Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Grammatical Gender and Number Agreement in Spanish Heritage Speakers: An ERP study
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
Alexandra Roman Irizarry1, Judith Kroll, Eleonora Rossi2; 1University of California, Irvine, 2University of Florida, Gainesville
Heritage speakers (HS) are bilingual users who typically acquire a minority language at home as their first language along with a societal language as the second language in formal school settings. Despite being the most representative group of bilinguals in the US (AAAS, 2017), they are the least studied group. HS may vary in proficiency in the heritage language due to limited opportunities to develop literacy skills in it. In Spanish, the heritage language under study, HS vary in their ability to establish grammatical gender agreement when compared to “native” homeland controls (Keating, 2022). What is often overlooked is how Spanish HS’ grammatical gender processing might compare to highly proficient L2 learners of Spanish. The comparison is of interest because both groups are English dominant, yet differ in how they acquired Spanish. While HS acquire Spanish early in life in a naturalistic environment, L2 speakers acquire Spanish beyond early childhood via formal education. We hypothesize that differences in Spanish acquisition will result in different neurophysiological patterns. Previous research on L2 Spanish learners suggests that highly proficient learners exhibit more native-like P600 responses to gender violations (Rossi et al., 2014). While studies that examine Spanish HS’ processing of grammatical gender agreement using online methods are currently emerging, electrophysiological studies on the topic are scarce. This study seeks to fill in that gap by examining Spanish HS’ sensitivity to grammatical gender and number agreement in clitic pronouns. A secondary goal of the study is to examine the role of individual differences in language proficiency in moderating the neural responses of different violations. Methods: 26 Spanish HS from Southern California read sentences in four conditions via RSVP as their EEG was recorded (Rossi et al., 2014). The conditions were: correct gender and number agreement (e.g., “Después de lavar los cuchillos, Andrea los colocó en la mesa del comedor” (After washing the knives, Andrea placed them on the dinner table), gender violation (e.g., … Andrea las…), number violation (e.g., … Andrea lo…), and a double violation (e.g., … Andrea la…). Each sentence was followed by a grammaticality judgment and a comprehension question. Preliminary results suggest that relative to the baseline condition, gender violations elicited a lexico-semantic N400 as well as a left-anterior negativity (LAN), while number and double violations elicited P600 responses. The LAN effect from the preliminary data suggests HS’ exhibit an automatic grammatical gender sensitivity but no later P600 reanalysis. In contrast, number violations and double violations elicited only later reanalysis, suggesting less reliance on faster automated processing. These findings evince the importance of using highly sensitive methods that capture the nuances that come with the heritage language experience. Future Directions: Data collection is ongoing to assess the effects of individual differences. We predict that individuals with higher Spanish proficiency will elicit more P600 effects, while individuals with lower Spanish proficiency will elicit more N400 effects.
Topic Areas: Multilingualism, Language Development/Acquisition