漢語「於是」和「所以」的語義、篇章、語用研究與教學啟示

Abstract

本研究以Halliday的三大語言純理功能（metafunction）—概念（ideational）、篇章（textual）、人際（interpersonal）為研究的架構，並以真實語料為分析來源，探討現代漢語「於是」和「所以」在語義以及篇章語用層面的語言現象，最後亦嘗試將研究結果應用於教學當中。 本研究採用臺灣地區的兩個語料庫，分別為「聯合知識庫」（UDN Data）以及「中文詞彙特性速描系統」（Chinese Word Sketch Engine），蒐集「於是」「所以」語料各200筆，一共400筆語料進行分析研究，本論文以前人的研究為基礎，探討兩詞的語義特徵，並分析兩者篇章功能的異同。 首先，在語義層面，「於是」須符合「已然態」、「動態性」、「變化性」等語義特徵，但「所以」則無以上特徵。而且，「於是」和「所以」在語義上基本的差異為「變化性」，前者必須具「變化性」語義特徵，後者則具「變化性」語義特性的比例較低。 其次，在篇章語用層面，「於是」呈現的範距較小，而「所以」則因不同語料庫而有不同結果：在語體極為書面的語料庫中，其範距較小；在語體為書面類型與較偏向口語類型兩者兼具的語料庫中，則範距較大。再者，在連詞語義弱化程度上，「於是」的語法化程度較低，其絕大多數用法為「邏輯連詞」，而「所以」的語法化程度則較高，其作為「弱化連詞」用法的比例非常高，且亦有少數「話語標記」的用例。而當「於是」和「所以」作為「弱化連詞」使用時，語料顯示「於是」的語氣較弱，而「所以」則無明顯的語氣強弱之別。 最後，筆者根據本論文的研究結果，評析四套通行於臺灣的對外華語教材，並建議將「於是」和「所以」的研究結果應用於教學當中。

This research adopted the theoretical framework of Halliday’s meta-function, encompassing ideational, textual, and interpersonal functions (Halliday 1985, 1994). Empirical data was collected to explore the semantic, discourse, and pragmatic aspects of Mandarin Chinese “yushi” and “suoyi”. Moreover, the results of this research findings are applied for pedagogical purposes. This thesis is substantiated by empirical data, composed of 400 tokens of data from two sets of corpora in Taiwan, in which 200 tokens are collected from UDN Data and 200 tokens from Chinese Work Sketch Engine. Based on former studies related to “yushi” and “suoyi”, the present research further explores the semantic features between the two and their differences in discourse are distinguished. With regard to the semantic aspect, the semantic features of “yushi” are “realis (actualization)”, “dynamics”, and “new change”. However, “suoyi” has none of the above core features. Moreover, the essential semantic difference between “yushi” and “suoyi” is in the aspect of “new change”. “Yushi” highly involves new change, whereas “suoyi” only shows very low percentage. At the discourse level, “yushi” has a smaller scope, but “suoyi” exhibits varying results across different corpora. In the corpus with mainly written style, “suoyi” has a smaller scope, but in the other corpus with both written and spoken styles, “suoyi” exhibits a larger scope. Additionally, in terms of semantic reduction in the two connectors, “yushi” has undergone less of the grammaticalization process, while “suoyi” has undergone more of it. Most of the use of “yushi” in the present study is that of a logical connector, thus rarely exhibiting the phenomenon of semantic reduction. On the other hand, a large number of semantic reduction is found in “suoyi” in this study. Moreover, some of the instances of the connector “suoyi” appear to function as a discourse marker in the present research, although the frequency is quite low. At the pragmatic level, when “yushi” and “suoyi” function as semantically reduced connectors, the former carries a weaker tone of speaking, while the latter shows no particular preference. Lastly, in terms of pedagogical purposes, four L2 Chinese textbooks commonly used in Taiwan are reviewed based on the findings of this study. And a series of teaching suggestions are further provided.