cSCAN Rounds

Towards a unified model of acquisition, production, and comprehension

One goal in the psychology of language is the development of a unified theory of language acquisition, production, and comprehension. Toward this goal, I will present a connectionist model of sentence-level processing. The model incorporates an implicit learning mechanism that acquires abstract syntactic representations and these representations allow the model to capture a wide range of structural priming phenomena (Chang, Dell, & Bock, 2006). These learning mechanisms are powerful enough to allow the model to learn English and Japanese representations, and to explain different incremental production biases in these languages (Chang, 2009). Finally, I will discuss a new version of this model, which incorporates spatial representations (e.g., object files, affordances) and these additions allow the model to account for visual world eye-tracking results in sentence comprehension. By linking language with implicit learning and spatial representations, it is possible to give a unified account of language development and language processing.