You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster B15, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Event-Related Potentials Indicate a Role for Word Frequency in L1 and L2 Grammatical Processing

David Abugaber1, Irene Finestrat1, Alicia Luque1, Kara Morgan-Short1;1University of Illinois - Chicago

Various approaches to psycholinguistics posit a critical role for the frequency with which particular linguistic forms are encountered in a language user’s input. Indeed, converging evidence from several methodologies, including from event-related potentials (ERPs), has found performance and processing effects for the frequency of individual lexical forms (Van Petten & Kutas, 1990; Van Zeeland & Schmitt, 2013; Webb & Chang, 2015). However, a common prediction of frequency-based accounts is that frequency would also have effects on grammar, such that grammatical processing would be less difficult for high-frequency forms than for lower-frequency forms (Ellis, 2002). Evidence for this has previously been found through corpus-based analyses (e.g., Ellis & Ferreira-Junior, 2009), longitudinal studies of L2 learners (e.g., Madlener, 2015), and behavioral experiments using both offline (e.g., Lee, 2002; Slabakova, 2015) and online (e.g., Ellis, Simpson-Vlach, & Maynard, 2009; Lehtonen et al., 2012) methodologies. However, to the authors’ knowledge, no study to date has examined these frequency effects at a direct, neurolinguistic processing level, as assessed by ERPs. Such an experiment would provide deeper insight into the nature of grammatical processing while overcoming certain issues with the aforementioned approaches (such as underreporting and reactivity) and being able to detect effects that might not be captured through behavioral measures alone (McLaughlin, Osterhout, & Kim, 2004; Tokowicz & MacWhinney, 2005). To address this gap in the research, an ERP study was conducted in which eight English-L1 Spanish-L2 learners (enrolled in intermediate-level Spanish foreign classes in a university setting) were asked to perform grammaticality judgment tasks on stimulus sentences that were either correct or contained a syntactic (phrase structure) violation. These were presented via a rapid serial visual presentation paradigm. The study was conducted in two separate sessions, with one session performed in L1 English and another in L2 Spanish. Critically, the stimulus sentences varied not only by their grammaticality status (phrase structure violation vs. correct) but also in terms of the critical words’ frequency in the language as a whole (low frequency vs. high frequency), as gauged using the BNC (BNC Consortium, 2007) and LexESP (Sebastián-Gallés, 2000) corpora. For both L1 and L2 processing, global ANOVAs revealed Violation by Frequency interactions in the 300-500 millisecond post-stimulus time window indicating marked negativities for grammatical violations involving high-frequency (vs. low-frequency) critical words. By contrast, in the 500-900ms time window, for both L1 and L2 there were trends towards Violation by Frequency interactions that were suggestive of positivity effects (P600s) towards grammatical violations with high-frequency (but not low frequency)critical words. Although data collection for this project is ongoing, these preliminary results seem to suggest a role for word frequency as a critical factor in grammatical (syntactic) processing, as indicated by differences in the ERP effects for low vs. high-frequency words. Beyond merely informing theories on the nature of both L1 and L2 neurolinguistic processing, such findings have practical implications for language teachers vis à vis manipulating input frequency so as to best foster students’ second language development.

Topic Area: Grammar: Syntax

Back to Poster Schedule