Development of word-finding

Abstract

The ability to produce names in response to three lexical access conditions (a semantic description, a rhyming form, and a picture) was studied in 50 normal children (10 each at 6, 8, 10, 12, and 14 years of age). Both accúracy and speed of word-finding were considered. Production of target names and response speed improved with age, although some age-related changes were nonlinear and some varied with the type of lexical access. The children's naming errors were generally real words in the language, produced with phonetic accuracy and bearing a semantic relation to the target name: even when naming pictures, errors were semantically rather than visually related to the target. The older the child, the closer was the error to the target, and, in the oldest children, error and target represented a minimally contrastive set of semantic terms. Failure to find a word in the youngest group was due to a lack of facility in organizing target descriptors, a tendency to ignore the syntactic information defining grammatical class of the target, and, perhaps most important, the lack of a fully hierarchical lexical organization. The data suggest that the accessibility of a name for children varies with the context in which it is accessed, but that lexical access is, for the most part, semantically based; and, that differences between younger and older children in naming skill concern the level, and the direction, at which semantic information is represented.