The Impact of pronunciation instruction on phonological global constructs: accentedness, comprehensibility, and intelligibility in L2 Spanish

Abstract

This study explores the effect of pronunciation instruction (PI) on three phonological global constructs––accentedness, comprehensibility, and intelligibility––in second language (L2) Spanish speech. These measures are based on listener judgments versus acoustic analysis as a result of the unlikelihood of ultimate attainment (Flege et al., 1995). An instructional intervention for intelligibility focusing on intonation, rhythm, stress, and fluency was carried out under a multiliteracies framework during a 15-week semester in a fifth-semester Spanish course. L2 learners completed pre-, post-, and delayed post-tests in which they responded to open-ended questions and read paragraphs. In addition to the L2 learners who received PI, a control group of L2 learners who received the same instruction without the intervention responded to the same tasks at the same time points, one semester apart. Speech samples from both groups of learners were distributed to listeners recruited through Amazon Mechanical Turk and Prolific who rated the samples for accentedness, comprehensibility, and intelligibility. The results reveal that PI yielded less accented speech over time, with reduced accentedness in extemporaneous speech. PI did not demonstrate any advantages in developing more comprehensible or intelligible speech throughout the course of the project. These findings are explained as a result of trade-offs and task effects. That is, L2 learners were more focused on phonetic accuracy in extemporaneous speech through the production of grammatical constructions and lexicon with which they are familiar. In tasks that employed structures with which learners were less familiar, a task effect was observed, negatively impacting listener evaluations. Pedagogical implications from this study emphasize the role of task as influential in L2 production and acknowledge that trade-offs may be occurring in L2 speech at the intermediate level as evidenced by the variable development of all three global constructs over time. Research implications exemplify the advantage of delayed post-tests in longitudinal assessment and the challenges of online data collection. This project concludes by calling for additional exploration of instructional methods to promote the development of global constructs of L2 pronunciation in the language classroom.