The Effect Of Segmental Information And Supra-segmental Information On 4 Year Old Children’s Spoken Word Recognition In Mandarin Chinese: Evidence From Eye Movements

Chinese is the most widely used language in the world,as a kind of tonal language,its has unique characteristics in pronunciation.The same pronunciation of Chinese words,if the tone is different,can form a different semantic.Spoken word recognition is an important cornerstone of spoken language comprehension in people’s daily life.The study of segmental information and supra-segmental information how to produced role in children’s spoken word is very few.Therefore,this study used eye movement method in visual world paradigm to explore the effect of Mandarin phonological information on spoken word recognition in 4 years old children.Specifically,two experiments were designed to explore the processing of segmental information such as consonant and vowel as well as supra-segmental information such as lexical tone during Mandarin spoken word identification in 4-year-old children.Participants listened to target words during looking at two pictures on screen.In experiment 1,participants listened to target words were written with bao/4 “leopard”,through headphones,for example,the displays contained vowel variable words(bang/4“stick”)and consonant variable words(hao/4“trumpet”).We manipulated word frequency and types of targets in the experimental stimuli,and analyzed the eye movement measures such as first fixation duration,total fixation duration,and number of fixations.The results showed that:(1)Consonants exerted more influence on children’s spoken word recognition than other phonological information,while the roles of vowel and lexical tones are similar during high frequency spoken word recognition;(2)There were no differences of total fixation duration between the consonants and lexical tonefor low frequency spoken word identification;(3)Both segmental information and supra-segmental information influence the processing of spoken word identification,but in different ways.Those results appear to provide evidence for Trace model in high frequency spoken word identification,while the processing of low frequency spoken word identification seems to support Cohort model.