Natural English as a query language offers the ultimate in directness for man-machine communications and has new and powerful potentials for interactive problem solving. The linguistic problems underlying machine interpretation and generation of natural English, however, have been slow in yielding to solution, though recent successes have raised new expectations. Speakers in the Query Language Session will explore these potentials and problems and discuss their approaches to the use of natural English. Two opening presentations by Dr. Vincent Giuliano and Dr. Christine Montgomery will set the theme of the session, and a panel of experts will then discuss their positions on the question.Natural English as a query language can provide interactive communication between men and computers of a kind and quality previously only possible between humans. Natural language is thought to be optimally coupled to the abilities and limitations of human information processing. Further, a machine that can understand and answer questions in natural English comes close to passing Turing's classic test of machine intelligence.However, natural English is not always formal English, and even when it is, the meaningful and the grammatical cannot be equated with any assurance. Nor can the strictures of a “constranied English” avoid the transforming colorations of context inherent in natural language. The very redundancy of English which permits ill-formed expression to convey clear messages to humans too often has resulted in a linguistic and programming quagmire of uncertainty and peril.Despite the considerable problems in using natural English as a query language, certain operational successes of significance have been achieved by session participants and by others, and these experiences will be discussed, as will recent advances in computational linguistics, that now show promise of closing the linguistic gap between form and meaning.