"The Role of Prediction in Second Language Processing"
Friday, February 11, 2022, 9:00am, via Zoom.
Trying to anticipate what will happen next is normal human behavior. The role of prediction and expectations in language processing has been a topic of central interest and debate in psycholinguistics. Yet it is only relatively recently that this perspective has been extended to investigations of second language (L2) processing (Kaan, 2014; Kaan & Grüter, 2021). Most L2 studies to date have looked at prediction at the level of lexical semantics and morphosyntax. In this talk, I will focus on expectations at the discourse level. I will present findings from recent work in our lab using both offline (written story continuation) and online (visual-world eye tracking) measures to examine to what extent native and non-native users of English create proactive expectations about upcoming reference based on linguistic cues such as grammatical aspect and implicit causality. Collectively, our findings indicate that non-native users engage in predictive processing to a lesser extent than native users, consistent with the RAGE hypothesis, which states that L2 users have Reduced Ability to Generate Expectations (Grüter et al., 2014, 2017). However, the consistent absence of L2 proficiency effects on predictive processing, both in our studies and other recent work, is inconsistent with the RAGE hypothesis. I will argue that the reduced engagement in prediction that we observe in L2 processing may not (only) reflect a lack of ability to engage in prediction in an L2, but differential utility of relying on potentially predictive cues to optimal processing outcomes in an L2 versus an L1. In other words, I will suggest we consider that not engaging in prediction may be a more successful processing strategy for some (second) language users.