A cross-sectional, correlational study of 30 children, 3 to 5 years ol
d, investigated relations between their theory of mind development and
social interaction, controlling for age and general language ability.
Children's overall performance on 4 standard false belief tasks was a
ssociated with their production of joint proposals and explicit role a
ssignments during a 10 minute session of pretend play. False belief ta
sk performance was not associated with the child's total amount of pre
tend play or with a measure of empathic concern. We discuss the signif
icance of an association between a laboratory measure of theory of min
d development and children's behaviour observed in a naturalistic sett
ing. We argue that the description and explanation of children's theor
y of mind and social understanding is best pursued through the combine
d efforts of experimenters and ethologists.