摘要
目的:使用事件相关电位技术考察人脑在视觉通道下汉语词汇识别中汉字正字法的作用。方法:实验于2004-04/07在徐州师范大学语言研究所事件相关电位实验室完成。①实验对象及分组:选择16名大学生(8女,8男,18～22岁)经书面同意后进行词汇判断实验。单音节词汇判断为测试组,双音节词汇判断为对照组。②实验语料:选自1986年北京语言学院出版社出版的《现代汉语频率词典》日常生活常用语部分,选词包括单音节词和双音节词。单音节假词通过调换真字相应位置的一个部件得到,双音节假词通过调换真词中一个音节得到。真词和假词的笔划数无显著差异。刺激词共264个,每个刺激呈现时间为200ms,刺激间隔在2.0￣2.5s。③实验过程:分两个序列,受试者被要求在保证反应正确的前提下尽可能迅速地进行按键操作。④实验评估:采用NeuroScan公司64导电极帽10/20系统记录脑电,离线分析处理事件相关电位数据。结果:名受试者进入结果分析。①行为学数据结果:真词的反应时均显著短于假词(P<0.05),真词的正确率均显著高于假14词(P<0.05)。②脑电数据结果:N400的波峰值假词显著低于真词(P<0.05),但只有测试组中N400的波峰单音节假词迟于真词出现;测试组中,560～660ms时间窗口内P600的平均幅值单音节假词显著高于真词(P<0.05),对照组中600～700ms时间窗口内P600的平均幅值双音节假词显著高于真词(P<0.05)。③假词减真词的差异波分析结果:测试组和对照组中均出现Nd400,Pd600,但Pd250却只在前者出现,且主要在额中央区、中央顶区分布;对照组中Nd400波幅值较小,脑区分布范围也较大,主要分布在双侧顶中央区、右半球颞顶区,测试组的Nd400主要在右半球颞顶枕区分布;测试组Pd600波幅值较大,脑区分布范围较广,主要分布于右半球中央及顶枕区,对照组则主要在右半球顶枕区分布。结论:①在视觉通道的汉语词汇判断实验中,差异波分析可观察人脑在词汇识别中正字法的作用。②Pd250是视觉词汇识别中汉字正字法作用的特异性成分,表明汉字正字法作用在刺激词呈现约250ms后开始体现,该成分可能反映了视觉词汇识别中人脑对汉字部件组合规则和汉字拼读规则的加工过程。③双音节词汇判断中Nd400可能反映了视觉词汇识别中人脑对汉字部件语义组配的加工过程;Pd600则可能反映了人脑重新分析汉字部件语义组配的加工过程。
AIM：To explore the effect of Chinese character orthography knowledge in visual lexical decision with an event-related potential （ERP） technique. METHODS： The experiments were carried out in ERP Laboratory, Institute of Linguistics, Xuzhou Normal University from April to July in 2004.①Totally 16 undergraduate students （8 males and 8 females, aged 18-22 years） wore enrolled as healthy examinees to make visual Chinese lexical decision after having signed the informed consents. The test group＇s task was monosyllable lexical decision, and the control group＇s task was bisyllable lexical decision.②The stimuli included monosyllable words and bisyllable words, which were selected from the part of common words used in daily life in Lexicon of Modem Chinese Frequency （Press of Beijing Languages College, 1986）; monosyllable pseudowords were made by replacing corresponding local radicals of real words, and bisyllable pseudowords were made by replacing corresponding local characters of real words. There was no significant difference in the strokes between real words and pseudowords. The experiment was divided into 2 sequences, which totally included 264 stimuli, each was displayed on the screen for 200 ms, and stimulus onset asynchrony was randomized between 2 and 2.5 s. ③The examinees were instructed to press the right keys as quickly as possible.④The ERPs wore disposed off-line analysis, brain electricity was recorded by American Neuroscan electric conduction evoked potential workstation. RESULTS： Experimental data of 14 examinees were all involved in the analysis statistics.①Behavioral data： Compared with pseudowords, the significantly shorter reaction time and higher accuracy were found for the real words （P 〈 0.05）.② Brain electricity data： N400 amplitude peak was significantly lower for pseudowords than for real words （P 〈 0.05）, and was later for monosyllable pseudowords than real monosyllable words only in the test group. Higher P600 mean voltages wore found significantly for pseudowords compared with real words （P 〈 0.05）, both in the test group （560-660 ms） and the control group （600-700 ms）.③Difference wave （pseudowords subtracted real words） analysis： Nd400 and Pd600 wore found in both the test and control groups, while Pd250 was only found in the former and distributed at the electrodes of the frontal-central and central-parietal area. Smaller amplitude of Nd400 with broader distributed areas were found in the control group, while larger amplitude of Pd600 with broader distributed areas were found in the test group. Nd400 was distributed in the right temporal-parietal-occipital area in the test group, but bilateral parietal-central area and right temporal-parietal area in the control group; while Pd600 was distributed mainly in right central and parietal-occipital areas in the test group, but right parietal-occipital areas in the control group. CONCLUSION： ①During the visual Chinese lexical decision, the difference wave analysis revealed the effect of orthography knowledge in word recognition. ②Pd250, as a specific component of visual Chinese lexical decision, indicated that the effect of Chinese character orthography came into being about 250 ms after the stimulus onset. It might reflect the processing of the rules of both Chinese character radical construal and Chinese character spelling by the human brain. ③Nd400 may reflect the processing of the semantic relations among the Chinese character radicals, while Pd600 may reflect the processing of the re-analysis of semantic relations among Chinese character radicals in the human brain.
出处
《中国组织工程研究与临床康复》
CAS
CSCD
北大核心
2007年第26期5097-5101,共5页
Journal of Clinical Rehabilitative Tissue Engineering Research
基金
国家社会科学基金重点资助项目(01AYY004)
江苏省"青蓝工程"(科技创新团队)资助项目(QL200618)
江苏省哲学社会科学十一五规划资助项目(06JSBYY006)
江苏省教育厅哲学社会科学基金资助项目(06SJB740002)~~
作者简介
顾介鑫,男,1977年生,江苏省连云港市人,汉族,南京师范大学文学院在读博士,讲师,主要从事神经语言学研究.guzhongzhi@163.com
杨亦鸣,通讯作者,教授,徐州师范大学语言研究所,江苏省徐州市 221116
