Visual-structural context effects in the spoken production of object names

2013
The processes underlying speech production, in particular object naming, are subject to many influential factors. The naming of a target object (e.g., DOG) in context with categorically related distractor words or pictures, results in slower response times in naming. This effect is known as semantic interference. Each model of speech productionaccounts for this interference effect with the assumption that a semantic category node is involved. (e.g. The node of ANIMAL encompassing exemplar items such as dog, or horse). The activation of this node results in co-activated lexical candidates within the category, that compete for selection, therefore increasing the time taken in naming a target. The assignment of a target picture concept (e.g., DOG) to a category within semantic memory(e.g., ANIMAL) often involves the recognition of visual-structural features that define the category itself (e.g., has a face, fur, four legs, tail etc), signifying the importance of visual features in classification. Yet no studies have focussed on the importance of visual features in the absence of categorically related concepts. To investigate this, a picture-word interference (PWI) task and a blocked cyclic naming task were utilised, in which visual similarity between target pictures was maximised while removing shared category membership and any semantic relationships. It was found that within the PWI task, the time taken in object naming was greater when both the target and distractor were visually similar compared to when they were dissimilar. Within the blocked cyclic naming task, it was found that within homogeneous blocks (visually similar); reaction time was slower across cycles than within heterogenous blocks (visually dissimilar). This evidence suggests that the activation of a category node may not be necessary to explain interference effects, rather shared feature overlap may provide a sufficient explanation for competitive activation of lexical candidates in selection.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []
    Baidu
    map