Communicated by Jeffrey Elman Parsing Complex Sentences with Structured Connectionist Networks

A modular, recurrent connectionist network is taught to incrementally parse complex sentences. From input presented one word at a time, the network learns to do semantic role assignment, noun phrase attach-ment, and clause structure recognition, for sentences with both active and passive constructions and center-embedded clauses. The network makes syntactic and semantic predictions at every step. Previous pre-dictions are revised as expectations are confirmed or violated with the arrival of new information. The network induces its own "grammar rules " for dynamically transforming an input sequence of words into a syntactickernantic interpretation. The network generalizes well and is tolerant of ill-formed inputs.