The effect of 'missing' information on children's retention of fast-mapped labels

This paper explores 'fast mapping', one of several processes that have been proposed to be involved in the rapid vocabulary expansion observed in the preschool years. An adaptation of a receptive word matching task examined how well children retained a just-mapped relation between word and referent when some information was later missing. Thirty-nine children between the ages of 3;0 and 5;6 (mean age 4;3) were taught to select a black square if the correct match for a spoken label was not visible in an array of pictures presented on a computer screen. This procedure allowed children to respond even when they perceived that the correct referent was absent. In experimental sessions, children experienced a single exposure to a word-referent relation. Then, under one condition they heard the just-learned label but were not presented with the matching referent; instead, a completely novel referent was visible along with the black square. Under another condition, they were presented with a just-learned referent (and the balck square) but heard a completely new label. The question of interest was whether the children appreciated that an earlier-learned map precluded re-assigning a label to a new referent or re-assigning a referent with a second new label. If so, they should select the black square under both conditions. The majority of children (69%) resisted re-assigning a just-mapped label to a completely novel referent and selected the black square, even when the original referent was not in sight. However, fewer of these children resisted accepting a second label for a just-named referent (46%). Older children were significantly more likely to adhere to original maps than were younger children.