From age two, children use pronouns to predict who will speak next in conversation

Children's exposure to language is shaped by interactional needs in conversation. Prior research has largely focused on how such language influences word learning and long-term lexical knowledge. However, the effects of interactional language are likely to go far beyond word learning. Earlier studies showed that in conversation, from around age 2;6 children who are watching a conversation are more likely to spontaneously switch their gaze to an upcoming responder when they hear a question compared to when they hear a non-question. However, what information drives these predictions remains unclear. Tracking the eye gaze behavior of Dutch children (1-4-year-olds) and adults, the purpose of this research was to examine whether participant's predictions are driven by individually informative linguistic cues, comparing two cues associated with interrogatives: one lexical cue (subject pronoun) and one canonically associated prosodic cue (utterance-final intonation). We find that from age children 2;0 make more and earlier anticipations of an upcoming addressee response when hearing the early lexical cue (you vs. I subject pronouns), but we have no evidence that their predictions are changed by the later prosodic cue. Further, we investigated how cue use depends on linguistic context by comparing semantically meaningful (Study 1) and non-meaningful (Study 2) context. Only in meaningful contexts did participants show a pronoun advantage in predicting conversational structure. This suggests that using these cues relies on broader linguistic context. The findings take us a step closer to understanding how linguistic and interactional skills become intertwined in development.