Automatic Inference of Semantic Deep Structure Rules in Generative Semantic Grammars

Abstract

Techniques and methodology for the automatic inference of semantic deep structure rules in generative semantic grammars. The key conceptual devices include a representation of semantic deep structure in the notation of a 4-dimensional network with properties of at least the 2nd-order predicate calculus, and also in the notation of a compiler-driven behavioral simulation language that describes and modifies the linguistic and extra-linguistic conceptual universe of speakers. The system is able to make grammatical-semantic inferences within the frameworks of all current generative semantic linguistic models, including the case grammar of Fillmore, the presuppositional model of Lakoff, and the 1972 semantic theory of Katz.