Abstract

We know very little about children’s ability to create complex mental models from verbal descriptions. This paucity might be explained by the difficulty of creating paradigms that would test analogous skills in this domain in children and adults. In two experiments we explore young children’s ability to take the perspective of a character central to a described scene and to dynamically update object relations when the character moves. In experiment 1, children were found to take the character’s perspective when they learned the layout of objects in a real-life model. In experiment 2 children learned the layout from text and gave responses to object location prompts in a computer based task measuring response times on a touchscreen. In line with predictions from adult spatial framework theory (Bryant, Tversky & Franklin, 1992), children recalled objects fastest and more accurately that were placed in front or behind the character, and slowest for objects placed left or right. Based on a novel methodology, these findings reveal that children take an internal perspective on a described scene, which differs from the perspective they learned the layout from, indicating that at a young age children form rich, dynamic mental models of described scenes.