Summary, in English
This research concerns lexical inferencing (LI), i.e., the ability to infer the meaning of unknown words in a running text (Haastrup, 1991). Despite its advantages, there is little research on LI as a collaborative effort. Further, virtually no studies compare paired and individual LI task performances. Based in Task-Based Language Learning (TBLL) and Vygotsky’s (1978) Sociocultural Theory (SCT), the present study seeks to explore the relationship between global second language (L2) proficiency and LI success. It further investigates the effect of (a) collaboration and (b) patterns of interaction on LI. The participants (N = 22) were adult learners of English as a second language (ESL) taught by the teacher-researcher. Their proficiency levels were determined by a series of tasks-referenced proficiency measures, namely Reading, Writing, Listening, Speaking and Vocabulary (Bygate, 2016). This enabled the creation of mixed-proficiency dyads, engaging in paired LI. These collaborative task outcomes were compared to the same students’ individual LI task performances. Both tasks amounted to verbally inferring the meaning of 12 carefully selected target words in a running text. The findings corroborate previous research (e.g. Haastrup, 1990, 1991) showing a positive effect of global L2 proficiency on LI, as a correlation between the proficiency task scores and the LI scores was established. However, no statistically significant positive effect of collaboration was found. A qualitative analysis revealed that ‘collaborative’ dyads generally performed better than pairs labelled as ‘expert/novice’ or ‘dominant/passive’ (Storch, 2002). Pedagogical implications and suggestions for future research are discussed.