CONCEPT
State-Space Semantics
Paul Churchland’s thesis that the brain represents the world not in sentences but in vectors—patterns of activation distributed across neuron populations, where meaning lives in geometric position within a high-dimensional space and cognition is the transformation of those positions through learned networks.
When you smell coffee, no inner word “coffee” is written anywhere in your brain. Instead, a large set of olfactory neurons fires, each at its own rate, and the particular profile of all those firing rates taken together is the brain’s representation of that smell. A different smell is a different profile. Meaning lives in the pattern, and the pattern is a point in a space whose dimensions are the neurons. This is
Paul Churchland’s state-space semantics in its simplest form, and the geometry is its heart: a population of a thousand neurons defines a thousand-dimensional space, every possible activation pattern is a location in that space, similar things sit near one another, different things sit far apart, and to learn a domain is to sculpt the space so that its structure mirrors the structure of the world being represented. Concepts are not definitions stored as sentences; they are prototype-shaped regions of activation