Previous work found that children as young as 15 months sometimes behave as if they understand wh-questions. For instance, 15-month-olds who hear questions like Which monkey is the frog feeding? will prefer to look at a video where a frog feeds a monkey, compared to a video where a monkey feeds a frog. We asked whether this behavior is driven by adult-like representations of the syntax of these questions, or by a strategy driven by verb knowledge. Specifically, if children know that the verb feed requires an object, and know that the frog is the subject, then they might be inclined to look more at a scene where a frog feeds someone, rather than a scene where a frog is getting fed. This hypothesis predicts that 15-month-olds will perform accurately only if they know enough verbs to deploy this strategy; therefore, their performance should depend on vocabulary.
We tested 15-month-olds with a range of productive verb vocabulary in a preferential looking task. Some infants heard wh-questions like Which monkey is the frog feeding? or Which monkey is feeding the frog? A second group of infants heard sentences with relative clauses, like Find the monkey that the frog is feeding and Find the monkey that is feeding the frog. Infants saw two videos on the screen at the same time: one where a monkey is feeding a frog, and one where the same frog is feeding a different monkey. We measured infants’ looking times towards each video in order to see how they interpreted these sentences. Confirming our predictions, we found that 15-month-olds only succeeded on this task if they had relatively high verb vocabulary. This is consistent with the hypothesis that infants at this age may not represent the full syntax of these questions, but can comprehend them by using their developing verb knowledge.