Computational Neurolinguistics - What Is It All About?

Abstract

Computational Neurolinguistics (CN) integrates art ifi cia 1 intelligence (AI) methods with concepts of neurally motivated processing to develop cognitive models of natural language processing. HOP~ is on~ ex amp 1e of a mode 1 deve 1oped to addres~ ~ssues 1n CN. The model is parallel, and exemp~1f1es language as the result of time synchronlzed processes which are asynchronous in nat~re. Furt_hermore, the model is substantially val1dated to 1nclude normal behavioral evidence in its design. In addition, it attends to aspects of 1anguage breakdown which are we 11 documented in the literature of neurolinguistics or aphasia. This paper discusses assumptions which und~rlie the_ CN approach to model development. It w1ll descr1 be the neurally motivated or "natural computational" processes which produce the model's observable and verifiable behavioral results. The differences in the CN approach to other models of parallel memory process and behavior will be presented . Finally, the contribution of the CN research approach as a tool for investigating the b~eakdown ?f l_anguage performance and its potentlal contr1but1on to understanding brain function will be discussed.