Age-related changes in representational flexibility are a characteristic feature of declarative memory development. The authors suggest that a qualitative shift in the nature of infants' memory representations accounts for increasing memory flexibility with age. We will argue that a comprehensive theory of declarative memory development must (1) account for the effect of experience on flexibility, (2) be empirically separable from more parsimonious explanations, and (3) propose a mechanism by which the transition takes place. We will argue that a converging-methods approach is necessary to understand not only when and what develops in declarative memory, but also how developmental change occurs.