Embodied AI Glossary中文

Embodied Cognition

具身认知Advanced

The view in cognitive science that thought is inseparable from the body and its interaction with the environment.

Embodied cognition is a family of theories in cognitive science and philosophy of mind holding that thought is not just symbol manipulation happening in the brain, but is shaped jointly by the body's form, its sensorimotor abilities, and the body's interaction with the environment. The idea traces back to the phenomenology of Maurice Merleau-Ponty and others; Francisco Varela, Evan Thompson, and Eleanor Rosch's 1991 book The Embodied Mind systematized it, while George Lakoff and Mark Johnson argued from a linguistic angle that abstract concepts derive from bodily experience. Its influence on robotics has been direct: Rodney Brooks abandoned the approach of building a complete world model before planning, letting sensors drive behavior directly and arguing that the world itself is the best model; Rolf Pfeifer and others likewise argued that genuine intelligence requires a body with sensorimotor capabilities. Today's term “embodied AI” descends from this line of thinking, and it provides the theoretical basis for the claim that intelligence must form through interaction with the physical world.

ExampleEnglish uses “grasp” to mean “understand”; Lakoff and Johnson treat expressions like this as evidence that abstract concepts borrow from bodily experience.

Also called
Embodiment (Cognitive Science)
Related
Embodied AI · Disembodied AI · Subsumption Architecture · Morphological Computation · Symbol Grounding Problem · Moravec's Paradox
Sources
Embodied Cognition (Stanford Encyclopedia of Philosophy)
Embodied cognition (Wikipedia)

See it in the full glossary →