INTRODUCTION: The theory of language embodiment posits that words' meanings are grounded in sensorimotor circuits in the brain. The cognitive processes involved in verbal material recognition would depend, then, on bodily actions and sensations. Embodied language processing was shown in neuroimaging (fMRI) studies for both the verb and noun classes of words.

AIM(S): We aimed to study the embodiment of the verb class based on the event-related potential of the brain. ERP technique, with its excellent temporal resolution and well-established correlates of semantic processing, can shed more light on the specifics of the processes involved in embodied words’ meaning recognition.

METHOD(S): 50 right-handed, native Polish speakers viewed 80 verbs presented centrally for 800 ms each. 40 of the verbs denoted body part movements (20 arm-related actions, e.g., “to catch”, and 20 leg-related actions, e.g., “to walk”), and 40 denoted abstract or “disembodied” concepts (e.g., “to think”, “to desire”). All the stimuli used were matched on length and word frequency values. After each presentation, the participants had to decide whether they saw a concrete or an abstract word. Brain activity was measured with a 64-channel Biosemi Active Two EEG apparatus. Amplitude differences between the verb categories in the N400 and LPP time windows were analyzed. Behavioral data of post-stimulus categorization were also analyzed.

RESULTS: Higher N400 amplitudes were noted for concrete (movement or action-related) verbs as compared with the abstract ones, whereas higher LPP amplitudes were noted for the abstract verbs as compared with the concrete ones. Action verbs were also classified faster and more accurately.

CONCLUSIONS: Embodied, action-related verbs were shown to be processed differently by the brain. More resources were engaged during their semantic elaboration, and seemingly less effort was involved in the decision-programming stage. This was followed by an overall better behavioral performance for that verb category.