Abstract
We investigated the effects of language on vision by focusing on a well-known problem: the binding and maintenance of color-location conjunctions. Four-year-olds performed a task in which they saw a target (e.g., a split square, red on the left and green on the right) followed by a brief delay and then were asked to find the target in an array including the target, its reflection (e.g., red on the right and green on the left), and a square with a different geometric split. Errors were overwhelmingly reflections. This finding shows that the children failed to maintain color-location conjunctions. Performance improved when targets were accompanied by sentences specifying color and direction (e.g., “the red is on the left”), but not when the conjunction was highlighted using a nonlinguistic cue (e.g., flashing, pointing, changes in size), nor when sentences specified a nondirectional relationship (e.g., “the red is touching the green”). The relation between children's matching performance and their long-term knowledge of directional terms suggests two distinct mechanisms by which language can temporarily bridge delays, providing more stable representations.
Get full access to this article
View all access options for this article.
