The effect of the number of labeled objects on novel referent selection across short and long time delays

Children often hear many new words in one conversation, and yet word learning research overwhelmingly focuses on how children learn and retrieve the meanings of single words. The current experiment tests how the number of labeled objects affects preschoolers’ novel word referent selection immediately after encoding and after a one-week delay. Seventy 3- to 6-year-olds were exposed to four novel objects. Half of the participants were given novel labels for two of the objects and half were given novel labels for all four. Label-referent mapping was tested with a four alternative forced-choice pointing task both immediately after exposure and one week later. Children performed worse overall after a week delay, replicating past work on novel word retention. While children performed significantly worse overall in the Four-Label condition, exploratory analyses revealed that this effect was driven solely by the second test trial immediately after exposure. Analyses suggest that referent selection is strongly influenced by in-the-moment constraints, such as label salience and pragmatic biases, and that these constraints are strongest immediately after novel word exposure.