A new false belief test for 36-month-olds

We report two studies that suggest that some 36-month-old (and younger) children understand others' false beliefs. In the false belief conditions, children and two adults (E1 and E2) watched as an object was put into a container. E I left the room, and E2 switched that object with another. El returned, expressed her desire for the object, and struggled to open the container (without succeeding). She spied both objects across the room and said to the child. 'Oh, there it is. Can you get it for me?' In other conditions, the object was not switched or E I witnessed the switch. Other variations included using a novel word for the object and removing the 'pull of the real'. Measures of children's latency and uncertainty were taken to determine whether those who were correct on the false belief tasks were guessing (luckily) or truly understanding. Results showed that between one-third and almost two-thirds of children took account of the adult's false belief when deciding which object the adult was requesting. We compare this task with other implicit and explicit tasks. We also conclude that certain task demands (e.g. the pull of the real) in traditional theory of mind assessments are unnecessary.