Multiword sequences in English language learners’ speech: The relationship between trigrams and lexical variety across development

Abstract

Languages have formulaic multiword sequences (MWSs) which occur repeatedly in speech and writing (e.g., Nattinger & DeCarrico, 1992; Siyanova-Chanturia & Pellicer-Sánchez, 2018). For learners, then, the production of MWSs is an important element in developing spoken language that is complex, accurate, and fluent. Though the use of MWSs is important for achieving spoken proficiency, it is unclear whether the production of MWSs supports or hinders another aspect of proficiency, lexical variety. This paper is an exploration of the production of MWSs (recurrent trigrams) and the development of lexical variety, found in 2-min speeches (n = 294) from English L2 learners (n = 66) over time in an intensive English program (IEP). Using hierarchical linear modeling and correlation analysis, we found different patterns of development for the two measures. The use of MWSs increased and then decreased while the lexical variety scores slightly decreased and then sharply increased over time in the IEP. Although the impact of MWSs on oral fluency has been studied, this seems to be the first study to consider how MWSs influence lexical variety across development.