太魯閣語鼻音韻尾之分佈與限制

Coda Condition in Truku

本文旨在探究太魯閣語鼻音韻尾之分佈與限制。文獻有關太魯閣語詞中鼻音韻尾的劃分主要圍繞在鼻音與詞綴以及其與後方輔音發音部位之關聯性。根據1813筆第一手語料，本文指出，太魯閣語的詞中鼻音韻尾除了可同時來自詞綴及詞根外，也可同時出現在發音部位相同或相異的輔音前。本文認為詞中鼻音傾向形成韻尾，只有位於重音節或特殊R輔音（即：喉音/ʔ，ħ/、響音/ɾ，ŋ，m，n，w/、及舌根擦音/ɣ，x/）前，鼻音才傾向形成韻首。本文指出，詞中鼻音之所以在重音節或特定輔音前才傾向形成韻首乃因重音節偏好能增進聽辨的開音節以及語言偏好在輔音和喉音、響音、及舌根擦音間加插元音所致，為語言常見之現象。本文以優選理論（McCarthy and Prince 1993，Prince and Smolensky 1993 [2004]）的架構，將所觀察到的通則具形化。

"This article provides a systematic investigation on coda conditions in Truku. Based on first-hand data, the findings of the study are demonstrated as follows. First, nasal codas come from both affixes and roots. Second, codas are the preferred positions for pre-consonantal nasals and that nasals only tend to form onsets when occurring in stressed position or before special segments, including gutturals /ʔ,ħ/, sonorants /ɾ,ŋ, m, n, w/, and velar fricatives /ɣ, x/. This study argues that nasals tend to form onsets in stressed syllables or before the special segments because stressed syllables prefer structures like CV that are perceptually salient (Smith 2002, 2004), and because crosslinguistically a sonorant or guttural tends to be separated from an adjacent consonant by vowel intrusion (Hall 2003). An OT analysis is proposed to account for the generalization."