Surprise or confirmation?: predictability helps young children remember new name-animal pairs

In order to investigate the relationship between children’s expectations and later memory for new information, in this study we tested how lexical predictions impact preschoolers’ new name-referent mapping. English speaking 3- to 5-year-olds (n = 60) were presented with 12 new names for familiar animals (e.g. “Clark is a … shark”) in a story-like context designed to encourage lexical prediction with rhyme and pausing. By manipulating the likelihood of each target animal to end the rhyme, children’s accurate predictions were encouraged or discouraged. Participants’ anticipatory looks to target animal pictures were coded, and children’s subsequent retention of the new name-animal pairs was tested. We found that regardless of whether children were in a presentation condition designed to invite predictions or a more neutral control condition, across conditions, children made anticipatory looks and remembered the new names of the more predictable, likely animals more frequently than the names for unlikely animals. Crucially, an exploratory probe of the direct connection between children’s anticipatory looking behavior and their later retention revealed that children were more successful in remembering the names of targets when they had anticipated them than when they had not anticipated them and been surprised, regardless of their assigned likelihood. Within the constraints of the limitations discussed, these findings demonstrate an advantage of predictability over surprise in this type of naturalistic early language learning task.