🧪 Linguistics Semantics Simulation

This simulation builds a small hierarchical semantic network — living things, animals and plants, their subcategories, individual concepts like canary and shark, and the properties attached to them — and renders it in 3D exactly as Collins & Loftus's classic spreading-activation model of meaning describes it. Pulse any concept and watch activation leak across is-a and property links to its neighbours, decaying over time, while live readouts track how many nodes are active, the network's total activation energy, and how long the "wave" took to reach each node from the last pulse. Because atypical members (an ostrich cannot fly) are wired without the direct property shortcut typical members get, the simulation reproduces real psycholinguistic findings — canary primes bird faster than animal, and typical exemplars activate their properties faster than atypical ones — as an emergent property of graph structure rather than a scripted animation.