Embodied cognition is the view that intelligence is not just abstract symbol processing carried out in a head, but something that depends on having a body, a sensory apparatus, and an ongoing physical engagement with the world. On this view, much of what looks like reasoning is actually offloaded to the body and the environment, and a system cut off from sensing and acting is missing something essential rather than merely lacking a peripheral.
The idea has older roots in phenomenology and in Hubert Dreyfus’s critique of symbolic AI, but its sharpest engineering statement came from the roboticist Rodney Brooks. In papers such as “Intelligence Without Representation,” hosted on his MIT site, Brooks argued against the dominant approach of building a rich internal model of the world and reasoning over it. His slogan was that “the world is its own best model”: rather than maintaining an internal map, a robot can sense the relevant feature of its surroundings directly, the moment it needs it. He built insect-like robots whose competent behavior emerged from layers of simple sense-and-act rules with no central representation at all.
Embodied cognition reframes the symbol grounding problem from the other direction: instead of asking how to attach meaning to symbols after the fact, it starts from a system already coupled to the world and asks how much intelligence can be built up from that coupling. The approach gave rise to behavior-based robotics and influenced how researchers think about perception, action, and the role of the environment in cognition.
Why business readers should care: embodied cognition is the intellectual backdrop to modern robotics and to skepticism about whether a system trained only on text, with no body or sensory experience, can ever fully understand the physical world it talks about.