In four volumes, Werner Leopold documented his first child's acquisition of German and English (e.g. 1939; 1949). In this article I problematize contemporary myths about bilingualism that partly date back to Leopold's pioneering work and his theorizing about the one-language/one-person method of language development. Notably, this method worked for his first-born, but not for his second child, whose very existence has often not been noted. A dyadic bias - privileging the study of one parent/one child - has led to a neglect of the role of siblings and peers. Moreover, a cognitive bias has led to an under-analysis of the role of play in language acquisition. On the basis of work on play practices and activity settings, this paper addresses these biases, highlighting the role of social demands and play communities for multilingual development. Several explanations are presented for why Leopold's second child did not speak German. In foregrounding the social situation of development and the role of play, this paper advocates more holistic approaches, including the study of hybrid improvisations in everyday practices.
Family and Community Involvement