http://symsys.stanford.edu/viewing/event/21495 Curiously I came immediately from a group at HP whose topic was technologies to simplify programming for parallel distributed applications. Three of the four speakers are among the five authors of this paper: http://www-psych.stanford.edu/~jlm/papers/McCEtAlIP_GazzVolChapter.pdf Parallel Distributed Processing (Approach to Semantic Cognition) Jay McClelland Buzz words and phrases: ‘propagation of activation’ The ‘silly’ methodology for discovering conceptual trees. You would ask a child of some known age whether it would be silly to say that a chair could swim. This allows recording the development of categories in children. The Rumelhart Model Do children need words for concepts in order to have concepts? (personal observation: but they are pleased to learn the word for a concept that they already had.) ------------ what we know about a frog is distributed to various sort specific sites in brain. The look of a frog is kept near the look of a sparrow. Semantic Dementia resulting from localized brain damage! Such people remember facts but are often unable to apply them to their own benefit. Perhaps there is a ‘analytic engine' in the head! --------- More ‘spreading activation’. I didn't see anything here that is not well covered with stereotype in the particular case the stereotype of family. Stereotype is a somewhat complex idea and decomposing it into simpler notions is good if you can do it. -------- Need for amodal ATL conductor More evidence of a super category of ‘man-made-object’. Some evidence that even monkeys know the difference. Evidence here of physical locality of brain part that applies such distinctions. www.psych-sci.manchester.ac.uk/naru/ -------- Overall I am impressed with the quality of brain architecture you can learn in the clinic. I also conclude that Kanerva's ideas are highly productive here. ----------- Bill mail: I compose this at intermission. The first two talks have been exemplary! There is much information here, new to me, that is highly relevant to how the brain actually works—experimental results with teeth that any brain theory must accommodate. Some evidence was presented for physical locality of cognitive processing of general but not specific categories. That was entirely new to me. --- Three excellent talks (so far). It is being video taped. Well worth watching. I will try to remember to rematch it too!