Erj{ors in Linguistic Performance

This paper examines three directions for the course of future research on speech errors: observational refinement (ways of improving how speech errors are collected and analyzed), experimentation (ways of testing hypotheses concerning the nature of speech errors), and theoretical integration (ways of improving and extending our theories). The paper also proposes a general theory of speech production at the semantic level, which summarizes a number of facts concerning errors in lexical retrieval. The years since Meringer's (1895) groundbreaking work have seen a growing interest in speech errors, beginning first in clinical psychology (see, for example, Freud, 1901), then in linguistics (e.g., Sturtevant, 1947; Wells, 1951; Hockett, 1967; Fromkin, 1971), and spreading now to experimental psychology (e.g., Boomer and Laver, 1968; MacKay, 1971; Motley and Baars, 1976). We all owe Meringer an indirect debt, and many of us (e.g., Freud, 1901; Sturtevant, 1947; MacKay, 1972a; Celce-Murcia, 1973) owe him a direct debt for providing our first source of c;lrefully collected and catalogued data. Vienna seems an especially fitting place to express our gratitude and to assess the accomplishments stemming directly or indirectly from Meringer's work. We can begin by commending the sheer courage of our enterprise. Following directly in Meringer's footsteps, we have risked the patience and amicability of our friends, families, and colleagues to collect speech errors whenever and :wherever they occur in our everyday lives. We have defied the Zeitgeist by advocating theories, metatheories, and paradigms that are somewhat ahead of their time. We have dared to advocate radically new methodologies within our respective fields, since speech errors do not fit the traditionally accepted techniques of either psychology or linguistics. We have reached an impressive degree of consensus on goals and methods: