The economic theory of rationality promises to equal mathematical logic in its importance for the mechanization of reasoning. We survey the growing literature on how the basic notions of probability, utility, and rational choice, coupled with practical limitations on information and resources, influence the design and analysis of reasoning and representation systems. 1 Introduction People make judgments of rationality all the time, usually in criticizing someone else's thoughts or deeds as irrational, or in defending their own as rational. Artificial intelligence researchers construct systems and theories to perform or describe rational thought and action, criticizing and defending these systems and theories in terms similar to but more formal than those of the man or woman on the street. Judgments of human rationality commonly involve several different conceptions of rationality, including a logical conception used to judge thoughts, and an economic one used to judge actions or...

The long-term goal of our field is the creation and understanding of intelligence. Productive research in AI, both practical and theoretical, benefits from a notion of intelligence that is precise enough to allow the cumulative development of robust systems and general results. This paper outlines a gradual evolution in our formal conception of intelligence that brings it closer to our informal conception and simultaneously reduces the gap between theory and practice. 1 Artificial Intelligence AI is a field in which the ultimate goal has often been somewhat ill-defined and subject to dispute. Some researchers aim to emulate human cognition, others aim at the creation of

this paper, we examine Gardenfors's criticisms of the foundations approach. We argue that the coherence and foundations approaches differ less than has been supposed, in that the fundamental concerns of the coherence approach for conservatism in belief revision apply in exactly the same way in the foundations approach. We also argue that the foundations approach represents the most direct way of mechanizing the coherence approach. Moreover, the computational costs of revisions based on epistemic entrenchment appear to equal or exceed those of revisions based on reasons, in the sense that any entrenchment ordering from which information about reasons may be recovered will be at least as costly to update as the reasons it represents. We conclude that while the coherence approach offers a valuable perspective on belief revision, it does not yet provide an adequate theoretical or practical basis for characterizing or mechanizing belief revision.

This article examines prospects for theories and methods of preferences, both in the specific sense of the preferences of the ideal rational agents considered in economics and decision theory and in the broader interplay between reasoning and rationality considered in philosophy, psychology, and artificial intelligence. Modern applications seek to employ preferences as means for specifying, designing, and controlling rational behaviors as well as descriptive means for understanding behaviors. We seek to understand the nature and representation of preferences by examining the roles, origins, meaning, structure, evolution, and application of preferences.

This paper investigates several methods for coping with inconsistency caused by multiple source information, by introducing suitable consequence relations capable of inferring non-trivial conclusions from an inconsistent stratified knowledge base. Some of these methods presuppose a revision step, namely a selection of one or several consistent subsets of formulas, and then classical inference is used for inferring from these subsets. Two alternative methods that do not require any revision step are studied: inference based on arguments, and a new approach called safely supported inference, where inconsistency is kept local. These two last methods look suitable when the inconsistency is due to the presence of several sources of information. The paper offers a comparative study of the various inference modes under inconsistency. 1 Introduction Inconsistency can be encountered in different reasoning tasks, in particular: - when reasoning with exception-tolerant generic knowledge, where ...

. The problem of how to remove information from an agent's stock of beliefs is of paramount concern in the belief change literature. An inquiring agent may remove beliefs for a variety of reasons: a belief may be called into doubt or the agent may simply wish to entertain other possiblities. In the prominent AGM framework [1, 8] for belief change, upon which the work here is based, one of the three central operations, contraction, addresses this concern (the other two deal with the incorporation of new information). Makinson [23] has generalised this work by introducing the notion of a withdrawal operation. Underlying the account proffered by AGM is the idea of rational belief change. A belief change operation should be guided by certain principles or integrity constraints in order to characterise change by a rational agent. One of the most noted principles within the context of AGM is the Principle of Informational Economy. However, adoption of this principle in its purest form has be...

2 The two stits 2

It is commonplace in artificial intelligence to divide an agent's explicit beliefs into two parts: the beliefs explicitly represented or manifest in memory, and the implicitly represented or constructive beliefs that are repeatedly reconstructed when needed rather than memorized. Many theories of knowledge view the relation between manifest and constructive beliefs as a logical relation, with the manifest beliefs representing the constructive beliefs through a logic of belief. This view, however, limits the ability of a theory to treat incomplete or inconsistent sets of beliefs in useful ways. We argue that a more illuminating view is that belief is the result of rational representation. In this theory, the agent obtains its constructive beliefs by using its manifest beliefs and preferences to rationally (in the sense of decision theory) choose the most useful conclusions indicated by the manifest beliefs. 1 Introduction Levesque (1984) introduced the notions of explicit and implici...

Theories of rational belief revision recently proposed by Alchourr'on, Gardenfors, Makinson, and Nebel illuminate many important issues but impose unnecessarily strong standards for correct revisions and make strong assumptions about what information is available to guide revisions. We reconstruct these theories according to an economic standard of rationality in which preferences are used to select among alternative possible revisions. By permitting multiple partial specifications of preferences in ways closely related to preference-based nonmonotonic logics, the reconstructed theory employs information closer to that available in practice and offers more flexible ways of selecting revisions. We formally compare this new conception of rational belief revision with the original theories, adapt results about universal default theories to prove that there is unlikely to be any universal method of rational belief revision, and examine formally how different limitations on rationality affe...

: This paper is concerned with the construction of a base contraction (revision) operation such that the theory contraction (revision) operation generated by it will be fully AGM-rational. It is shown that the theory contraction operation generated by Fuhrmann's minimal base contraction operation, even under quite strong restrictions, fails to satisfy the "supplementary postulates" of belief contraction. Finally Fuhrmann's construction is appropriately modified so as to yield the desired properties. The new construction may be described as involving a modification of safe (base) contraction so as to make it maxichoice. Descriptors: belief, change, contraction, revision, base, theory. We often change our beliefs. We learn new things, occasionally things that conflict with our current beliefs. On such occasions new beliefs replace the old ones. It is as if this process is completed in two steps: (1) first we identify and throw out the beliefs that conflict with the new information and t...