Putting Meaning into Grammar Learning

This paper proposes a formulation of grammar learning in which meaning plays a fundamental role. We present a computational model that aims to satisfy convergent constraints from cognitive linguistics and crosslinguistic developmental evidence within a statistically driven framework. The target grammar, input data and goal of learning are all designed to allow a tight coupling between language learning and comprehension that drives the acquisition of new constructions. The model is applied to learn lexically specific multi-word constructions from annotated child-directed transcript data.