A Study of the Semantics of Mandarin Cleft Constructions and Their Exhaustivity


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
Award date14 Sep 2017


Shi…(de) sentences, e.g. (1) below, were generally considered to be the Chinese counterparts of English it-clefts (see Teng 1978; Huang 1988; Lee 2005; Cheng 2008; Paul & Whiteman 2008).

(1)Shi Xiaogao chidao le.be Xiaogao late SFP‘It is Xiaogao who was late.’Exhaustivity: ‘Besides Xiaogao, no one else was also late.’

This dissertation investigates the semantics of Mandarin shi…(de) clefts with much attention drawn to the exhaustivity feature shown in (1). It aims to address three important questions: (i) what is the semantics of shi…(de) clefts? (ii) is exhaustivity of shi…(de) clefts a semantic or a pragmatic inference? and (iii) what is the analysis towards exhaustivity of shi…(de) clefts?

The investigation begins by re-examining several relevant structures of shi…(de) clefts. Based on Lee (2005), Cheng (2008), Paul and Whitman (2008), among others, I sub-categorized the eight relevant structures into two types, namely, the interpretative parallels and the non-interpretative parallels, according to whether they could yield a cleft-like interpretation. Contrary to previous analyses, it is argued that the cleft semantics of shi…(de) sentences is derived from the semantics of shi but not from the sentence final de. Therefore, bare shi structure and shi…de construction are semantically equivalent.

It then centers on the exhaustivity of shi…(de) sentences. Previously, scholars agree that English restrictive particle only asserts exhaustivity, whilst wh-elicited plain focus sentence conversationally implicates exhaustivity. However, cleft receives some controversies: it is considered (i) as an assertion, on a par with only (see É Kiss 1998; Lee 2005); (ii) as part of presupposition (see Velleman et al. 2012); or (iii) a conversational implicature (see Horn 1981). In this context, I present three experiments examining how shi…(de) clefts encode exhaustivity. It is found that exhaustivity of shi…(de) clefts is interpreted differently from restrictive particle zhi-you ‘only-have’ or a wh-elicited plain focus sentence but is interpreted on a par with that of definite specificational sentences. These results suggest that exhaustivity of shi…(de) clefts is not asserted. However, whether it is a semantically encoded presupposition or a pragmatically derived implicature remains unclear. To solve this problem, a large-scale corpus and web-based questionnaire study was conducted. It is found that exhaustivity of shi…(de) sentences is less robust and more heterogeneous than expected by previous analyses. The fact that it reflects contexts-dependency further suggests that it is a pragmatically derived implicature.

Based on the above empirical results, I proposed a QUD (Question under Discussion) based analysis. Following Velleman et al. (2011), I argue that shi…(de) clefts are sensitive to QUD, and it is QUD that determines whether and how exhaustivity arises. A shi…(de) cleft presupposes that there is a minimal answer to the Current Question. While asserting its prejacent, it further faces a discourse constraint: as an inquiry terminator, a shi…(de) cleft is required to provide a maximal answer to the CQ; and it is from such a maximality of the answerhood that exhaustivity is further derived.

    Research areas

  • shi...(de), exhaustivity, clefts, experimental semantics