An ERPs Study Of Possible Feature Extraction In Processing Chinese Hyperbole

There has been a long history for scholars to do researches on hyperbole at home and aboard. Hyperbole is often studied as a figure of speech which has a great popularity among scholars and writers. The researches of hyperbole can seldom be found in cognitive semantics and neuro-psychological linguistics. Even no study on possible feature extraction in processing Chinese hyperbole has been conducted from the perspective of ERPs.Nowadays, with the development of science and technology, the technique of ERPs provides a new empirical way for linguistic research. This paper, based on the ERPs study, with preliminary findings and the inadequacies of previous researches, aims to explore the psychological reality of possible feature extraction in processing Chinese hyperbole. Thus, the research questions is:Whether there exists a psychological reality of possible feature extraction in processing Chinese hyperbole?This experiment is completed in the Key Lab of Cognitive Neuroscience & Foreign Language Learning in SISU, Chongqing. Twenty four postgraduate native speakers of Chinese from SISU volunteer to take part in the experiment. And the experimental materials contain thirty one groups of sentences in total. The materials will be presented randomly and doubled for one time. After seeing and understanding the background sentence and key sentences, a probe word and declarative sentence will be given to judge whether they are real words or non-words.The experiment is designed with software E-prime 2.0, and this experiment also use it to present the materials. The data are recorded by the software Neuroscan 4.5. Then, make the off-line analysis of the ERPs data, and put the data into software SPSS 16.0 for statistical analysis. The research obtains the following results:(1) The reaction time of judging the probe word of the unrelated sentence is the longest among these three types of sentences, the literal sentence is the shortest, the hyperbolic sentence is longer than that of the literal sentence. And there is a slight difference between unrelated sentence and hyperbolic sentence.(2) During the time window of 280-550 ms, the N400 amplitude of the unrelated sentence is higher, the literal sentence is lowest, and the hyperbolic sentence is slightly lower than that of the unrelated sentence. Almost all the p value of hyperbolic statement and literal statement, literal statement and unrelated statement are less than.05, and there are significant differences between literal statement and the hyperbolic statement, literal statement and unrelated statement.Based on the results, it indicates that:(1)The process of possible feature extraction in hyperbole makes the judgment of the probe word in hyperbolic sentence become longer, thus there exists a process of possible feature extraction(2) Higher N400 amplitude of the Chinese hyperbolic sentence and the unrelated sentence means the processing of them involves more mental and cognitive resources than literal sentence. And the hyperbolic sentence just gets a slightly higher N400 amplitude than that of the unrelated sentence. The reason may be that all the experimental materials of the hyperbolic sentences themselves are far away from the normal expression, or deviating from the people's general cognition, so that it requires more neural activities and cognitive resources to get a better understanding at the later stage of semantic integration.To sum up, inferred from the experimental results, the conclusions of present study can be summarized as followings: this study has proved there exists a psychological reality of possible feature extraction in processing Chinese hyperbole. The results of the present study also provide evidence for psychological reality of the possible feature extraction and support the construal mechanism model of figurative languages proposed by Liao Qiaoyun.