Young instructed learners of a second language are known to rely extensively in the early stages on rotelearning and formulaic language; the relationship between this formulaic knowledge, and the eventual emergence of productive morphosyntax, is still poorly understood.

This paper draws on data from a longitudinal study of 73 classroom beginner learners of French, aged 5, 7 and 11. Divided by age, each group received 38 hours of instruction by the same teacher over a period of 19 weeks. All lessons were captured on video and transcribed, providing complete documentation of all L2 French classroom input and interaction. Children’s developing knowledge of French was regularly tested using a variety of receptive and productive tasks, including an elicited imitation test, a receptive vocabulary test, and a role play task.

Previous analyses have shown that the 11 year old beginners made faster overall progress in morphosyntax than the younger children. Here, we explore the relationship between use of formulaic language and the emergence of productive morphosyntax, for the different age groups, in order to explain the apparent advantage of the older group. We analyse children’s French oral productions in two datasets: a) the group role play tasks, and b) the elicited imitation test. We depart from established practice in the scoring of EI tests, which is primarily meaning-based and provides information on test-takers’ overall proficiency (Tracy-Ventura et al 2014), and instead focus on formal features of children’s production (the reproduction of NPs and VPs: McCormick & Zach, 2016). We explore the relative abilities of the different age groups in use of formulaic expressions and in the (re)production of non-formulaic morphosyntax, and discuss the implications for young learner pedagogy.