Studies within evolutionary musicology and ontogenetic development propose an intimate relation between the quality of the human voice, the rhythm of interactional patterns (e.g. the alternation between repetition and improvisation), the origins of aesthetics, and the characteristics of performances within the temporal arts. Focusing on the role of auditory perception in children's development of narrative skills, this article similarly proposes an intimate relation between children's voices in interaction, their imitative use of formulaic and genre-specific language, and their creative and aesthetically attuned written compositions. The notion of voice opens up a productive and coherent approach to investigating how children interactionally and imitatively come to develop a command and reflexive understanding of spoken and written genres. The discussion is based on a full ethnography of children's acquisition of written language during their second school year.