For example, "John loves Mary" is an unstructured activation pattern that can be traced back to earlier causal effects of "John" "loves" and "Mary" activation patterns, and perhaps even further back to microfeatures of those.
By virtue of causal history, unstructured connectionist representations can account for systematicity and related phenomena.
So, Fodor & Pylyshyn are right to claim that connectionist representations are unstructured, but wrong to claim that such representations can't account for systematicity.
Keith Butler, 1991. |