Abstract
Word learning poses a referential problem: when a label is presented, there are potentially many objects in view, each with several different properties that may or may not be relevant to the category being learned by the child. Most theories of word learning solved this problem by proposing that children have language-specific knowledge that guides their learning. In this set of experiments, however, we explore an alternative framework: by keeping track of regularities in the environment (e.g. most object categories are organized by shape, words and shapes co-occur), the learner’s attention can be rapidly cued to an object by a word. Under this attention-based account, verbal labels should show clearly measurable effects on the rapid deployment of attention in visual search. We developed a search task suitable for young children using a computer touchscreen, and tested the prediction that words can cue children’s attention to the shape of the objects. In the first experiment we used a conjunction search (with half the distractors sharing the shape but not the color with the target, and vice-versa) and in the second experiment we manipulated the degree of shape similarity among target and distractors. Both experiments varied the number of distractors (2 to 12 in experiment one; 3 to 12 in experiment two). Across experiments we found a decrease in search times (the intercept but no interaction with number of distractors) when the label that designates the target object was presented. The presentation of an unrelated word (‘go’) did not facilitate search times, ruling out sound-related arousal effects. The results suggest that nouns rapidly cue attention to shape in 3 years-old children, providing a stepping stone to a mechanistic account of how words organize attention – and in so doing, organize early word learning.
Meeting abstract presented at VSS 2012