台灣韓中空耳之音韻分析

Abstract

本研究主要探討台灣的韓文歌曲中以空耳方式對譯成中文的音韻現象。藉由11首韓文歌中收錄955個音韻詞（2690個音節）作為分析之語料。本研究之語料顯示，韓語源詞會調整成中文之語音系統（Mandarin phonotactics），而其中8.07%使用注音符號或英文字且違反中文之語音系統。語音違規（phonotactic violation）還可分為兩類：一、按照韓語語法系統違反中文語音系統。二、按照中介語語法違反中文語音系統。\n\n而大部分音譯符合中文之語音系統，結果如下：緊張閉鎖音（tensed obstruent）做鬆弛化，並以中文的不送氣閉鎖音取而代之。韓語中的音節尾 /p、t、k、m、l/ 被刪除，或被中文之合法音節尾 /n、ŋ/ 所取代。滑音 /j/ 會插入清齦顎塞擦音（post alveolar affricates）/tɕ、tɕh、tɕ’/的後面。韓語元音 /ɯ/ 被降音到 /ɤ、ə/，而單元音 /o、ʌ/ 和 /e/ 被雙元音化為 [ou] 和 [ei]。除了按照中文語音系統音譯的現象之外，本文還觀察到隨著韓語語音系統音譯的 CV序列（sequencese），也觀察到隨著中文與韓文之間的中介語音譯的序列。隨著韓語（L2）語音系統的序列是：軟齶音（velars）/ k、kh、k’、h / 和 [+high, -back]元音 /i/，以及牙音/n/ 和 [-high, +back]元音 /ʌ、o/。這兩個序列會保留韓文語音系統的CV組合。另一方面，諸如軟齶音 /k、kh、k’、h/ 和唇音（labial）/p、ph、p’、m / 等非舌冠音（coronal）在母音/jʌ/前出現，會隨著中介語語音系統音譯。此外，齒齦音/s、s’/ 和 元音/e/ 會隨著中介語語音系統輸出。\n\n本研究基於優選理論（Optimality Theory）和Ranking Ordering Model of Eval 探討韓中空耳之音譯變體（variations）。結果顯示，在韓文文法中做top-ranked的制約會被降級，這支持了Miao（2005）和Broselow（2004）的中文語音系統比韓文語音系更重要的作用。然而，一些音譯違反中文之語音系統，以保持類似於源詞的音。因此，結果反映出韓文語音系統和中介語語音系統也影響韓中空耳音譯過程。

This thesis investigates the Korean-Mandarin Kong-er in Taiwan. 955 phonological words (2690 syllables) from 11 songs are collected. The major part of the corpus shows that Korean source words are adjusted to the Mandarin phonotactic system, while 8.07% of the whole show phonotactic violation by using either Zhuyin (the Chinese transliteration system for Taiwan Mandarin) or English. The phonotactic violations also can be divided into two groups: one for violations that follow Korean grammar and the other for violations that show the emergence of an interlanguage grammar. \n\nThe majority of the patterns of the adaptations conform to Mandarin phonotactics. Tensed obstruents are de-tensed and replaced by unaspirated counterparts in Mandarin. Illicit codas such as /p, t, k, m, l/ in Korean are deleted or replaced by the Mandarin licit codas /n, ŋ/. The glide /j/ is prone to be inserted after the alveolar-palatal affricates /tɕ, tɕh, tɕ’/, which follows Mandarin phonotactics. The Korean /ɯ/ vowel is prone to being lowered to /ɤ, ə/, and the monophthongs /o, ʌ/ and /e/ are diphthongized as [ou] and [ei]. \n\nApart from the adaptations that follow Mandarin phonotactics, this thesis also observes CV sequences that follow Korean phonotactics and also an interlanguage between Mandarin and Korean. The sequences that follow Korean (L2) are the velars /k, kh, k’, h/ and the [+high, -back] vowel /i/, and the dental /n/ and the [-high, +back] vowels /ʌ, o/. These two sequences are prone to preserving CV manners and features. On the other hand, non-coronal consonants such as the velars /k, kh, k’, h/ and the labials /p, ph, p’, m/ mostly induce phonotactic violation following interlanguage phonology before /jʌ/. Also, the alveolars /s, s’/ and /e/ induce interlanguage output, which is not allowed in either the L1 or the L2. \n\nThe patterns and sequences are analyzed based on the Optimality Theory and Rank Ordering Model of EVAL to account for the variations. The results explain the fact that the Korean top-ranked constraints are demoted in the adaptation, supporting both Miao (2005) and Broselow (2004) that Mandarin phonology plays a more important role than Korean phonology does. However, some of the adaptations violate Mandarin phonology in order to preserve a sound similar to that of the source word. Thus, the results reflect that Korean phonology and the emergence of an interlanguage have a partial influence on KM Kong-er adaptation.