To say it in more plain words, "spreading activation" describes the activity through a weighted graph, that could be a good, old-fashioned, semantic network (bu

To say it in more plain words, "spreading activation" describes the activity through a weighted graph, that could be a good, old-fashioned, semantic network (but with weights) or a neural net. It's fundamental and foundational connectionist stuff and for once I'm not using the term dismissively. It seems like spreading activation goes to the heart of connectionist ideas about storing knowledge in a graph, what used to be known as a Parallel Distributed Process model, apparently.