A Speech Learning Model Perspective Of English Monophthongs By Native Speakers Of Mandarin And Tibetan

According to Speech Learning Model (SLM) proposed by Flege (1995),successful second language production relies on the construction of new phoneticcategories for the segments in L2but do not have any counterpart in the L1and thatL2learners will be likely to establish a new category for the L2sound if they can beaware of the phonetic differences between the L1and L2sound systems. Similarsounds in L1and L2vowel system are difficult to produce.Under the framework of Flege’s (1995) Speech Learning Model, this paperinvestigates the acoustic features of monophthongs produced by20Chineseuniversity students who speak Chinese Mandarin(5male and5female) and Tibetan(5male and5female) and who have learned English for at least8years.With the help of PRAAT software, the first and second formant values of seventarget vowels produced by subjects are measured in connected speech. Afterregulation, data measured by PRAAT are compared with the standard data measuredby Deterding(1997) in his study for vowels produced by British SouthernBroadcasters. To make the comparison more theoretical and reliable, vowel spaces oftarget vowels produced by Chinese students--students whose first language isstandard Mandarin Chinese and whose first language is Tibetan are drawn.Results of the comparative studies are as follows:For Chinese Mandarin students results are:1. Under the influence of StandardChinese vowel system, English vowel space for Mandarin Chinese is relativelydispersed.2. Female Chinese Mandarin failed in the production of vowel contrast/i:/and/i/./(?)/and/u:/produced by female Chinese are less fronter than that of RP/(?)/and/u:/; the distinction of vowel contrast/i:/and/i/produced by male ChineseMandarin is vowel frontness,/ɑ:/is much fronter and lower than RP/ɑ:/. Moreover,both/(?)/and/u:/produced by male Chinese Mandarin are lower than that of RP.For Tibetan students results are:1. English vowel space for Tibetan iscentralized and smaller than that of RP.2. Female Tibetan cannot distinguish vowel contrast/e/and/(?)/very well. For English vowels produced by female Tibetan,/i:/islower than that of RP,/i/and/ɑ:/are fronter than RP/i/and/ɑ:/. What’s more,/u:/isless fronter than RP/u:/; male Tibetan recognize vowels/e/and//successfully. ForEnglish vowels produced by male Tibetan,/i:/is less fronter than RP/i:/,/ɑ:/is muchfronter than RP/ɑ:/, and/u:/is less fronter and lower than RP/u:/.The results that female Chinese Mandarin students and female Tibetan studentscannot successfully distinguish vowel contrasts/i:, i/and/e,/respectively confirmthe hypothesis proposed by Flege in his Speech Learning Model that L2learnershave difficulties in differentiating similar vowels in their first and second languagevowel categories for the interruption of the first language and the difficulty degreevaries with the similarity degree of two counterpart vowels in the first and secondlanguage vowel systems. Influenced by vowel systems of Chinese Mandarin andTibetan (Lhasa), vowels produced by Chinese students are L1-like. Other similarvowels produced by Chinese students formed a new L2vowel category that differsL1vowel system. Which is in accord with Flege’s hypothesis that L2learners arelikely to establish a new category for the L2sound if they can be aware of thephonetic differences between the L1and L2vowel system.