10 - Language and communication

Summary

Introduction

Language and communication, considered as relevant to artificial intelligence (AI) in general, I take to refer to the issues that cluster round the representation of language and meaning so as to enable language processing and the communication of meaning by a computer, an area of research roughly captured by the fields of Natural Language Processing (NLP) and Computational Linguistics (CL).

A remarkable feature of the fifty-year history of those two related fields is how much of what we now take as topics of current interest was there from the very beginning; all the pioneers lacked were real computers. In the fifties and sixties, Gilbert King was arguing for machine translation by statistical methods, which is only now a reality, Margaret Masterman for the power of meaning-based structures in programs, and Vic Yngve, still working at the time of writing, had designed COMIT, a special programming language for NLP, and had stated his famous claim that limits on the way computers process language should reflect the way the syntax of a language is structured. This last project brought Yngve into direct conflict with Noam Chomsky over permissible ways of drawing syntactic tree structures, which can now be seen to have constituted a defining moment of schism in the history of NLP in its relationship to mainstream linguistics. Chomsky has always denied any relevance of computation to the understanding of language structure, and this foundational schism was not healed until decades later, when Gerald Gazdar became the first major linguist to embrace a computational strategy explicitly.