Presents a model of the structure of logic considered necessary for computer processing of metaphorical language. Formally states and diagrams the algorithm for metaphors, isolating domain distance, predicate inequality, and hyperbole as particularly important factors. Distinguishes explanatory literal analogies from expressive metaphors,…

Examines the use of computers for error detection in natural language texts. Focuses on a computer program designed to teach students to edit their papers for errors using pattern matching in error detection. Describes a "pedagogy for editing" and speculates on ways to improve computer detection of errors in natural language texts. (AEM)