漢語V得動詞重複結構的衍生

Abstract

本論文旨在探討漢語V得動詞重複結構的衍生。V得動詞重複句式的表層結構含有兩個型態相同的動詞，其中第一個動詞組由中心語和補語組成，第二個動詞組則由V-得和動後補語構成。前人研究中的衍生方法舉凡動詞移動（Cheng 2007, Bartos 2019）、基底生成（Hsu 2012）及其他生成法如動詞組並列（Fang 2006）和部分刪略（Meadows and Yan 2022）；然在這些分析中，「複製」時常被用以指涉表層出現重複動詞的描述性現象而非句法上的複製（參見Chomsky 1995, Nunes 2004）。此外，在這個句式中第一個動詞可以不出現且在不同動詞組又有不同表現的特徵鮮少被提及或討論。針對動詞重複結構，我提出平行合併的生成法（Citko 2005, 2011），重複的動詞位於共享節點，分別和補語以及動後「得」子句接合；而促使平行結構聚合的機制是對稱性的破壞，這可以透過第一動詞組的動詞組或名詞組移動（參見Moro 2000, Citko 2005）或轉化來達成（參見Epstein, Kitahara, and Seely, 2015）。我主張這些選項能解釋第一動詞在不同動詞的不同表現，有鑑於此，我循內部補語可提取性的基準，進一步將動詞分類為「開車」、「跑步」、「哭」三類；其中「開車」含有可提取補語、「跑步」包含不可提補語，「哭」則不帶補語。我也提出，補語的句法地位是影響動詞組進行對稱性破壞的關鍵—「開車」的補語屬名詞組故能提取，「跑步」的補語屬中心語，故不允許提取。總括而言，平行合併的分析本著聚合結構的句法性策略而生成重複的動詞並能遵循衍生經濟性，並且實證性地爲動詞組和其句式變異的關聯帶來嶄新的啟示。The aim of this thesis is to investigate the derivation of the V-de verb copying construction (V-de VCC) in Mandarin Chinese, which involves a double occurrence of verb instances. Specifically, the first VP in linearization contains a head-complement structure, and the second occurrence of VP is the V-de sequence. Previous approaches include verb movement (Cheng 2007, Bartos 2019), base generation (Hsu 2012), and other possibilities such as VP coordination (Fang 2006) and partial deletion (Meadows and Yan 2022). Among these analyses, the idea of copying is often loosely referred to as a descriptive term instead of syntactic copying (cf. Chomsky 1995, Nunes 2004). Besides, the property that different verbs behave differently regarding the flexible presence of the first verb instance is unaccounted.I propose a novel approach based on parallel Merge (Citko 2005, 2011), by which a shared verb head merges with a complement and the postverbal de sequence in a parallel manner. The convergence of such a parallel structure is motivated by symmetry-breaking, which can be fulfilled through VP or DP movement (cf. Moro 2000, Citko 2005) or Transfer (Epstein, Kitahara, and Seely 2015) of the first VP as an adjunct. I argue that these options can account for the syntactic behavior of different verbs, which I categorize based on complement extractability—kaiche type with extractable complements, paobu type with unextractable complements, and ku type without complements. I further argue that the availability of symmetry-breaking options lies in the syntactic status of the complement. Specifically, the kaiche type contains a phrasal DP and allows DP movement. In contrast, the paobu type contains a head noun, disallowing DP movement.The proposed analysis generates copies through syntactic-constrained strategies and displays derivational economy. Empirically, it provides new insight into the (im)possible variants of different VPs in V-de VCCs.