汉语否定词 "没(有)" 和 "不" 对焦点敏感度的异性

“不” 和 “没( 有) ” 是汉语中两个主要的否定词，以往文献把它们视为对焦点敏感的算子，并不区分它们对焦点的敏感度。Herburger( 2000) 提出通过事件量化( event quantification) 分析英语否定词，重新检视非焦点成分对句子所述事件的量化影响，认为事件量化词的量化是受限于其辖域内的非焦点成分，该成分给出的是一个背景事件描述。Beaver 和Clark( 2003) 也认为即使对焦点敏感的算子的释义可能与焦点直接或间接有关，也不能采取一刀切的方法处理。根据Herburger 及Beaver 和Clark 的分析，我们提出 “不” 和 “没( 有) ” 对焦点的敏感程度并不一样: “不” 是一个对焦点敏感的否定算子，与焦点直接关联得出 “不” 字句的基本释义; 而由于句法上的限制，“没” 必须依附于 “有”，“没” 的释义只能是对焦点间接敏感。此外，通过事件量化分析，“没” 可以被看作是否定情状存在的量化词，它的释义显示，量化结构中的背景或限定部分可以通过句子去掉焦点部分得出，而焦点在 “没” 字句中的角色是给出非焦点或背景部分，以对与 “没” 有关的背景事件进行限制。
Bu (不) and mei(-you) (没(有)), both translated into English as “not,” are the two main negators in Chinese. Previous analyses generally assume bu to be focus-sensitive, with mei(-you) taken as equivalent in terms of focus-sensitivity. This paper re-examines the interpretation of focus for the two Chinese negators, bu and mei(-you), with particular focus on the latter, and argues that although falling under the same category, bu and mei(-you) demonstrate a non-uniformity towards focus. Herburger (2000) proposed an event quantification analysis to account for the English negator “not” and negative quantifiers, arguing that the non-focused part of a sentence is to provide a background event description for quantification. This analysis of mei(-you) works similarly to Herburger’s event quantification analysis. It is widely acknowledged that bu is a focus-sensitive operator, and its interpretation is dependent on the placement of focus (if there is one). When there is focus within the focus negation domain of bu, bu will associate with the focus, triggering a focus-background partition. I argue that mei(-you) should be made distinct from bu in its degree of dependence on focus: unlike bu, mei does not directly associate with focus, with focus-background partition rendering misinterpretation of the sentence. Since the mapping of mei(-you) is not focal mapping, an immediate difference between bu and mei(-you) is the absence of an exclusive reading on the focus in mei-negation, which is nevertheless found in bu-negation. The underlying difference between the two in terms of focus dependence is due to the failure of semantic focus to override the syntactic constraint of mei, which requires its syntactic attachment to you, making mei(-you) inherently a negative existential quantifier of situations. As a quantifier, interpretation of mei is only indirectly affected by focus, with mei exhibiting the following quantificational structure: material within the TP scope of mei, excluding the focus, will be structurally mapped to the background part, to set up its restrictive domain and everything within its scope to the nuclear scope. The event description in the background part thus indicates what the sentence is about, with the nuclear scope asserting that for some (relevant) events denoted by the background part, there exists no such an event denoted in the nuclear scope. The role of focus in mei sentences is therefore to provide the non-focused part to contribute the backgrounded event description. This should be made distinct from bu, of which the mapping mechanism setting up the background part of bu is focal mapping. Moreover, such a quantificational structure of mei reveals another role of focus in contributing to the restrictive part: unlike focal mapping which is in the non-focused or background part, the focused item is replaced by a variable f, the restrictive part of mei is made up by deleting the focus, with the residual mapped to it to contribute to the backgrounded event description. Since what constitutes the restrictor part is the event description and focus is removed altogether when structuring the restrictive domain of mei, the only semantic presupposition triggered in mei-negation is the existence of at least one relevant event, which is non-defeasible, with no existential import of at least one alternative to the focus. Finally, the variation in terms of their focus-sensitivity has determined the two negators to be inherently different in their prototypical readings: bound reading for bu and structured-wide reading for mei (cf. Herburger, 2000), which is in contrast with English “not” where ambiguity between the two readings arises.