The interpretation and processing of quantifiers in structurally ambiguous sentences: insights from child language

众所周知，像*所有海盗没有回到船上*和*船长没有和美人鱼跳舞*这样的句子会产生所谓的范围歧义：它们产生不同的解释，取决于运算符all/a和not是按照它们出现的顺序(表面范围)还是相反的顺序(反向范围)解释的。因此，前一句可以解释为*并不是所有的海盗都回到了船上*或者*回到船上时没有海盗*，后一句可以解释为*船长没有和任何美人鱼跳舞*或者*有一条美人鱼船长没有和他跳舞*。值得注意的是，从所有出现在NOT之后和A出现在NOT之前的相似句子中获得反范围解释要困难得多，如果不是不可能的话。句子*船长没有吃掉所有的草莓*和*美人鱼没有和船长跳舞*似乎只允许表面上的解释。有没有一个普遍的原则来解释这种模式？正确预测反作用域可用性的理论概括建立在逻辑强度的概念上。只有在逻辑上不比相应的表面范围读数强的情况下，反范围解释才是可用的。也就是说，“并非所有海盗都回到船上”在逻辑上并不比“没有海盗回到船上”强，因为至少有一种情况下，前一句是真的，而后一句是假的，例如只有一个海盗回到船上，而其他海盗没有。在这些观察的基础上，我们在这个项目中解决了以下关键问题：(I)逻辑强度为什么以及如何决定局部解释的可用性？(Ii)是否有语义过滤器来阻止切换操作符范围的句法机制？(Iii)这种行为是由于语用推理(例如，标量暗示)的相互作用造成的吗？这些推理被认为是受逻辑力量支配的。(4)韵律语调最终是否表明说话者传达的是哪种解释？对语言习得的研究是获得针对上述问题的见解的最佳手段，因为它允许确定儿童习得相关语法方面的不同阶段可能控制范围歧义(语用强化、韵律等)。因此，范围转移与语用推理、韵律等加工因素的关联程度可以通过儿童在互动中如何应对这些现象来揭示。本研究涉及一系列六个实验，以成人和4-5岁儿童为实验对象，收集线下语义判断和在线眼动数据。研究结果将有助于我们更好地理解范围歧义及其与句法、语义、语用原则和加工因素之间的相互作用。

It is a well known fact that sentences like *All the pirates did not go back to the ship* and *The captain did not dance with a mermaid* generate so-called scope ambiguities: they yield different interpretations depending on whether the operators all/a and not are interpreted according to the order in which they appear (surface-scope) or the reversed order (inverse-scope). As a result, the former sentence can be interpreted as *not all the pirates went back to the ship* or *no pirates when back to the ship* and the latter sentence can be interpreted as *the captain did not dance with any mermaid* or *there is a mermaid with whom the captain did not dance*. Remarkably, to obtain the inverse-scope interpretation from analogous sentences in which all occurs after not and a occurs before not is much more difficult if not impossible. Sentences *The captain did not eat all the strawberries* and *A mermaid did not dance with the captain* only seem to allow the surface-scope interpretation. Is there a general principle that explains this pattern? The theoretical generalization that correctly predicts the availability of inverse-scope builds on the concept of logical strength. Inverse-scope interpretations are available only if they are not logically stronger than the correspondent surface-scope reading. I.e. *not all the pirates went back to the ship* is not logically stronger than *no pirates went back to the ship* because there is at least one situation in which the former sentence is true and the latter is false, such as a situation in which only one pirate went back to the ship but other pirates did not. Built on these observations we address the following key questions in this project: (i) Why and how does logical strength determine the availability of scopal interpretations? (ii) Is there a semantic filter that blocks the syntactic mechanism switching the scope of the operators? (iii) Is this behavior due to the interplay of pragmatic inferences (e.g. scalar implicatures) that are known to be governed by logical strength? (iv) Does prosodic intonation ultimately signals which interpretation is conveyed by the speaker? Research on language acquisition constitutes an optimal mean to obtain insights aimed to addressing the questions above, in that it allows to identify the different phases in which children acquire relevant grammatical aspects may control scope ambiguities (pragmatic strengthening, prosody, etc.). So, the extent to which scope shift is linked to pragmatic inferencing, prosody and other processing factors can be revealed by how children cope with these phenomena in interaction.The present proposal involves a series of six experiments with adults and 4- to 5-year-old children to collect offline semantic judgments and online eye tracking data. The results will lead to a novel and better understanding of scope ambiguity and its interaction with syntactic, semantic, and pragmatic principles and processing factors.