霧台魯凱語的空間概念研究

Abstract

空間概念如何在不同語言中表達，一直是許多學術領域長期關注的主題。然而，相較於印歐語系等為人知悉的語言來說，較少研究著重在探討南島語的空間概念。其中，霧台魯凱語作為古南島語的主要分支之一，對其空間概念的研究也未曾在過往文獻中所提及。因此，本論文欲探討霧台魯凱語的空間概念，特別著重在跟空間相關的用法及其在動作事件的描述。 首先，本文的第一部分討論了跟空間相關的詞彙與詞綴，包含地方名詞 (Locative noun)、基本方位 (Cardinal directions) 等詞彙及空間詞綴 (Spatial prefix)。地方名詞按照 Zlatev (1997) 對於區域 (Region) 的分類，分成了五大類。每一類地方名詞皆可搭配前綴i- 及sa- 一同使用，並作為方位句中的述語。兩者只在上/下區域 (UP and DOWN Region)及裡/外區域 (IN and OUT Region) 才會產生差異。除此之外，另一個與空間相關的前綴tali-「方向」也可與前綴i- 搭配使用，指涉背景 (Ground) 中較寬廣的一個區域。跟賽夏語、郡群布農語、排灣語和賽考利克泰雅語等臺灣南島語相似，霧台魯凱語的基本方位詞之語意來源與太陽的東昇西落相關。最後，第一部分也討論了四種在Blust (2003) 的研究中所提及的空間詞綴，如mu-, i-, pu-, pi-，這些詞綴常見於排灣語、邵語及卑南語中，在霧台魯凱語的能產性 (productivity) 也相當高。前兩個詞綴mu-及i-分別與動態與靜態空間相關，後兩個詞綴pu-及pi-則與動態和靜態空間的使役語意相關。在這四個詞綴中，更可將mu-依照構詞和語意的差異分成五小類。 基於Huang 和Tanangkingsing (2005) 及Slobin (2004) 兩篇研究的觀點，本論文的第二部分進一步探討霧台魯凱語青蛙故事語料中的動作事件 (Motion event)。七個故事長度共計45分39秒，涵蓋900個語調單位。按照前者的分析架構，霧台魯凱語比許多Huang 和Tanangkingsing (2005) 所分析的南島語更為路徑顯著 (path-salient)，因為在故事中，跟動作事件相關的子句常使用路徑動詞 (Path verb) 作為唯一的動詞或作為連動結構中其中之一個動詞。按照後者的分析架構，霧台魯凱語可被視為相當典型的動詞框架化語言 (Verb-framed language)，因為他通過了Slobin (2004) 所提供的四個對動詞框架化語言的診斷測試。從本篇的研究結果顯示，兩篇研究的觀點相互吻合，因為動詞框架化語言一定是較為路徑顯著的語言。

How space is conceptualized in different languages has long been considered a highly researchable topic in various academic fields. However, studies exploring the spatial conceptualizations in Austronesian languages are relatively less than those investigating the same topic in Indo-European or other renowned languages. As one of the main offshoots of Proto Austronesian, spatial conceptualizations in Budai Rukai has not been thoroughly investigated yet. Therefore, the present thesis aims to study the spatial conceptualizations, especially spatial expressions and motion events in Budai Rukai, a Formosan language spoken in the mountainous areas in Pingtung of Taiwan. On the one hand, a detailed investigation into spatial expressions in Budai Rukai is conducted, including locative constructions with various locative nouns and cardinal directions as predicates and spatial prefixes. With respect to locative nouns, five types of Regions are discussed in terms of Zlatev’s (1997) classification. Data show that locative nouns of each type of Region can all be prefixed by i- and sa- to be predicates in locative constructions, which convey a slightly different meaning only in the UP and DOWN and the IN and OUT Regions. In addition, another prefix tali- ‘direction’ can be used with i- to denote a broader region of the Ground. Regarding cardinal directions, Budai Rukai exhibits the sun model, similar to some Formosan languages such as Saisiyat, Isbukun Bunun, Paiwan and Squliq Atayal. Lastly, four spatial prefixes, mu-, i-, pu- and pi-, attested in Thao, Puyuma and Paiwan by Blust (2003) are shown to be highly productive in Budai Rukai. Whereas the first two prefixes indicate motion and location respectively, the latter two prefixes indicate the causative of motion and location. In particular, the mu- prefix of motion can be classified into five types in Budai Rukai based on its morphological formation and semantic functions. On the other hand, this research also examines the motion events in Budai Rukai Frog narratives based on viewpoints from both Huang and Tanangkingsing (2005) and Slobin (2004). Seven narratives running to 45 minutes and 39 seconds for a total of 900 Intonation Units (IUs) formed the database. Firstly, our data show that Budai Rukai is more path-salient than the other six Austronesian languages examined in Huang and Tanangkingsing (2005) because Path verbs are frequently used in motion clauses, as either the only verb or one of the verbs in serial-verb construction. Secondly, Budai Rukai is also a purely verb-framed language from Slobin’s (2004) perspective because it passes four diagnostic tests for verb-framed languages. The results show that these two perspectives echo with each other because verb-framed languages are path-salient.