漢藏語之否定暨迦龍語個案研究：用歷時術語對其分布異象的闡釋

Sino-Tibetan Negation and the Case of Galo: Explaining a Distributional Oddity in Diachronic Terms

絕大部分漢藏語言（或跨喜馬拉雅語言）的謂語中心詞都帶前置否定標記，然而，在漢藏語言中，也有一小部分語言的謂語中心詞只帶後置否定標記，達尼語便是其中之一。這種否定標記具有某種特殊的性質——既與派生變化相似，又與屈折變化相似，在謂語主體中處於派生變化和屈折變化的「中間」地帶。通過對以情態助詞結尾的連動結構形成的單個謂詞的再分析，本文認為早期前置的情態否定標記由於在謂語主體中位置左移、重新排列而形成了謂語後綴，並且提出對上述兩種現象的普適性解釋。這與在有些藏緬語，譬如，庫基－－欽語和那加語中發現的另一種變化途徑有著相似性，在這些語言中，否定前綴與句尾情態助詞融合而形成了後綴。然而，筆者認為上述現象在本質上有些許的不同。本文提供了比 Post (2007) 文關於迦龍語（一種藏緬語族達尼支語言，位於東喜馬拉雅）否定結構更為詳盡、綜合的描寫，基於此，筆者得出上述論斷。因此，本文的另一目標在於豐富亞洲語言否定結構的類型學特徵。

While the vast majority of Sino-Tibetan (=Trans-Himalayan) languages have a pre-head predicate negator, Tani is one of a small handful of subgroups whose languages display an exclusively post-head negator. This negator, furthermore, is somewhat unusual in having both derivation-like and inflection-like properties, and in occupying an ‘intermediate’ position between derivations and inflections in the predicate stem. This article proposes a common explanation for both facts, by hypothesizing that reanalysis of an AUX-final serial verb construction as a single predicate word has resulted in realignment of an earlier pre-head auxiliary negator as a predicate suffix with leftward scope over the predicate stem. This is similar to another channel found in some Tibeto-Burman languages in which a prefixal negator fuses with a clause-final auxiliary to become a suffix (as in Kuki-Chin and ‘Naga’); however, I argue it to be ultimately somewhat different. These arguments are made on the basis of a more comprehensive description of negation in Galo (Tibeto-Burman > Tani, Eastern Himalaya) than was provided in Post (2007); as such, a second goal of the paper is to contribute to the typology of negation in Asian languages more generally.