Intervention Effects and the Covert Component of Grammar

干涉效應與語法隱性部門

楊中玉(Barry Chung-Yu Yang)

指導教授 ： 蔡維天

國立清華大學/人文社會學院/語言學研究所/博士(2009年)

https://doi.org/10.6843/NTHU.2009.00067

摘要

疑問詞的干涉效應(wh-intervention effect)最主要的問題在於跨語言間和語言內部的分佈上都有著不一致的表現，而說話者的語感判斷亦有不一致的情形發生(Kuno & S.-Y. Kim 2004, Miyagawa & Endo 2004)。雖然近年來的研究傾向於專注在焦點(focus)干涉效應這一層面以避免上述問題(Kim 2002, 2005, Beck 2006)，然而仍有研究指出焦點干涉效應還是有其問題存在(Miyagawa & Endo 2004, Tomioka 2007)。本研究有三個目的：一、將分佈紛亂的干涉效應化約為二個獨立運作的句法效應，即「近距效應(Minimality Effect)」與「競爭效應(Competition Effect)」；二、印証大句子分裂假說(Split-CP Hypothesis)的有效性；三、擴大干涉效應至否極詞項(negative polarity item)與自由選項詞(free choice item)的允准(licensing)環境。
首先本文指出漢語在干涉效應現象上清楚的分佈為這方面的研究提供很好的切入點。我們認為應該將干涉效應化約成「近距效應」與「競爭效應」二類。近距效應係指從Y到X的特徵移位(feature movement)受到干涉成分Z所帶的同類特徵[α]阻擋而產生(cf. Starke 2001, Rizzi 2002)。我們採用Tsai (1994)的參數原則(parametric approach)處理在位疑問結構，使得近距效應能應用並釐清干涉效應的諸多分佈。「競爭效應」則是從一個傳統術語「一位一補」(one-slot-per-Comp)衍生而來，此一簡單觀念可以而清楚地解釋焦點干涉效應的嵌入句問題(Tomioka 2007)和語用效應的言談連結(D-linking)問題(Miyagawa & Endo 2004)以及其他幾項相關議題。
其次，我們發現上述二種效應也可以在大句子層次上體現。該發現成功地將看以無關的反問句、反駁句、評註副詞、和疑問論元/附加語交互作用等現象以統一的手法在句子的左緣結構上加以分析，同時也直接印証了大句子分裂假說的有效性。
接著，本文將干涉效應擴大到否極詞項與自由選項詞的認可環境，同時藉著干涉效應指出這些詞項和無定疑問詞(wh-indeterminate)的異同之處，並進一步點出干涉效應的發生並非因為允准詞(licensor)和否極詞項與自由選項詞間的C-統制/約束關係被阻斷，而是因為否極詞項與自由選項詞的特徵移位受到阻擋，此一現象再度印証本研究方向的正確性。
總結來說，本文指出干涉效應紛亂的現象可以化約為二項單獨運作的句法效應，該效應能進一步驗証大句子左緣結構的存在，同時也能擴大到否極詞項與自由選項詞的允准關係。本研究成果期望一方面進一步瞭解干涉效應的內涵，一方面也能具體有效的掌握在位疑問詞的內部結構。

關鍵字

干涉效應 ； 焦點效應 ； 大句子分裂假說

並列摘要

The exploration on wh-intervention effects generally suffers from distributional variations both across languages and within languages. Meanwhile, speaker judgment is also reported to be inconsistent (Kuno & S.-Y. Kim 2004, Miyagawa & Endo 2004, among others). Though recent researches tend to limit their discussions to the focus intervention effect (Kim 2002, 2005, Beck 2006) since it enjoys a more stable distribution, new language data show that such an approach still runs into some problems (Miyagawa & Endo 2004, Tomioka 2007). This thesis has three aims: i) to subsume the variations of intervention effects into two types of syntactic effects: Minimality Effect and Competition Effect; ii) to testify the Split-CP hypothesis (Rizzi 1997, 1999, 2002); iii) to extend the domain of intervention effects to other syntactic dependencies.
To begin with, I show that Mandarin Chinese provides a clear picture for us to sort out the puzzling variations of intervention effects. I propose that the above variations can be best captured if we categorize them into two types of syntactic effects: Minimality Effect and Competition Effect, each of which is triggered by independent factors. The Minimality Effect occurs when the feature movement of [α] from Y to X is blocked by an intervening Z which bears the same feature as X and Y do (cf. Starke 2001, Rizzi 2002). Assuming Tsai’s (1994) parametric approach toward wh-in-situ construal, we can easily extend the Minimality Effect to account for various distributional differences. The Competition Effect is derived from a traditional notion “one-slot-per-Comp”. This simple idea elegantly explains both the embedding problem of the focus effect (Tomioka 2007), the d-linking problem of the pragmatic effect (Miyagawa & Endo 2004), as well as several other puzzles.
Next, I show that the Minimality Effect and the Competition Effect can also be located in the CP domain, which brings together several seemingly unrelated phenomena involving the rhetorical wh, the refutatory wh, the evaluative adverb, and the wh-argument/adjunct interaction, with a uniform analysis. The analysis developed in this study thus lends further support to the Split-CP hypothesis (Rizzi 1997, 1999, 2002).
Then, I proceed to probe the intervention effects in wider contexts and show that the licensing of negative polarity items (NPIs) and free choice items (FCIs) are also subject to intervention effects. Specifically, evidence from the contrast between the NPI/FCI renhe ‘any’ and the wh-indeterminate shows that the intervention effects occur not from the interruption of c-commanding/binding relation between the licensor and the NPI/FCI or wh-indeterminate, but from the interruption of feature movement. This further testifies that the Minimality Effect is on the right track.
To conclude, I propose that the puzzling distributions of intervention effects can be reduced to two independent effects which in turn can be extended to the left periphery of CP and further to the licensing of the NPI/FCI. Hopefully, this study can bring a step closer to the understanding of intervention effects on the one hand and to the characterization of wh-in-situ construal on the other.

並列關鍵字

Intervention Effect ； Focus Effect ； Split-CP Hypothesis

參考文獻

Beck, Sigrid. 1995. Negative islands and reconstruction. In On Extraction and Extraposition in German, Uli Lutz and J

Aoun, J., N. Hornstein, and D. Sportiche. 1981. Some Aspects of Wide Scope Quantification. Journal of Linguistic Research 1, Vol 1, No. 3, 69-93.

Aoun, Joseph, and Audrey Li. 1993a. Syntax of scope. Cambridge, MA: MIT Press.

Aoun, Joseph, and Audrey Li.. 1993b. Wh-elements in situ: Syntax or LF. Linguistic Inquiry 24: 199-238.

Beck, S. and S.-S. Kim. 1997. On wh- and operator scope in Korean. Journal of East Asian Linguistics 6:4, 339-384.