Towards a graded lexical inventory of multi-word combinations: the distribution of collocations across Spanish CEFR levels

Abstract

Collocations have become increasingly important in our understanding of foreign language learning. When it comes to setting vocabulary learning goals, concerns about how to address collocations still arise today. This article explores the distribution of collocations in L1 and L2 Spanish production with the ultimate goal of informing the design of graded lexical inventories of multi-word combinations. To do so, we explore three defining properties of collocations in L1 and L2 production data, and across different levels of L2 proficiency: syntactic structure, semantic transparency, and the strength of association. Results indicate that there is an increase of collocational density and diversity, but that isolated features of collocations fail to predict L2 proficiency. Findings suggest the need to evaluate collocation use at a high level of granularity.