A Synthesis of Some Recent Work in Sentence Production

Research into the psychological mechanisms underlying the syntactic processes deployed during sentence production seems to have begun in earnest with the appearance of Garrett (1975). While previous works had noted the existence of speech errors produced by normals involving linguistic units larger than phonemes, those works tended to focus on sound errors (additions, deletions, etc. of phonemes) and on the implications of such errors for phonological processing (cf., for instance, Fromkin, 1971). In contrast, Garrett’s (1975) article is chiefly concerned with distinguishing properties of word and stem errors from those of sound errors in order to argue for the deployment of two specifically syntactic levels of representation in production. It is a tribute to the insight and carefulness of Garrett’s research that nearly every investigator working on syntactic production since that article has felt compelled either to base their own analyses on it or to contrast their accounts with those that Garrett has presented.