Perspective-taking in referential discourse among Mandarin-speaking children during three-party conversations

During conversations, individuals continuously need to take others’ perspectives to constrain the referential domain. The present study specifically investigates how 3-year-old Mandarin-speaking children take the visual perspectives of two partners in three-party communications. In two experiments, two speakers took turns instructing children to identify objects on a shelf, using critical nouns that semantically matched two identical referents from the child’s visual perspective. However, one of these referents might not be visible to the speakers. The critical nouns were presented in the form of bare nouns, allowing for definite or generic interpretations in the Mandarin context, leading to a competition between single and dual references. In Experiment 1, when both speakers only saw one shared candidate object, children preferred definite interpretations, using their partners’ perspectives to narrow the referential domains. However, in Experiment 2, one of the speakers could see both candidate objects while the other could see only the shared one. Results showed that children tended to interpret both partners’ discourse similarly and preferred generic interpretations. Additionally, eye-tracking data indicated that children implicitly processed each partner’s visual perspective. This research sheds light on visual perspective-taking strategies among 3-year-old children in a Mandarin context and provides insight into how children integrate multiple perspectives in communication.