阿美語的“語態”標記： 一種角色指稱語法的分析

“Voice” Markers in Amis: A Role and Reference Grammar Analysis

本文採用角色指稱語法 (Role and Reference Grammar, RRG) 的理論架構，重新分析阿美語的語態系統（或稱焦點系統）。文中首先指出，一般所認為的阿美語的四種語態（即主事語態、受事語態、工具語態、處所語態），實際上包含兩種系統：語態系統及應用句式系統。前者僅包含兩種語態，即行動者語態 (actor voice, AV)，與承受者語態 (undergoer voice, UV)，後者則包含工具應用句式 (instrument applicative) 與處所應用句式 (locative applicative)。應用句式可視為是一種特別的承受者語態。就RRG 的聯結系統而言，這兩種系統分別作用於兩種不同的聯結階段，應用句式標記指出特別的承受者選擇，而語態標記則是指出該句的特權語法論元 (privileged syntactic argument)是行動者還是承受者。本文接著將三個最常見的語態標記形式mi- (AV)，ma-(AV, UV)，及 -en (UV) 加以解構，並以邏輯結構呈現其語意內容來解釋這些標記的特性，像是具有隱含的時貌意義或是主事性指標等。本文指出前綴 mi- 的語意是一個活動述語，並帶有一個可省略的移動性／目的性成分，後綴 -en 的語意則分析為一個具主事性的成就述語標記，而 ma- 的語意型態最為複雜，可用四種邏輯結構來呈現：活動述語、結果狀態、一般狀態、及活動／使動成就述語。

This paper reanalyzes the voice markers in Amis within the RRG framework. The traditional four-voice system in this language is first regrouped into a voice system and an applicative system. The former contains two voice operations: actor voice and undergoer voice, while the latter subsumes the instrument applicative and the locative applicative constructions. The applicative constructions follow the undergoer voice pattern. These two systems operate at the two different phases of the linking from semantics to syntax; the applicative markers indicate a marked choice of the undergoer, while the voice markers influence the selection of the privileged syntactic argument. Three most productive voice forms, mi- (AV), ma(AV and UV), and -en (UV), are further decomposed into logical structures to account for their derivational functions and other semantic features such as TAM marking and agentivity indication. The prefix mi- contains an activity predicate with an optional motional/purposive component. The suffix -en is analyzed as an agentive accomplishment marker. As for the form ma-, it exhibits four different logical structures: activity, result state, transient/plain state, and active/causative accomplishment.