Context matters: How do task demands modulate the recruitment of sensorimotor information during language processing?

Front Hum Neurosci. 2023 Jan 17:16:976954. doi: 10.3389/fnhum.2022.976954. eCollection 2022.

Abstract

Many theories of semantic representation propose that simulations of sensorimotor experience contribute to language processing. This can be seen in the body-object interaction effect (BOI; how easily the human body can interact with a word's referent). Words with high BOI ratings (e.g., ball) are processed more quickly than words with low BOI ratings (e.g., cloud) in various language tasks. This effect can be modulated by task demands. Previous research established that when asked to decide if a word is an object (entity condition), a BOI effect is observed, but when asked to decide if a word is an action (action condition), there is no BOI effect. It is unclear whether the null behavioral effect in the action condition reflects top-down modulation of task-relevant sensorimotor information or the absence of bottom-up activation of sensorimotor simulations. We investigated this question using EEG. In Experiment 1 we replicated the previous behavioral findings. In Experiment 2, 50 participants were assigned to either the entity or action conditions and responded to the same word stimuli. In both conditions we observed differences in ERP components related to the BOI effect. In the entity condition the P2 mean amplitude was significantly more positive for high compared to low BOI words. In the action condition the N400 peak latency was significantly later for high compared to low BOI words. Our findings suggest that BOI information is generated bottom-up regardless of task demands and modulated by top-down processes that recruit sensorimotor information relevant to the task decision.

Keywords: EEG; ERP; body-object interaction; embodied cognition; semantic representation.

Grants and funding

This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), in the form of a Postgraduate Scholarship–Doctoral to EM and a Discovery Grant (RGPIN/03860-2018) to PP.