Year: 2012
Type: article
Abstract: How Japanese Learners Learn to Produce Authentic English Vowels(KITAGAWA) phones share a single category in L1 and L2 phonological space, which is called equivalence classification.The possibility of forming a new L2 phonetic category was explored in various studies.Lambacher, Martens, Kakehi, Marasinghe, and Molholt (2005), for example, conducted an experiment with native speakers of Japanese to assess how an L2 vowel category was created through training.They trained the participants for six weeks and performed an identification task and a production test of American English vowels, , as a pretest and posttest.According to the results, the identification of greatly improved, the ratings of the native speakers of American English confirmed the effectiveness of the training on and the accuracy of even without training, and the acoustic analysis of the first three formant frequencies showed the improvement of due to the training.This led them to conclude that the effectiveness of the training depended on the task, but the training contributed to the formation of an L2 new category for because of their auditory and phonetic distinctiveness from the Japanese .On the other hand, the similarity of to the Japanese hindered them from forming these categories.That is, the similarities and differences influence the formation of a new category, as predicted in the SLM.Another piece of substantial research on the formation of a new category by Japanese learners of English was done by Ingram and Park (1997).In their study, they performed three experiments with Japanese and Korean learners of English perceiving and producing Australian English vowels, How Japanese Learners Learn to Produce Authentic English Vowels(KITAGAWA) based on the other two methods.It has five vowels, (the Japanese back high vowel is also transcribed as , but will be used throughout this paper unless the previous studies applied it).Tsujimura (2006) described each Japanese vowel with reference to English as follows; the high front vowel is close to , but lacks in lip spread: the mid front vowel is a little higher than English (the mid front monophthong in American English is usually transcribed as , but will be used throughout the paper except the previous studies on American English): the low central vowel resembles , but is slightly higher: the mid back vowel is near , but is higher to some degree and a bit more front: the high back is like , but requires the lips to be unrounded.Although these descriptions might not apply to a particular accent of English or a mixture of English accents some Japanese learners speak because the descriptions refer to American vowels, they help predict the similarities and differences between the two languages.The second method of measurement was employed by Strange et al. (1998), where they directly examined the perceptual assimilation of American English vowels, , to the five Japanese vowels.Their results showed that when focusing solely on the quality, Japanese leaners assimilated American vowels into Japanese vowel categories like into , into , into consistently and into , into , into , into less consistently.Additionally, the consistency in assimilating into some Japanese vowel category varied from disyllable to sentence conditions.Nishi, Strange, Akahane-Yamada, Kubo, and Trent-Brown (2008), on the other, experimented in the opposite direction of the perceptual assimilation, i.e.Japanese vowels into American English vowel categories, .According to their analyses, each of the Japanese vowels was assimilated into fairly constantly, whether the tokens were long or short vowels, and they were given in citation or sentence forms. revealed the difference in the consistency of the assimilation depending on the length, but not on the form.Whereas some of the short and were each assimilated into and to a moderate degree, long and were highly consistently assimilated into and respectively-their native subjects confused and , so the results of pooled and as were reported.Nishi et al. also studied the similarities between Japanese vowels and American English vowels using the third method of measurement, a cross-language acoustic comparison.They found out that exactly like the perceptual assimilation, each of , and [ was consistently classified into , and only by spectral cues, no matter what length and what form. was, however, classified mainly into except long in citation form into .Likewise, most were classified into , but short in citation form was classified into . HypothesesThis study aimed to examine the English vowels produced by Japanese learners not specifically How Japanese Learners Learn to Produce Authentic English Vowels(KITAGAWA) level, p < .05,and were categorized into homogeneous subsets with it.The equivalent number of the subjects was also selected from the bottom in the ranking, who were grouped into low JL.The data was then submitted to two statistical tests on IBM SPSS Statistics 19.One was a multiple discriminant analysis.After observing how each token was classified within each of four groups, this analysis, was furthermore, carried out with the BN and AN data as the input set and the high JL and low JL data as the test set.This made it possible to see directly how the tokens produced by high JL and low JL were classified under the classification rules of BN and AN.The other test was a 4 × 2 mixed design ANOVA, which illustrated whether the location of each vowel category differed significantly between high JL and low JL, referring to the difference from BN and AN.The difference of the classification and location implies the developing process in the English vowel learning that the Japanese learners of English would follow. Results and discussion BN and AN: The Distribution of F1/F2 and the Correct Classification Rate
Author: Aya Kitagawa
Language: Japanese
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W396212057