In-situ or Ex-situ: the Wh-Adjunct Hé ‘how, why’ in Old Chinese

Abstract

This thesis explores the syntactic behavior of the wh-adverbial hé ‘how, why’ in Old Chinese by focusing on the constructions “VP-hé” and “hé-VP”. Unlike the wh-nominal hé ‘what’, whose wh-fronting has been well discussed in the literature, the wh-adverbial hé receives little attention, especially in the construction “VP-hé”. The main innovation of this thesis is to correlate the syntax of hé with its reconstructed pronunciation *gˤaj, which motivates a morphological decomposition of hé within the framework of feature movement and the Copy Theory of Movement. A concise answer to the question posed in the title is that the how-wh-hé is an in-situ wh-word but the why-wh-hé can be analyzed either as an in-situ or an ex-situ wh-word. Chapter Ⅰ introduces the motivation for this study, which comes primarily from the fact that wh-adverbials cannot surface sentence-finally in Mandarin, whereas they can do so in Old Chinese. After a brief literature review, a methodological issue of doing historical generative syntax, namely, how to compensate for the lack of native intuition is discussed. Chapter Ⅱ provides descriptive syntax and semantics of the wh-word hé, revealing that both types of the wh-adverbial hé can surface at either the left or the right periphery. Then, the etymology of hé is clarified in a comparative Sino-Tibetan perspective. Some OC-internal evidence supporting the decomposition of hé is provided in this chapter. Chapter Ⅲ compares four solutions to the structure of wh-adverbial hé in “VP-hé”: object analysis (Aldridge 2010, Shi 2015), adjunct analysis, predicate analysis, and resultative complement analysis. This study chooses to analyze this kind of hé as a wh-adjunct for both empirical and theoretical reasons. Chapter Ⅳ subcategorizes the wh-adjunct hé into two types: why-wh-hé (i.e. high wh-adjunct) and how-wh-hé (i.e. low wh-adjunct) based on OC-internal evidence (i.e. Intervention Effect) and cross-linguistic evidence from Mandarin and Japanese. This chapter reaches a preliminary conclusion that both types of the wh-adjunct hé base-generate at the right periphery but the why-wh-hé moves in LF to [Spec, CP]. Chapter Ⅴ starts with a clarification of the necessity of a movement analysis for the why-wh-hé. To derive a feature movement analysis of hé, the remaining part of this chapter focuses on two kinds of question particles: Q[y/n] and Q[wh] in Old Chinese and other languages. Chapter Ⅵ temporarily turns its attention to modern Chinese dialects in order to compensate for the lack of certain syntactic diagnostics in OC. It is assumed that the preverbal Q-particle kəʔ5 in Jiānghuái Mandarin is etymologically related to the OC wh-word hé. Since kəʔ5 is demonstrated to undergo the rightward Neg0-to-C0 movement, hé can also reasonably move rightwards. This chapter provides a different solution to the analysis of the why-wh-hé, in contrast to the preceding two chapters. Chapter Ⅶ summarizes main findings and conclusions of this thesis and gives further clarifications on some contentious issues.