A mechanistic model of three facets of meaning

摘要

This chapter presents a physical-computational model of sensorimotor grounded language interpretation for simple speech acts. It discusses that the interpretation of directive and descriptive speech acts consists of translating utterances into updates of memory systems in the controller. It explains that the same memory systems also mediate sensorimotor interactions, thus serving as a cross-modal bridge between language, perception, and action. It suggests that a viable strategy for modelling language use is to focus on simple language use, such that of young children. It defines three facets of ‘meaning’ that need to be explained, leading to functional specifications for a model of language use framed as semiotic processing. It describes the embodiment behavior of Ripley, a conversational robot built in the laboratory that serves as a concrete launching point for a more general model.