Modality and nature of speaking task as mediators of the effects of collocation knowledge on perceived speaking proficiency in English as a second language

Abstract

Formulaic sequences refer to “a sequence, continuous or discontinuous, of words or other elements, which is, or appears to be prefabricated” (Wray, 2002, p.9). It is an umbrella term that encompasses a myriad of linguistic phenomena such as collocations, idioms, binomials and sentence stems. A growing body of research has shown that knowledge and use of formulaic sequences may enhance overall speaking proficiency in various first language (L1) and second language (L2) contexts. Nevertheless, the relationship between formulaic sequences and overall L2 speaking proficiency is still inconclusive and some issues remain unaddressed. First, the effects of formulaic sequences and single words on L2 speaking proficiency have not been directly compared. Second, knowledge of formulaic sequences has exclusively been measured through the visual modality, with no research on the predictive power of aurally measured knowledge of formulaic sequences on L2 speaking proficiency. Third, overall L2 speaking proficiency has been inconsistently conceptualized and defined, with different subconstructs used to represent learners’ overall speaking proficiency. This inconsistency could be due to a wide range of simplistic and less communication-intensive speaking tasks used to measure L2 speaking proficiency. It is therefore unclear if formulaic sequences can exert positive effects if more communication-intensive and demanding speaking tasks are administered. This study adopts a mixed-methods approach to address the above issues. 112 L2 users of English studying in a university in Hong Kong were recruited. Quantitatively, the participants took 4 types of vocabulary tests (visual single word knowledge, aural single word knowledge, visual collocation knowledge, aural collocation knowledge) and a multiple regression analysis was run to determine the type of knowledge that best predicts speaking performance in an academic group discussion. Qualitatively, assessors of the participants’ discussion took part in a modified think-aloud session to elicit the speaking proficiency constructs invoked when assessing and to compare their relative importance. This study found that collocation knowledge is a stronger predictor of L2 speaking proficiency than single word knowledge. Whilst both visually and aurally measured knowledge of collocations predicted overall L2 speaking proficiency significantly, visual collocation knowledge emerged as a slightly stronger predictor. The modified think-aloud elicitation sessions with the assessors revealed that when a more interactional and demanding speaking task (i.e. a university level academic discussion) was administered to measure the participants’ speaking proficiency, a wider range of speaking constructs were invoked by the assessors. To be deemed proficient, learners must not only show fluency, accuracy and idiomaticity in their speech but also show pragmatic competence. This study is amongst the first to statistically show that collocation knowledge is more predictive of L2 speaking proficiency than single word knowledge. The qualitative data yielded also suggest that collocations, or formulaic sequences in general, may need to be used with awareness of their contextual appropriateness in a complex and communication- intensive speaking task, an issue rarely explored in the literature. This study asserts the utility of incorporating collocations (or formulaic sequences) in L2 speaking curricula, together with awareness of their socio-pragmatic appropriateness.