An ERPs Approach To The Possible Feature Derivation In Metaphor Processing Of Chinese Idioms

Metaphor is a distinctive language phenomenon frequently used in daily life.Many scholars have done follow-up studies based on the construal of metaphor and a series of construal models have been put forward.With the help of ERPs(Event-Related Potentials)techniques,many scholars have conducted researches on metaphor processing,among which empirical studies on the difference on metaphoric meaning processing and literal meaning processing are in the majority.Thus studies on metaphor processing are still a matter of great concern.In recent years,empirical studies on metaphor construal mechanism are weighted in conventional metaphors or metaphors without a dimension of familiarity.Liao Qiaoyun(2015,2018)proposed a comprehensive construal framework of Chinese semantic rhetorics,dubbed Annotation-Denotation Relevance-Inheritage Model(ADRIM).It is believed that there exists possible feature derivation in construing semantic rhetorics,which provides a new theoretical base for metaphor construal.Pan Xiang(2016)and Wu Dengju(2017)have verified this model and proved its psychological reality.However,novel metaphor,especially metaphor in Chinese four-character idiom(metaphor in Chinese idioms)is rarely taken into consideration in empirical studies on metaphor construal.Therefore,the present study has testified the electrophysiological reality of possible feature in the processing of metaphor in Chinese idioms with ERPs techniques.The present experiment is conducted in the Key Lab of Cognitive Neuro-science and Foreign Language Learning in SISU,Chongqing with 24 students of SISU participating in the experiment.A 2(probe types: related probe words and pseudo words)× 4(sentence conditions: novel metaphor condition,conventional metaphor condition,literal condition and unrelated condition)within subjects’ paradigm is adopted.All the materials are made in Chinese and stimuli are Chinese idioms with novel metaphoric,conventional metaphoric and literal meanings.Participants are required to carefully read and fully understand the sentences and judge the true or false of the probe words.The experiment is programmed by using the software E-prime 2.0.The behavioral data are recorded by the computer and the ERPs results are collected,recorded and off-line analyzed by the software Neuroscan4.5.Afterward,the ERPs data are analyzed by statistic software SPSS22.0.The results of the present study are demonstrated as follows:(1)As seen in behavioral data,response time of probe word judgment in unrelated condition(841.96ms)is longer than that in conventional metaphor(828.32ms)and literal conditions(823.83),and that in novel metaphor condition(862.07ms)is the longest,while the accuracy rate of the judgment of probe words in four conditions shows no significant difference.(2)ERPs data show the N400 amplitudes in literal condition is significantly smaller than that in novel metaphor condition [F(1,17)=19.468,p=.000<.05] and conventional metaphor condition [F(1,21)=6.726,p=.017<.05] over frontal and central regions,and the reason might be interpreted as deriving the possible feature between tenor and vehicle is more effort demanding.So the N400 amplitude which reflects processing effort in metaphor condition is more negative.Novel metaphor condition also differ significantly from conventional condition [F(1,21)=8.561,p=.008<.05] and unrelated condition [F(1,17)=10.056,p=.005<.05] over frontal,central and middle parietal areas.Besides,both hemispheres are activated in metaphor comprehension,but no significant lateralization is observed.(3)The LPC amplitude shows a marginal difference in condition × region interaction [F(1,21)= 1.977,p=.09 <.1].A reduced LPC in response to novel metaphor conditions relative to nonsensical conditions is observed over the frontal region.On the basis of the above results,the conclusions can be illustrated as follows:(1)Behavioral data results corroborate more mental effort is needed in possible feature derivation between tenor and vehicle and further testify the electrophysiological reality of possible feature derivation in metaphor comprehension.(2)ERPs data suggest possible feature derivation in metaphor condition is more taxing than that in the literal condition.The difference mainly lies in the front-central area and both hemispheres have been activated.(3)Familiarity modulates both N400 and LPC amplitude.The results of the present study provide the electrophysiological evidence for the possible feature derivation in processing metaphor in Chinese idioms and lend some support for the Annotation-Denotation Relevance-Inheritance Model proposed by Liao Qiaoyun(2015,2018).