To move or not to move: An ERP study on the processing of fictive and literal motion expressions

Abstract

This study used electroencephalographic evoked response potentials to investigate the processing of fictive motion and literal motion during natural language comprehension. A hypothesis is that the motion component of a verb is preserved in both literal and fictive motion constructions (‘The army/The bridge crossed the river’). However, the incorporation of a motion-event frame into fictive motion constructions requires reanalysis or reconstruction both syntactically and semantically. Comparing fictive motion constructions to literal motion constructions revealed a larger P300 at presentation of the subject noun phrase, a larger P600 at motion verbs, and an N400 modulation on the sentence-final complement noun phrases. These results suggest that the processing of fictive motion requires increased cognitive effort relative to literal motion condition, particularly computations related to thematic role assignment and semantic unification to construct a motion-event frame involving mental simulation of motion.