A finite and real-time processor for natural language
People process natural language in real time and with very limited short-term memories. This article describes a computational architecture for syntactic performance that also requires fixed finite resources.