The Focal Structure in Mandarin VP-Ellipsis: A Cross-Linguistic Perspective

Abstract

本文論證Soh（2007）之ΣP分析，只能解釋漢語動詞刪略句中極性運作的部分事實。我們以「也」和「卻」的新事證為基礎，提出動詞刪略句中有兩個焦點投射的分析；一為以「也」或「卻」為主要語的對比焦點投射FocP，其位置比TP還高；另一為以「正的極性焦點」及「負的極性焦點」為主要語的極性投射PolP，其位置比TP低；FocP及PolP會在極性上互相呼應，形成「也」類前後極性一致的句式，及「卻」類前後極性相斥的句式；此互動可透過「呼應運作」（Chomsky 2000，2001）的理論架構，同時採用Watanabe（2004）的「否定一致」特徵複製來達成。此外我們檢視各類語序的語言，包括英語（SVO）、日語（SOV）、泰雅語（VOS）、布農語（VSO），來驗證此焦點分析的合理性。

This paper argues that Soh’s (2007) ΣP analysis only partially explains polarity operation in Mandarin VP-ellipsis. With new examples of the use of the particle que ‘however’, a polarity contrast of ye ‘also’, we propose that there are two focus projections in VP-ellipsis. One is the contrastive FocP headed by ye or que higher than TP and the other is the polarity PolP headed by an affirmative polarity focus shi ‘be’ or a covert negative polarity focus lower than TP. Foc interacts with Pol by a way of polarity concord, which is responsible for the polarity symmetry or asymmetry across two conjuncts. We suggest that the polarity concord is achieved via the Agree operation (Chomsky 2000, 2001) in line with Watanabe’s (2004) feature copying analysis of the negative concord. A cross-linguistic investigation of languages of various word orders, including English (SVO), Japanese (SOV), Atayal (VOS), and Bunun (VSO) lends support to this focus account.