The Phenomenology of Role Play: Are Children “Thinking-As-If” or “Behaving-As-If”?

Abstract

When children role play, they do things such as change their tone of voice or take on the character’s emotions or needs. These behaviours make it appear as though children adopt the mental perspective of the character they are pretending to be, which has led some researchers to argue that children “think-as-if” they are the character while role playing (e.g., Harris, 2000). However, it is questionable whether these sorts of behaviours can really be taken as strong evidence that children are “thinking-as-if”; children might simply be imitating the distinctive behavioural characteristics of the character they are pretending to be – or “behaving-as-if”. In two studies, I attempted to obtain better evidence that children “think-as-if” while role playing. To do this, I developed a novel paradigm called the Pretend Self-recognition (PSR) task that examines how children refer to photographs of themselves while they are pretending to be someone else. I reasoned that if children were truly “thinking-as-if”, they might see the photograph of themselves as a third person would, and thus refer to the photograph of themselves using their own proper name. In contrast, if children were simply "behaving-as-if", they might continue to refer to the photograph with the personal pronoun “me”. In Study 1, approximately half of 4-year-old children labeled their own photograph from the perspective of the character they were pretending to be, and thus showed evidence of "thinking-as-if". This finding was replicated when children were given the PSR task at two time points along with measures of theory of mind, executive functioning, pretense understanding, and narrative absorption. Results showed that PSR performance was stable across testing period, but was not related to any of the other constructs that were also measured. Taken together, the findings reported in this dissertation show that the PSR task is a reliable measure of perspective taking in role play, though the source of individual differences in the measure remains a target for future research. The lack of relation between individual differences might suggest that PSR performance is orthogonal to the other constructs and is something in its own right.