An ERP Study On English Nominal Metaphor Comprehension By Chinese EFL Learners

Metaphor is one of the most interesting phenomena of human language not only because of its pervasive use in daily communication but,more importantly,because of its fundamental role in structuring our perceptions and understandings of the external world(e.g.,Arzouan et al.,2007;Glucksberg,2003;Lakoff&Johnson,1980).The central concern about metaphor comprehension lies in the argument over whether there exists any cognitive differences between processing metaphors and literal expressions,on which several models or hypotheses have been proposed and widely discussed.Moreover,a great amount of experimental studies have been reported,including behavioral studies and more recent ones using event-related potentials(ERPs)and fMRI technology.However,findings in previous literature have been contradictory and compared with the in-depth discussion of ERP studies on first language(L1)metaphor comprehension,neural mechanisms of second language(L2)or foreign language metaphor processing have yet to be fully explored.The present study aims to investigate via event-related potentials(ERPs)the cognitive mechanism underlying English metaphor processing by Chinese EFL learners,and tries to find ways to facilitate EFL teaching.Two experiments were designed to explore:1)whether the models proposed by previous studies are applicable to the Chinese mind;2)whether language proficiency would influence the underlying comprehension mechanism;3)whether a supporting context could"shortcut" conceptual mapping and thus cost fewer cognitive resources for Chinese EFL learners during English metaphor processing;and 4)whether the contextual influence on metaphor comprehension is the same between subjects with different levels of English proficiency.Experiment one was to answer the first two questions.Thirty participants were grouped based on their English proficiency,those who had passed TEM-8 or scoring more than 600 in CET-6 being in Group H and those who had not passed or scoring lower than 450 in CET-4 being in Group L.A sentence-final word paradigm was adopted with stimuli being of the X(s)be Y(s)form.Through a semantic judgment task,ERPs elicited by anomalous,metaphorical and literal sentence endings by subjects in both groups were compared and analyzed.Amplitudes of the N400 component(300-500 ms)formed a gradient for Group H,with anomalies eliciting the most negative N400s and literals the least,while in Group L those elicited by metaphors and literals did not differ significantly.Amplitude patterns in the later window(500-800 ms)remained similar for the two groups.Metaphorical and literal expressions triggered more enhanced LPC amplitudes than did anomalies,though not differentiated from each other.The results partially support conceptual blending theory in that all languages are processed with the same mechanism though costing different cognitive effort,but English metaphors and literal expressions are not clearly distinguished cognitively for Chinese EFL learners with low level of proficiency.It is also suggested in the present study that learners with higher English proficiency spent relatively more cognitive resources in metaphor processing.Experiment two was based on conclusions from the first experiment,answer the last two questions.Thirty participants were grouped into Group H and Group L using the same criteria as in the first experiment.An S1-S2 word-to-sentence paradigm was adopted with S1 being a word relevant or irrelevant to S2,a metaphorical sentence in X be a Y form.Participants were instructed to decide whether the word was relevant or not to the sentence and ERPs elicited by sentence endings under both conditions by subjects in the two groups were compared and analyzed.For Group H,amplitudes of the N400 component(300-500 ms)were less negative under relevant conditions,indicating reduced cognitive effort.However,relevant condition triggered more enhanced LPC amplitudes than did irrelevant one in Group H,suggesting more cognitive resources needed for information integration.On the contrary,Group L saw no such main effects whatsoever.ERPs results showed great difference between groups.Compared with Group L,Group H elicited more enhanced N400 at the midline and left posterior brain sites without relevant context,and larger LPC in midline and right hemisphere at central and posterior brain sites when given supporting context.Although the underlying mechanisms of metaphor processing may not differ between groups,the results suggest a stronger contextual effect for learners with higher English proficiency.