試論表永恆性質及表短暫狀態關係子句的次序問題及優越效應

The Order of Stage-level and Individual-level Relatives and Superiority Effects

這篇論文提出證據證明Larson&Takahashi(2004)、DelGobbo(2005)及Hsieh(2005)所主張的名詞組雙層修飾結構並無法真正解釋中文表永恆性質及表短暫狀態關係子句的句法分布，因此提出另一解釋，主張表永恆性質的關係子句是名詞論元結構的一部分，而表短暫狀態的關係子句則是不折不扣的附加語。附加語在結構上總是處於論元之外，因此表永恆性質的關係子句必須比表短暫狀態的關係子句更靠近中心語名詞。不過，中文的關係子句也可以移位至DP指示語的位置，但是移動時必須遵守句法上的最短距離條件，結果，移位後的關係子句的結構高低位置仍然會保留未移動前的結構高低位置，顯現了所謂的優越效應。

This paper argues that the two-layer modification of noun phrases as proposed in Larson & Takahashi (2004), Del Gobbo (2005), and Hsieh (2005) cannot account for the distribution of the order of stage-level and individual-level relatives in Mandarin. Alternatively, it is suggested that I-level relatives must occur closer to the head nouns than S-level relatives because I-level modifiers are arguments of “augmented nouns”, whereas S-level modifiers are true adjuncts. It is always the case that adjuncts are base-generated outside arguments. It is also shown that relative clauses in Mandarin may be attracted to the specifier position of DP. When they move, however, they have to obey syntactic economy conditions such as Shortest. Consequently, the hierarchical positions of an S-level and I-level relative after the movement must preserve their original base-generated hierarchical order before movement.