Developing objective indexes of phonological similarities for two-kanji compound words used in both the Japanese and Vietnamese languages

本研究では，日越両言語間で使用される2字漢字語の音韻類似性を，音韻的距離と音素類似性の2つの指標で客観的に計算した。日越で共通に使用される2字漢字語（訓読みを含まない）1,475語の音韻的距離は，Rで提供されたcba パッケージ（Buchta & Hahsler, 2016）で，音素類似性は，Rのphonosimパッケージ（Yu, 2016; Version 0.1）で，自動的に計算した。その結果，日越両言語間の音韻的距離の平均は6.06で，標準偏差は2.52であった。音韻的距離が小さければ小さいほど，音韻類似性が高くなることを示す。一方，日越両言語間の音素類似性の平均は0.5で，標準偏差は0.2であった。音素類似性は0から1まで分散し，1に近いほど，音韻類似性が高いということを示す。音韻的距離と音素類似性の相関係数は非常に高く（N=1,475, r=-0.92, p
In this study, phonological similarities of two-kanji compound words used by both the Japanese and Vietnamese languages were objectively calculated by two indexes of phonological distance and phoneme similarity. The phonological distance of 1,475 (71,67%) of two-kanji compound words used in both languages (71,67%) was automatically calculated by R cba package (Buchta & Hahsler, 2016), and by the R phonosim package (Yu, 2016; Version 0.1) for phoneme similarity. The mean of phonological distance was 6.06 with a standard deviation of 2.52. This index indicates that the smaller the phonological distance, the higher the phonological similarity. On the other hand, the mean of phoneme similarity between two languages was 0.5 with a standard deviation of 0.2. Phoneme similarity was dispersed from 0 to 1, and the closer it was to 1, it showed that phonological similarity was higher. The correlation between phonological distance and phoneme similarity was very high (N=1,475, r=-0.92, p