| The merging of phonemes that the two phonemes are phonetically contrasted in one dialect of a specific language,but in the other dialect of the language these two phonemes are assimilated and cannot be distinguished as different phonemes.This phenomenon is called phonological merger or phonemic assimilation.Numerous studies has investigated the differences between individuals difficult in identity phonetic contrasts.Evidence from behavioral,electrophysiological and neuroimaging studies showed that in phonemes speech sounds processing,difference in phonetic perceiving stem from differences in language-specific memory traces.These studies have confirmed that phonological mergers not only show differences in behavioral responses of subjects,but also have different brain mechanisms for processing.Participants with phonological mergers were difficult to discriminate well between the two phonemes,and have longer reaction times in the discrimination unmerged group.Further ERP and neuroimaging studies indicated that phonological processing is different between two groups in the auditory cortex of the left hemisphere including superior and medial temporal areas,as well as bilateral cortical and subcortical regions.As reflected by the mismatched negative wave(MMN)of event-related potentials components.However,previous studies have explored the neural mechanism of this phenomenon in the auditory modality.It is not clear whether the phonological merger exists under the visual modality.In event-related potential(ERP)studies of phonetic contrasts,most of the research mainly concentrated on auditory MMN components and investigate neural responses in auditory cortex.This study use behavioral and event-related potential(ERP)to explore whether this phenomenon also exists in the process of visual word processing and the neural mechanism of this processing.The properties of vMMN rapid automatically detection novel stimuli and the widespread use of homophone character in Chinese provide the possibility for this study.Experiment 1 explored the merged group the brain activity and the neural mechanism of processing the n/l phoneme was combined with the text of the homonymous variant word.17 n/l phonemes merged subjects(9 females,8 males)and 15 n/l phoneme unmerged(8 females,7 males)were selected in this experimental,of which 2 data were invalid of merged group.Because they had too high accuracy in behavioral task.In the ERP experiment,subjects were presented with visual word in a passive oddball paradigm.There were two blocks and the block order was fully random between subjects: block 1(lán_deviant vs nán_standard);block 2(nán_deviant vs lán_standard).The subjects were required press a button as accuracy and quickly as possible when the target stimuli appeared.In the behavioral experiment,subjects were presented with n/l consonant Chinese character speech and asked to press a button as accurately and quickly as possible when hear speech sounds,meanwhile recorded the accuracy and reaction time.The brain electrical and behavioral data of the two groups were analyzed by SPSS18.0.The results of the study found that the accuracy of merged group significant lower than unmerged group,and reaction time is also longer.The results of ERP found that when nán was used as a bias stimulus,the mean amplitude of vMMN in the two groups was significantly different in the time window(350-450 ms),but this difference did not exist when lán was used as a bias stimulus.There were significant differences in the accuracy rate and response time between the two groups.This shows that the phenomenon of phoneme merger not only exists in the auditory processing process,but also exists in the process of visual word processing.At the same time,it also shows that speech sounds information affects the processing of written word to some extent.Failure of the merged group vMMN activation may be due to the lack of specific phoneme memory traces.The experiment 2 explored the difference of the neural mechanism of processing n/l phoneme combined with multiple tones in the visual word between merged group and unmerged group.In the experiment,15 n/l phonemes merged(8 females,7 males)and 15 n/l phonemes unmerged(8 females,7 males)subjects were recruited from experiment 1,and two of merged group did not meet the experimental requirements due to the trial were rejected over 40% in block;one subject of unmerged group rejected in the ERP experiment because of low accuracy in behavioral task.Visual word were presented to the participants in a passive oddball paradigm.In the block 1 of experiments,the /n/ consonant initials Chinese character were used as standard stimuli,and the /l/ consonant initials Chinese character were used as deviation stimuli and n consonant initials Chinese character were used as standard stimuli.The block 2 in turn,the n consonant initials Chinese character were used as deviation stimuli and l consonant initials Chinese character were used as standard stimuli.Two groups of subjects' ERP and behavioral data were analyzed by SPSS18.0.The results showed that the average amplitude of vMMN was not significantly different between the two groups of subjects in the ERP experiments.There were significant differences in accuracy rate and reaction times between the two groups in behavioral experiments.The lack of ERP results between two groups may be that the diversity of the stimuli leads the subjects to perceive the standard stimuli of the same initials as transformed novel stimuli during processing.When the deviation stimuli appear,the perception has been suppressed,deviation stimuli are no longer processed as new stimuli automatically process by brain.In this study,behavioral and ERP studies were used to explore the neural mechanism of processing the n/l phoneme merger in visual word.The results of the study found that the merged group showed low accuracy,longer response time compare to unmerged group in the behavioral phonetic discrimination experiments.This difference can be reflected by the amplitude of the electroencephalogram component vMMN.The EEG experiment found that the phonemes merged group was relatively weak activity in processing of the merging phonemes characters compared to the unmerged participants.And the ability of the merged group to automatically process the speech sounds information embedded in the visual word is poor than unmerged group.The findings indicate that phoneme merger exists in the processing of visual words,suggesting that the phonemic assimilation in merged dialect affects neural responses to phonological information embedded in visual words.Language understanding involves a series of language-specific phoneme trace processing.However,the memory traces of the n/l phoneme contrast is absent in the merged group mother tongue,and the system of the native speech sounds memory trace was determined by the early language experience,which led to the phonemes merger.This study extends the research scope of the phenomenon of phoneme merger.The research results have great practical significance and further deepen the understanding of phoneme merger. |
PDF Full Text Request