Case-based reasoning (CBR) is a prolific research program in AI and cognitive science. Originally motivated by the study of natural-language processing, it has been applied to computer models of story-understanding, learning, and planning, as well as a host of commercial, manufacturing, legal, and educational areas such as multimedia instruction and design.

As its name implies, case-based reasoning involves the interplay of two major themes: a (positive) theme about “cases” and a (negative) theme about “reasoning.” The core intuition in CBR is that people reason and learn from experience, and that the knowledge acquired from experience, captured in cases, is central to their understanding of the world. In the domain of language, this intuition gives primacy to meaning and “semantics,” as opposed to form and “syntax,” which were of prime concern in the logicist approach. As a negative claim, therefore, CBR presents a challenge to (formal) logic. It promotes a different kind of reasoning than deductive inference, and emphasizes the role of memory in human cognition.

This chapter is structured to reflect these two themes. Starting with a brief overview of classical models and the treatment of meaning in the logicist tradition, it provides the background conditions for the later discussion of linguistic approaches to meaning, which sets the stage for CBR.