The phonological loop and the central executive as predictors of explicit and implicit knowledge of English passive voice

Working memory (WM) is a key cognitive factor determining the outcomes of second language (L2) learning (e.g., Suzuki & DeKeyser, 2017). Research indicates that two components of WM, the phonological short-term memory (PSTM) and the working memory capacity (WMC), are implicated in learning L2 grammar (e.g., Li, Ellis & Zhu, 2019; Martin & Ellis, 2012; Suzuki & DeKeyser, 2017). However, it still remains unresolved which of them plays a more important role in the comprehension and production of grammar or how they affect different types of L2 knowledge. The paper reports a study which investigated the relationship between PSTM and WMC, and explicit and implicit knowledge of English passive voice (DeKeyser, 2009, 2017; Ellis, 2009), taking into account the mediating effect of overall mastery of grammar (i.e., grades on the final exam) and self-reported length of study. Participants were 152 Polish university students majoring in English. PSTM was assessed with the use of the Polish Non-word Span (PNWSPAN) and a digit span, based on Wechsler Adult Intelligence Scale (WAIS-R(PL). WMC was tapped by means of the Polish Listening Span test (PLSPAN) and the Polish Reading Span (PRSPAN). Receptive and productive explicit knowledge was measured by means of an untimed grammaticality judgment test and a test requiring paraphrasing sentences. As regards implicit knowledge, its receptive dimension was assessed by means of a timed grammaticality judgment test, and its productive dimension though a focused communication task (Ellis, 2003). Evaluation focused on accuracy and suitability in a given context. Canonical analysis revealed that mainly WMC as measured by listening span affected both types of knowledge but primarily with respect to its productive dimension. At the same time, a much weaker effect of PSTM was observed.