Related Experiment Videos
Deictic codes for the embodiment of cognition.
D H Ballard1, M M Hayhoe, P K Pook
1Computer Science Department, University of Rochester, NY 14627, USA. dana@cs.rochester.edu
The Behavioral and Brain Sciences
|March 31, 1999
Summary
Computational brain models use deictic computation, linking body movements to cognitive programs at specific time scales. This "embodiment level" approach explains how pointing actions enable natural task performance and working memory.
Area of Science:
- Cognitive Science
- Computational Neuroscience
- Embodied Cognition
Background:
- Brain models require multiple abstraction levels for diverse time scales.
- Orienting body movements are crucial at ~1/3 second time scales for cognition.
- The "embodiment level" bridges natural phenomena and high-level cognitive processes.
Purpose of the Study:
- To explore the role of deictic computation in linking sensory data, cognitive programs, and motor actions.
- To explain how deictic binding facilitates natural task execution.
- To relate working memory to body feature dispositions.
Main Methods:
- Modeling computational processes at the "embodiment level" (~1/3 second time scales).
- Analyzing the synergy between body movement sequentiality and decision systems.
- Investigating deictic reference systems for binding external objects to internal programs.
Main Results:
- Deictic binding, using pointing movements, links external objects to cognitive programs.
- This mechanism enables the performance of natural tasks by matching body movements to computational economies.
- Working memory is associated with dynamic body features like eye and hand movements.
Conclusions:
- Deictic computation offers a mechanism for integrating sensory input with cognitive and motor outputs.
- The "embodiment level" provides a valuable framework for understanding cognition at specific time scales.
- Body movements and deictic referencing are fundamental to cognitive operations and working memory.