A new multiword naming deficit: Evidence and interpretation

Abstract

This study analyses the naming performance of MP, a nonfluent aphasic speaker, and demonstrates that she has a context-sensitive naming deficit, which manifests in worse naming of a target set when multiple targets are depicted at once and a multiname utterance is prepared. Experiments 1 and 2 documented the effect in composite-picture descriptions, compared to standard naming. Experiment 3 documented it in the naming of picture lists of length 2 and 3, compared to 1; and Experiment 4 documented it in the naming of one of two pictured items, under conditions that encouraged preparation of both item names. Several alternative accounts are considered and an argument is made for a type of multiword naming deficit that has not been described before. The deficit involves interference at the lexeme stage of retrieval, arising when multiple lemmas are co-activated during the preparation of a multiword utterance.