Moving Generative SLA from Knowledge of Constraints to Production Data in Educational Settings

Abstract

This paper begins by reviewing generative-based research from the past ten years. It will highlight some of the key findings and methodological tools from sentence processing research (e.g., Juffs & Rodríguez, 2014) and morphological processing (e.g., Clahsen et al, 2010; Cunnings, 2016; Diependale et al., 2011). Based on these findings, I will suggest that many questions that have concerned formal SLA (‘access to UG’, issues regarding ‘shallow’ processing and morphological ‘insensitivity’) are being resolved in favor of a view that, given time and exposure, adult L2 learners demonstrate knowledge that approaches (bilingual) ‘native speakers’. However, I argue that to remain relevant to a broader audience o f applied linguists, L2 research based on linguistic theory should account not just for knowledge of constraints on language, but also for what students comprehend and produce in over time in instructed contexts. Based on data collected in the English Language Institute at Pitt, I will make the case that corpus-based research that is informed by (formal) linguistics is necessary in a field that is increasingly dominated by scholars who see input frequency and saliency as the sole determining factors in acquisition. I will discuss studies from lexical, phonological, and morpho-syntactic development from students who have studied in an intensive English program over three semesters and show that the level of detail that a careful linguistic analysis provides is one that is useful to SLA research and language educators alike.