The Development of Formulaic Sequences in Dutch L2 Learners of English

This article reports on a longitudinal study of the development of formulaic sequences in 22 teenage Dutch learners of English in high and low input conditions (Dutch bilingual TTO and Gymnasium). Written samples of language were collected monthly over a period of 3 years and analyzed using a holistic, integrated approach to the identification of formulaic sequences (chunks). Our results have revealed different developmental tendencies in our low and high input groups, even though both started off with no significant differences in the overall number of chunks (p>.05) and with a comparable proportion of different chunk types. While low input learners developed a limited range of mostly grammatically-based chunks, high input learners show increase in a wide range of chunks types. These are more lexical, idiomatic, longer and more complex in structure, which is reflected in numerous overlaps and nested sequences. High input learners also show considerable increase in chunks which can be used to construct longer and more complex sentences. However, although the contrast between the two groups in the increase of chunks is striking (p=.001), clear individual differences in developmental paths have been detected. These will be explored in further studies.