ABSTRACT
In this chapter the author intends to deal with language, and concentrate mainly on it from the artificial intelligence viewpoint, since the issues raised by philosophers and semanticians are so complex that they have been the subject of a whole host of studies from various points of view. Whatever the value of attempting a context free grammar for other purposes, pragmatic considerations suggest it is wholly untenable as a simulation of human conversation and therefore almost certainly undesirable as a syntactic base for computer-man conversation. The author accepts the usefulness of syntax in the attempts to provide computer languages as approximations to natural languages. Roger Schank at Stanford has been working to develop a theory of human language understanding. It contrasts with SHRDLU in that it is more versatile and flexible, although it is less precise. His basic premise is that people understand language by mapping utterances onto a conceptual base.
