| Budget Amount *help |
¥2,500,000 (Direct Cost: ¥2,500,000)
Fiscal Year 2006: ¥500,000 (Direct Cost: ¥500,000)
Fiscal Year 2005: ¥2,000,000 (Direct Cost: ¥2,000,000)
|
|---|
| Research Abstract |
Teachers of foreign languages often have their students repeat their target pronunciation individually or in chorus. Because repetition after a teacher and repetition with a teacher are the most widely used ways of pronunciation practice in a language classroom, the aim of this paper was set to measure the effectiveness of the two types of pronunciation practice, using four experiments which focus on the temporal factor. In Experiment 1, Japanese learners of English were asked to imitate English and nonsense syllables, and their reaction time was measured. Average latency before launching repetition fell between 600ms and. 900ms, varying with the phonological length of the test words. In Experiment 2, pairs of nonsense and English words, which included difficult phonemic distinctions for Japanese learners of English such as between /l-r/ and /v-b/, were both a-repeated and w-repeated. In Experiment 3, pairs of English sentences, which had strong-weak and weak-strong stress patterns, we
… More
re a-repeated and w-repeated. In both experiments, learners' practices were recorded and evaluated by British teachers of English. In Experiment 2, the results of nonsense words revealed that there was no significant difference in naturalness between a-repeat and w-repeat. However, a-repeat surpassed w-repeat in Experiment 3. In Experiment 4, subjects were presented the identical stress-patterns of test sentences but reproduced with short sinusoid waves. The learners a-repeated and w-repeated the patterns by pronouncing the monosyllable "ta." The results showed no difference between a-repeat and w-repeat, which implies that repetition of sentences is different from that of nonsense words.
The naturalness of English sentences pronounced by learners was examined in Experiment 3, and the rhythm patterns, generated with spotted sine waves based on the power contours of the original test sentences, were investigated in Experiment 4. The results in Experiment 3-were based on naturalness scores given by native British listeners; the scores obtained in Experiment 4 were calculated from differences in time of speakers' productions. Comparison of the timing structure of the test sentences repeated by the participants in Experiment 3 with those of the results of Experiment 4 was done in order to ascertain to what degree is the timing structure of the original test sentences maintained by learners.
Two more experiments succeeded in proving the effect of Time Order Error (TOE) on perception of Japanese geminated consonants. Test words considered to evoke the TOE were synthesized by adding a short syllable ba to the bisyllabic contol word baba. In Experiment 1, the test word bababa had a short-long timing structure among three ba syllables, which was considered to lower the perceptual threshold of geminate consonants. Fundamental frequencies of the same test words were manipulated in Experiment 2. The result indicated that stepped pitch counteracted the effect of TOE. Less
|
