從優選理論分析南王卑南語重疊詞

Abstract

本篇論文目的在重新分類南王卑南語重疊詞，並從優選理論分析之。文獻中提到南王卑南語呈現豐富的重疊詞形式，依據前人文獻，卑南語重疊詞可分成七類：詞彙化重疊、第一音節重疊、連續重疊、右向重疊、Ca- 重疊、CVC 重疊與CV.CV重疊。本研究經重新分析語料，認為卑南語重疊詞分類應分為六類：詞彙化重疊、第一音節重疊、連續重疊、完全重疊、Ca- 重疊與右向重疊。此分類認為文獻中提到的CVC重疊應分析成C.CV重疊，且C.CV重疊應與CV.CV重疊共屬右向重疊一類。本文論述優選理論中的標記性制約（markedness constraint）及信實性制約（faithfulness constraint）可為卑南語右向重疊與Ca-重疊提出合理解釋。右向重疊中有一類特殊的重疊詞必須遵守韻腳條件制約，形成CV重疊詞。此外，優選理論中的標記性制約（markedness constraint）可預測 Ca- 重疊中的預設元音（fixed vowel)。最後，本論文採用ROE模型（Rank-Ordering Model of EVAL）解釋牽涉指稱人類複數型的重疊詞時可能產生的變異現象（variation）。ROE模型可預測Ca- 重疊詞與右向重疊詞的使用頻率，後者的頻率高於前者。

The thesis aims to re-categorize the reduplication in Nanwang Puyuma and provide an analysis within the framework of Optimality Theory. Previous works dedicated to Puyuma phonology have discovered diverse reduplicant patterns, including lexicalized root reduplication, first syllable reduplication, serial reduplication, rightward reduplication, Ca- reduplication, CVC- reduplication, and CVCV- reduplication. The thesis argues that CVC- reduplication should be re-classified into CCV- reduplication. Furthermore, CCV- reduplication and CVCV- reduplication should be subsumed under rightward reduplication. An OT analysis is presented to show that the reduplicant shape, copying direction, and reduplicant placement are attributed to interactions between markedness and faithfulness constraints. An uncommon undersized CV reduplicant is a result of obedience to RimeCond. The fixed vowel in Ca- reduplication can also be accounted for from an Optimality Theoretic perspective. Finally, with regard to the special case where two reduplicant patterns share the same function, Rank-Ordering Model of EVAL (ROE) is able to help predict the variants and their relative frequencies.