Abashidze, D. & Trofimovich, P.
Concordia University, Montreal
Recently seen events influence gaze patterns during language comprehension. When monolingual participants see an action and they hear a related sentence, they preferentially inspect targets of recent events and delay to predict plausible future-event targets, independent of tense information [1, 2]. Bilinguals, who experience increased processing demands due to competition between their two languages [5], may have weaker lexical access [3] and weaker predictive processing abilities [4] relative to monolingual speakers. It is currently unclear how bilinguals can use recent event information as predictive cues and how much they rely on visual cues during spoken sentence processing.
The current eye-tracking study (N = 32) examined English-French bilinguals? reliance on recently seen events, focusing on their ability to predict a plausible future action. Early simultaneous Bilinguals, all exposed to both languages before the age of five, saw a videotaped actor performing an action (e.g., sweetening strawberries, see [1] for sequence of events in an experimental trail) and then listened to an English (NP1-Aux-Verb-NP2) sentence (e.g., The experimenter has sweetened the strawberries) referring to that recently performed action or heard an alternative sentence (e.g., The experimenter will sweeten the pancakes) referring to an equally plausible action that the actor would perform next (e.g., sweetening pancakes). Eye movements to the recent and future objects were analyzed during the sentence (Fig. 1).
Preliminary results indicate that bilinguals performed similarly to monolinguals, preferring to inspect recent-event targets when exposed to both tenses. Although bilinguals showed a decrease in eye-gaze frequency toward recent-event targets (compared to monolinguals [1, 2]) when listening to future sentences, they inspected the recent target less than the plausible future-event targets in this condition (only in the NP2). Bilinguals and monolinguals thus appear to be constrained by similar processing biases in their comprehension of spoken discourse in the presence of visual information.