全民英檢中高級與高級寫作中從屬關係命題結合之研究

Abstract

英文寫作中很重要的部分在於如何將想法用適當的文法結構表達。然而， 以英語為外語的習得者通常未曾被教導過如何選擇適當的文法句型來精確地陳述命題 (proposition)。文獻上也鮮有以修辭及篇章分析的角度來討論如何將抽象想法轉化為合邏輯的命題，並以合乎文法的書面形式表達。通常一個命題可以用一個句子來寫，但更多的情況是把多個命題整合在一個句子裡。本論文旨在探究英語為外語習得者如何在作文當中，使用文法結構來結合具有從屬關係之命題。分析的資料來自通過全民英檢中高級與高級寫作測驗的100篇作文。根據Halliday 與 Hasan (1976) 及 Stalter (1978) 提出的四種語義關係與七項文法結構為理論架構。這四種語義關係為闡述、因果、時間與逆轉；七種文法則為形容詞子句、副詞子句、名詞子句、名詞化結構、分詞結構、不定詞、介系詞片語。兩位具有語言學背景的分析者先就100篇作文中挑出內含兩個以上命題的句子，然後分析句子內命題之間的語義關係，再檢視其語義關係所使用的文法結構。\n研究結果顯示第一、闡述關係為資料中最常使用的命題結合關係，在中高級與高級中都佔了五成以上。此外，闡述與因果則在全部的語義關係中佔了90%。第二、闡述關係最常用不定詞、名詞子句及介系詞片語來表達。這些結構通常用來表示一個人的態度、介紹新訊息及總結重點。因果類則最常使用副詞子句及介系詞片語來陳述。這兩個結構用於表示動作發生的理由、條件及目的。時間與逆轉類也都大量使用副詞子句及介系詞片語來表達動作的時間框架或背景。第三、在闡述及因果兩類中，高級與中高級的應試者傾向使用不同的文法結構。在闡述類中，高級的應試者使用更多種類的介系詞片語及名詞化結構；在因果類中，中高級的應試者使用較多的限定類形容詞子句及動作類名詞化結構。第四、藉由分析四種語義結合所使用的句法結構，本研究顯示出闡述類的語義結合通常在詞組內發生；因果、時間及逆轉三類則大多是子句的組合。最後，本論文建議在寫作教學中，命題結合應從修辭的角度切入，並向學生解釋文法結構的篇章功能。未來的研究則可進一步探討不同英語能力者使用不同文法結構來結合相同語義的原因，且比較英語為外語習得者及母語者結合命題的方式。

In the process of translating ideas to language, writers search for syntactic structures, or sentence patterns, to indicate propositions. Nevertheless, the choice of an appropriate sentence structure to match the intended meaning is a challenge to L2 writers. Little research has been conducted from rhetorical and discourse perspective to explore how L2 writers transform abstract ideas into logical propositions, and further translate these propositions into sentences. To respond to the issue, the present study investigates how L2 writers use language forms to combine propositions with subordinate relations in English writings. The analysis data are 100 writings of test takers passing the high-intermediate and advanced levels in GEPT. The analysis framework includes both semantic relations between subordinate and superordinate propositions and their syntactic structures. The semantic analysis is based on Halliday and Hasan (1976) and Stalter’s (1978) coherence relations that include four basic semantic relations of elaboration, causal, temporal and adversative used or found in text. The syntactic analysis covers seven syntactic structures of adverbial, noun and relative clauses, participles, prepositional phrases, infinitives and nominalizations. Two analysts with TESOL background code the writings with the two systems: the first is the semantic relations between the subordinate and superordinate propositions in sentences; the other is the syntactic structures used to realize the semantic relations.\nThe results show, first, elaboration accounts for over 50% of the total semantic relations at both proficiency levels. Along with the causal relation, they constitute around 90% of the semantic relations. It suggests that elaboration and causal are the major relations employed in the data. Second, in terms of the grammatical structures employed to perform the four semantic relations, elaboration is mainly represented by infinitives, noun clauses and prepositional phrases. The structures are largely used to show ones’ attitude, introduce new ideas or summarize main points. The causal relation is dominantly fulfilled by adverbial clauses and prepositional phrases to manifest the reason, condition, method and purpose of an action. As for temporal and adversative, both are mostly performed by adverbial clauses and prepositional phrases to give either a time frame or a contrasting idea as background. Third, there are some distinct differences in the structures used to realize the relations of elaboration and causal between the two proficiency levels. In elaboration, writers at the advanced level used more diverse types of prepositional phrases and nominalizations. For causal, the writers at the high-intermediate level employed more restrictive relative clauses and more gerundive nominals to indicate the cause. Finally, the analysis of the syntactic levels at which proposition combination occurs reveals a hierarchical scale of the semantic relations. It is found that elaboration in the data is prone to happen at the within-phrase level, while causal, temporal and adversative are mostly composed of the above-phrase combination. The findings suggest that proposition combination involves rhetorical consideration and discourse functions. From the pedagogical viewpoint, syntactic structures used to combine propositions should be explicitly explained to learners in L2 writing classes. Future research can be conducted to further explore the contributing factors to the differences between two proficiency levels in proposition combination. A cross-cultural comparison can also be done to see if and how Mandarin learners of English and L1 English writers use different strategies to combine propositions.