‘Not my cup of tea’: The effect of proficiency, frequency, and congruency on L2 knowledge of idioms

Abstract

Factors that predict second language (L2) collocation knowledge have been extensively studied, but research on predictors of L2 knowledge of idioms is still limited. The aim of the present study is to examine the effect of various factors established in L2 collocation research on idioms. These include L2 proficiency, idiom frequency, and congruency status (whether the idiom can be literally translated into the first language or L1). To that end, 225 L1-Arabic–L2-English speakers completed three measures (familiarity rating, meaningfulness rating, and meaning recall) involving 72 English idioms, half congruent and half incongruent. Additionally, the participants completed a vocabulary size test as a rough measure of their English proficiency. We initially examined the association between the ratings provided by the L2 participants and those of L1 speakers, which are available through a norming study. Then, we fit mixed-effects models to examine predictors of the three outcome measures. Raw meaning recall scores indicated that only 33% of the target idioms had their meanings correctly recalled. The statistical analysis showed a strong positive interaction between the familiarity and meaningfulness ratings provided by L2 speakers. However, the association between L1 and L2 ratings was weaker. Results of mixed-effect modelling showed that for all three measures, only estimated proficiency and congruency were significant determinants of L2 idiomatic knowledge. Additionally, estimated proficiency modulated the congruency effect only for the rating measures, with a larger difference between congruent and incongruent idioms as proficiency increased. We discuss the implications of these findings to the teaching and learning of L2 idioms.