Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Accounting for variability in event-related potentials reflecting second language processing: A role for memory?
Poster Session A, Friday, September 12, 11:00 am - 12:30 pm, Field House
This poster is part of the Sandbox Series.
Kara Morgan-Short1, Ana Rodríguez Gallego1, Alexis Berles1, Marina Ridchenko1; 1University of Illinois Chicago
Substantial variability among individuals has been evidenced in second/additional language (L2/A) neurocognitive processing as evidenced by electrophysiology (Luque & Covey, 2023). We have a limited understanding of the various factors that account for this variability, although recent evidence suggests a role for cognitive factors, such as working memory (Gabriele et al., 2021). Indeed, individual differences in both working memory and long-memory systems, such as declarative and procedural memory, have been shown to play significant roles in the development of L2/A knowledge (Linck, 2014, Morgan-Short et al., 2021). Yet, little research has examined their role in neurocognitive L2/A processing. Thus, the current study examines whether individual differences in working memory, declarative memory, and procedural memory can account for variability in the electrophysiological processing signatures in L2/A learners. The current study employs an event-related potential methodology to examine this question in regard to the P600 ERP effect, which associated with grammatical processing in first and L2/A languages. We tested 37 L2/A learners of Spanish, who were enrolled in intermediate to advanced Spanish college courses at a large midwestern University. Participants completed two tasks to assess their Spanish proficiency level (a written vocabulary/grammar test and an oral elicited imitation task). They also completed tasks that measured working memory capacity (shortened complex span tasks, Oswald et al., 2015) and declarative and procedural memory learning ability (an encoding and recognition task, Hedenius et al., 2013, and a serial reaction time task, Nissen & Bullemer, 1987, respectively). Finally, electrophysiological data was recorded (ANT Neuro) as participants judged the grammaticality of 244 Spanish sentences (half correct and half violation sentences, including phrase structure (PS), subject-verb agreement (SV), noun-phrase number and gender agreement (NPn, NPg) conditions. EEG was processed following standard lab procedures (Morgan-Short et al., 2022) and averaged for the P600 time window (600-1000 ms). Participants demonstrated intermediate levels of proficiency and reasonable levels of accuracy on the judgment task, with accuracy ranging from 61% to 82% across the four conditions. ERP analyses demonstrated small positivities for all conditions, although the only clear P600 effect was found for the NPn condition. Correlational analysis found that (a) working memory was positively correlated with the P600 effect for PS (r=.37, p=.03), but was not associated with the P600 for the other conditions; (b) declarative memory was not associated with the P600 for any condition; and (c) procedural memory was positively correlated with NPg (r=.37, p=.04), negatively correlated with NPn (r=-.35, p=.05), and not associated with the other conditions. Overall, these preliminary results do not yet show a clear pattern between the development of the P600 effect and cognitive abilities within these memory systems. Although two positive associations were found between the P600 effect and working memory and procedural memory, other correlations were not statistically significant, and perhaps somewhat surprisingly, we found a significant negative correlation between the P600 and procedural memory for the NPn condition. With data collection ongoing, we consider potential interpretations of the associations found in this preliminary analysis in light of participants’ L2/A proficiency level.
Topic Areas: Multilingualism, Language Development/Acquisition