Lexical access and representations in children : naming and word learning

This thesis is concerned with naming and word learning in typical children. The investigations of naming concerned lexical processes and representations associated with children’s quick and accurate picture naming. The investigations of word learning concerned the development of naming processes and representations during word learning. The aim of the investigations of naming was to better understand the way that different cognitive processes contribute to the speed and accuracy of discrete and serial naming. The contribution of age, gender, semantic abilities, phonological abilities and speed of response to non-lexical stimuli was analysed in relation to children’s speed and accuracy of naming. The findings indicated that age, but not gender, was related to the naming process; that different processes are involved in discrete and serial naming, and that speed of non-lexical information processing speed appeared to be an important variable to accurate (and fast) naming. Additional analyses on a sample of children with Word Finding Difficulties concerned errors and the role of lexical factors in the naming process. These findings revealed that lexical factors were not as important as participant characteristics, such as age or language ability. The aim of the investigations of word learning was to assess the role of semantic and phonological information in relation to the development of lexical representations. Children learnt a set of nonsense words and their knowledge of these words was assessed using a comprehensive range of tasks involving different lexical abilities. The children had greater difficulty with assessments of production than assessments of comprehension, and did poorly on tasks involving the retrieval of semantic representations. In addition, comparisons of the effectiveness of different types of input revealed that visual input was most effective and that children were able to use information from one modality to form lexical representations in another modality. The implications of these findings for existing knowledge about typical populations, as well as for understanding atypical development, are discussed