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

 
Poster A17, Wednesday, November 8, 10:30 – 11:45 am, Harborview and Loch Raven Ballrooms

Noun and verb processing in French during sentence comprehension – an event-related potential study

Lauren Fromont1,2, Phaedra Royle1,2, Karsten Steinhauer2,3;1Université de Montréal, 2Centre for Research on Brain, Language and Music, 3McGill University

Syntactic and semantic processing are essential to sentence comprehension: syntactic categories (SCs) and lexical semantic information help integrate each incoming word into the sentence context. How these two processes interact is still under debate. Serial accounts such as Friederici’s (2002, 2011) extremely influential “syntax-first” model posit that SCs are analyzed first. SC violations were shown to elicit early left-anterior negativities (ELANs) and appeared to ‘block’ semantic N400 effects, suggesting that SC analysis is instrumental to semantic processing (Friederici et al., 1999). However, some studies in other languages (Yang et al., 2015 in Mandarin), could not replicate either the ELAN or the semantic blocking effect. The authors suggested that SC analysis is not automatic in languages with poorer morphological marking. In addition, Steinhauer and Drury (2012) argued that most previous ELAN studies involved pre-target context differences that may have caused ELAN-like artifacts as well as the absence of N400s. To avoid context effects, we created a balanced design that controls for both the context and the target word (verb or noun). In French, definite articles le/la/les (= the in (1)) and accusative clitics le/la/les (= him/her/them in (2)) are homophographs and appear before a noun or a verb, respectively. We selected verbs in (1) and (2) that subcategorize specifically for either noun or verb complements. In our design, only target verbs bear inflection. (1) Elles ôtent le crapeau. ‘They remove the toad’ (2) Elles osent le plaquer. ‘They dare tackling him’ Syntactic category violations (3) and (4) are created by cross-splicing the two sentences before the target word. (3) Elles ôtent le *plaquer. ‘They remove the *tackling’ (4) Elles osent le *crapeau. ‘They dare *toad him’ All experimental sentences (1-4) were preceded by a lead-in sentenc” that a) licensed clitics and definite determiners with a referent noun (e.g., “Mary and Jane are playing hockey with their friend” for sentence (2)), and b) established degrees of semantic priming, allowing us to study semantic N400s in the presence of SC violations. Event-related potentials across sub-conditions (1-2 vs. 3-4) revealed no ELAN for SC violations, suggesting that there is no early automatic stage of syntactic processing, and that previous ELAN findings may be context-driven artifacts. Instead, we found an N400 and a late P600, showing that lexical semantic processing is not blocked but takes place in parallel to syntactic processes. N400 responses were still present when the target was primed, suggesting that the two processes take place in parallel. Comparing between sub-conditions (1 vs 2, 3 vs. 4) revealed no difference between the violation conditions, but additional N400 effects on the correct noun (1) versus verb condition (2). Such differences may be attributed to the absence of inflection in the noun condition. Friederici (2002) TiCS 6 (2):78-84. Friederici, Steinhauer & Frisch (1999) Memory and Cognition 27 (3):438-53. Steinhauer & Drury (2012). Brain and Language. 120 (2), 135-162. Yang, Wu & Zhou (2015) PLoS ONE. 10(6)

Topic Area: Grammar: Syntax

Back to Poster Schedule