Recent advancements in the field of infant false-belief reasoning have brought into question whether performance on implicit and explicit measures of false belief is driven by the same level of representational understanding. The success of infants on implicit measures has also raised doubt over the role that language development plays in the development of false-belief reasoning. In the current paper, we argue that children's performance on disparate measures cannot be used to infer similarities in understanding across different age groups. Instead, we argue that development must continue to occur between the periods when children can reason implicitly and then explicitly about false belief. We then propose mechanisms by which language associated with false-belief tasks facilitates this transition by assisting with both the processes of elicited response selection and the formation of metarepresentational understanding.