The Time-Course Of Metaphor Comprehension: An Event-Related Potential Study

We can easily understand familiar metaphors,but have difficulty in understanding novel metaphors.How can we get metaphoric meanings?Are both metaphoric and literal meanings processing concurrently and involving the same mechanism,or else ,we should first try to access the literal meaning and only after failing we try to find a metaphorical one? Researchers then proposed three hypothesis : three-stage model ,parallel-processing view and the graded salience hypothesis. The present paper verified the above assumption experimentally.Two experiments investigated whether metaphor process different from literal. In Experiment 1,reaction time was recorded while subjects were reading short familiar metaphors novel metaphors, literal sentences or error sentences presented in the screen of computer. Participants performed a semantic judgment task in which they decided whether each sentence conveyed a meaningful expression .In Experiment 2,the same stimuli and presentation was examined using Event-related potentials. All the participants are undergraduate student of shanghai normal university, 20-25years old.Experiment 1 , the effect of sentence type was not significant, which supported parallel-processing view.Experiment 2 , ERPs behavioral and potential data revealed that (1) ERPs behavioral data is consistent with Experiment 1, it supported parallel-processing view;(2) N400 reflects the difficulty of language processing. The terminal word of error sentences elicited the largest N400 amplitude, followed by novel metaphors, familiar metaphors, with literal sentences smallest. The effect of sentence type was significant,F=8.32, p<0.01.This shows that the difficulty of understanding novel metaphors is significantly greater than familiar metaphors;(3) The third finding of this study is the P600 amplitude gradient, starting with small P600 for error sentences and increasing for novel metaphors, familiar metaphors, and finally largest for literal sentences. The effect of sentence type was significant ,F=8.36,p<0.01.P600 reflects semantic integration of later period. Familiar metaphors aroused subjects many memories and images. Subjects stopped processing novel metaphors and error sentences ,because they had already judged that theses sentences didn't convey anything meaningful;(4)We found N200,but the effect of sentence type was not significant, F=2.29,p>0.05;(5)Although understanding metaphors requires both hemispheres, the relative contribution of each hemisphere at specific processing stages depended on stimulus type .In the initial stages of word recognition, the left hemisphere has a privileged role in the processing of familiar metaphors .In the semantic understanding stage and semantic integration phase, the two hemispheres both participate in processing.