An Acoustic Analysis of Fricatives in Hakka Infant-Directed SpeechBased on Dapu Hakka in Dongshi Township

• Main Authors: Kuo Shu-Chu, 郭淑珠
• Other Authors: 鄭明中
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/13384637275890318715

碩士 === 國立聯合大學 === 客家語言與傳播研究所 === 100 === This acoustic study of fricatives of IDS (infant-directed speech) in Dongshi Dapu Hakka is designed to examine the similarities and differences between IDS and ADS (adult-directed speech) regarding Frication duration, spectral peak and consonant/ syllable ratio, also to view MIPhI Model applicability of IDS fricative in Hakka and to compare the performances of IDS in different identities . In this study, the observation of nine groups of adults and infants speaking Dongshi Hakka lasts for one year. The six life-common two-syllable words, consisting of Dongshi Hakka fricative [s, S] and corner vowels [i, a, u], are used as target words. The record was made when infants were 3, 6, 9, and 12 months old with the participating adults freely interacted with the infants and said the target words by showing toys and/or pictures provided by the researcher. With PRAAT voice analysis software, we conducted the acoustic analysis (measurement of the fricative duration and spectral peak) with the selected samples. The SPSS statistical software package was adopted for statistical processing and plotting graphs. The results are as follows: (A) In terms of frication duration, from the overall trend, fricative duration of [s, S] in the IDS are significantly greater than in the ADS, and the duration of [s] in IDS, although longer than [S] , but the difference is not significant. Further analysis shows that in the ADS there is significant difference in duration of [s] and [S] when followed by the vowel [i, a, u], but the order is not consistent, and difference of the duration of [s] and [S] in IDS is not significant; Durations of [s, S] in IDS are always longer than in ADS. Looking at the entire study period, except the 12th month, there is obvious difference in duration of [s] and [S] in ADS, regardless of which speech register, durations of [s, S] are similar, the difference is not significant. In addition, the trend shows durations of [s, S] are extended in IDS in the 3rd, 6th, 9th , and 12th month . (B)In terms of spectral peak frequency, from the overall trend, regardless of IDS or ADS, the spectral peak of [s] is higher than [S] and reaches the significant statistical level ; the spectral peak of [s, S] in IDS are significantly higher than in the ADS. Further analysis, when followed by the vowel [i, a, u] , the spectral peak of [s] is significantly higher than the [S] regardless of IDS or ADS; In addition, the spectral peaks of [s, S] in IDS are higher than in ADS, but the spectral peak of [s] in IDS is significant higher than in ADS only when followed by vowel [i] while spectral peak of [S] followed by the vowel [i, a, u] in IDS are all higher than in ADS significantly. Throughout the study period, no matter in which speech register, spectral peak of [s] is significantly higher than [S], and spectral peaks of [s, S] in IDS are higher than in ADS, the difference is more significant in [S]. Conclusion from the above analysis is that spectral peak of [S] in IDS expands more significantly than that of [s].(C) In terms of consonant/Syllable ratio, the overall trend, regardless of [s] or [S] ,consonant/syllable ratio in ADS is bigger than in IDS with significant difference. Further analysis, the consonant/syllable ratio of [s] followed by the vowel [i, a] in IDS is significantly smaller than in ADS and not so obvious if followed by the vowel [u]; the consonant/syllable ratio of [S] followed by vowel [i] in IDS is smaller than in ADS,yet,when followed by vowel [a, u], the difference is very limited.Checking on months, in most months, consonant/syllable ratio of [s] in ADS is narrowly bigger than in IDS except in 6th month; In 3rd and 6th month, consonant/syllable ratio of [S] in ADS is significantly greater than in IDS, however, the difference is not obvious in 9th and 12th month. Thus we found that consonant/syllable ratio is decreasing, vowels extend greater than consonants, but in general the vowels and consonants expand in accordance with a certain proportion,therefore,the consonant/syllable ratio is close in ADS and IDS. (D) In terms of different identities, among dad, mom and nanny, the fricative duration of [s] in ADS is about the same or slightly longer than [S]. In IDS, the fricative duration of [s] and [S] are close; the fricative duration of [s, S] in IDS expands significantly with all three identities. From the spectral peak : regardless of IDS or ADS, the spectral peak of [s] is significantly higher than [S]; the difference of spectral peak in IDS and ADS is obvious among dad, mom and nanny while the spectral peak of [S] is only obvious with mom. From the Consonant/Syllable ratio : mom has obvious different consonant/syllable ratio of [s, S] in both IDS and ADS, the dad and the nanny show no obvious different consonant/syllable ratio of [s, S] in both IDS and ADS. From the spectral characteristics and consonant/syllable ratio, it explains well enough the influence of mom in forming the characteristics of IDS. In the end, according to the above results, this article strives to deliver the descriptions and remediation for the appropriateness of the Mother-Infant Phonetic Interaction Model, also to provide a satisfactory explanation of performance differences of [s, S] in IDS .