This is an excerpt from the Handbook of Artificial Intelligence, a compendium of hundreds of articles about AI ideas, techniques, and programs being prepared at Stanford University by AI researchers and students from across the country. In addition to articles describing the specifics of various AI programming methods, the Handbook contains dozens of overview articles like this one, which attempt to give historical and scientific perspective to work in the different areas of AI research. This article is from the Handbook chapter on natural language understanding. Cross-references to other articles in the handbook have been removed-terms discussed in more detail elsewhere are italicized. Many people have contributed to this chapter, including especially Anne Gardner, James Davidson, and Terry Winograd. Avron Barr and Edward A. Feigenbaum are the Handbook's general editors.

Births are always interesting affairs. According to some, births are always traumatic — a shock to come from the womb to the new world. The birth we give witness to here is that of a new society, the American Association for Artificial Intelligence — AAAI. It has not seemed to me traumatic, but rather almost wholly benign. In a world where not much is benign at the moment, such an event is devoutly to be cherished. The proper topic for this initial message is talk about beginnings and circumstances, goals and aims, character and style. My premier duty as president of AAAI, it appears, will be to give a presidential address at the upcoming annual meeting. Specific precedents being absent, I need to give thought to what belongs in an AAAI presidential address. But one thing I already know: That talk should be devoted to our science, not our society. It should be substantive , not procedural. It should look inward at the state of what we know about intelligence and computers, not outward at our place in the larger society. It is in this message that earthly matters belong.

This paper describes a generative theory of bugs. It claims that all bugs of a procedural skill can be derived by a highly constrained form of problem solving acting on incomplete procedures. These procedures are characterized by formal deletion operations that model incomplete learning and forgetting. The problem solver and the deletion operator have been constrained to make it impossible to derive “star-bugs”—algorithms that are so absurd that expert diagnosticians agree that the alogorithm will never be observed as a bug. Hence, the theory not only generates the observed bugs, it fails to generate star-bugs.

In Lea, W, (Ed.). Trends in speech recognition. Englewood Cliffs, N.J.: Prentice-Hall, 1980, 39-89

The performance of a voice- and touch-driven natural language editor is described as subjects used it to do editing tasks. The system features the abilities to process imperative sentences with noun phrases that may include pronouns, quantifiers and references to dialogue focus. The system utilizes a commerical speaker-dependent connected-speech recognizer, and processes sentences spoken by human subjects at the rate of five to seven sentences per minute. Sentence recognition percentages for our expert speaker and for subjects, were 98 and around the mid 70s, respectively. Subjects had more difficulty learning to use connected speech than had been the case in earlier experiments with discrete speech.

This paper reviews the history of process-dependent reasoning in AI systems, and argues that it represents an essentially different approach to non-monotonic reasoning from other formalizations. Much of the paper is a basic level tutorial, explaining the issues and providing a framework for understanding the essential features of non-monotonic reasoning.

'Non-monotonic' logical systems are logics in which the introduction of new axioms can invalidate old theorems. Such logics are very important in modeling the benefits of active processes which, acting in the presence of incomplete information, must make and subsequently revise assumptions in light of new observations. We present the motivation and history of such logics. We develop model and proof theories, a proof procedure, and applications for one non-monotonic logic. In particular, we prove the completeness of the non-monotoic predicate calculus and the decidability of the non-monotonic sentential calculus. We also discuss characteristic properties of this logic and its relationship to stronger logics, logics of incomplete information, and truth maintenance systems. Artificial Intelligence 13:41-72.

This paper describes a model of cooperative behavior and describes how such a model can be applied in a natural language understanding system. We assume that agents attempt to recognize the plans of other agents and, then, use this plan when deciding what response to make. In particular, we show that, given a setting in which purposeful dialogues occur, this model can account for responses that provide more information that explicitly requested and for appropriate responses to both short sentence fragments and indirect speech acts.

In this paper we explore the number of tree search operations required to solve binary constraint satisfaction problems. We show analytically and experimentally that the two principles of first trying the places most likely to fail and remembering what has been done to avoid repeating the same mistake twice improve the standard backtracking search. We experimentally show that a lookahead procedure called forward checking (to anticipate the future) which employs the most likely to fail principle performs better than standard backtracking, Ullman's, Waltz's, Mackworth's, and Haralick's discrete relaxation in all cases tested, and better than Gaschnig's backmarking in the larger problems.

This paper explores the truism that people think about what they say. It proposes that, to satisfy their own goals, people often plan their speech acts to affect their listeners' beliefs, goals, and emotional states. Such language use can be modelled by viewing speech acts as operators in a planning system, thus allowing both physical and speech acts to be integrated into plans. Methodological issues of how speech acts should be defined in a plan-based theory are illustrated by defining operators for requesting and informing. Plans containing those operators are presented and comparisons are drawn with Searle's formulation.