This article presents a theory of sentence production that accounts for facts about speech errors-the kinds of errors that occur, the constraints on their form, and the conditions that precipitate them. The theory combines a spreading-activation retrieval mechanism with assumptions regarding linguistic units and rules. Two simulation models are presented to illustrate how the theory applies to phonological encoding processes. One was designed to produce the basic kinds of phonological errors and their relative frequencies of occurrence. The second was used to fit data from an experimental technique designed to create these errors under controlled conditions.
ASJC Scopus subject areas