Saltmarsh R, Mitchell P, Robinson E
Department of Psychology, University College of Swansea, UK.
Cognition. 1995 Dec;57(3):297-325. doi: 10.1016/0010-0277(95)00675-5.
In a standard deceptive box procedure, children aged around 3 years typically fail to acknowledge their own prior false beliefs. For example, they judge incorrectly that they had initially thought a Smarties tube contained pencils after discovering these to be the actual content. Wimmer and Hart (1991) showed that children were more likely to answer correctly in a variant of this task known as a "state change", procedure. In this task, they saw that a container held its expected content (so the initial belief was true) before this was exchanged for something atypical. This appears to offer powerful evidence suggesting that children who fail the standard task do not understand about belief. However, we argue against this view. In a series of 4 experiments, we show that when children see the expected contents before these are swapped for something atypical, this not only makes it easier to report their own and a puppet's initial true belief but also a puppet's current false belief. The results are consistent with the "reality masking hypothesis", according to which facilitation is due to the belief option being linked with a physical counterpart in the state change procedure.
