, which identified the habit system solely with procedural learning and implicit improvements in task performance, the model presented here attributes a direct role in declarative-memory tasks to the habit system. Furthermore, within the model, the computational competencies of the two systems are used to construct purposeful sequences of actions—that is, skills. Human cognition is the result of the performance of these skills. Thus, voluntary action is central to human cognition.
Original language | English (US) |
---|---|
Pages (from-to) | 2192-2216 |
Number of pages | 25 |
Journal | Attention, Perception, and Psychophysics |
Volume | 81 |
Issue number | 7 |
DOIs | |
State | Published - Oct 1 2019 |
All Science Journal Classification (ASJC) codes
- Language and Linguistics
- Experimental and Cognitive Psychology
- Sensory Systems
- Linguistics and Language
Keywords
- Attention: Interactions with memory
- Attention: Neural mechanisms
- Attention: Theoretical and computational models