Lexical Inference vs Scalar Implicature

格莱斯的隐含概念以区分语用派生的内容和词汇定义的内容为前提。这一区别是所有关于含义的描述的核心，但准确的描述仍然很困难，而且往往存在争议。手边的一个案例是通过句法结构中的无声穷举运算符来分析标量含义。在这样的分析下，词汇内容和含义都有助于真值条件意义，并且两个意义成分之间的差异应该较小，例如在它们的加工、可消除性和可嵌入性方面。同时，穷举分析的力量越大，越有必要更好地理解词汇和派生内容的区别，以及对穷举机制的限制，以避免过度生成。这个项目的目标是对词汇内容和含义之间的这种区分至关重要的一系列案例。我们使用理论和实验诊断学研究了四种不同类型的语言。此外，我们还比较了与数量和方式相关的效果所需的备选方案集。这些经验发现将使我们处于有利地位，能够对标量可选集合和穷举进行新的描述。我们的工作假设是，派生特定意义成分的选择集合的构成与相关意义成分的解释和推导中可识别的独特的心理语言学特征相关联。具体地说，该工作假设假设了三类意义成分：A)独立于选择集合的词汇内容，B)仅从次成分和数词选择中产生的强含义，以及C)基于词汇尺度的弱含义，例如‘一些’-‘全部’尺度。这一主要假设将通过回答八个开放的经验问题来检验：1)所有自由选择析取的情况都能得到语用解释吗？2)自由选择推理能被其他类型的暗示启动吗？如果是的话，哪些含义？3)一方面是Warlpiri和日语中的可变力连接词，另一方面是ST‘at’imcets中的可变力情态动词，表现出其他含义的心理语言学特征？5)不确定项的自由选择效应何时被儿童习得？6)连接词的变化力何时被儿童习得？7)当嵌入非单调逻辑运算符时，哪些现象可以表现出非局部含义和含义取消？8)涉及相同选择的方式效应和标量含义是否表现出相同的加工特征？在追求这八个经验目标的过程中，我们将在词汇意义与含义意义的争论中建立重要的新证据，使我们能够在不同和广泛的观察基础上修改我们的工作假设。

Grice's notion of implicature presupposes a distinction between pragmatically derived and lexically defined content. This distinction is central to all accounts of implicature, but the precise delineation remains difficult and often controversial. A case at hand are analyses of scalar implicature via silent exhaustification operators in the syntactic structure. Under such an analysis, lexical content and implicature both contribute to truth-conditional meaning, and fewer differences between the two meaning components are expected to be observed, for example in their processing, cancelability, and embeddability. At the same time, the greater power of an exhaustification analysis amplifies the need to better understand the distinction of lexical and derived content and the restrictions on the exhaustification mechanism to avoid over-generation. This project targets a broad array of cases central to this distinction between lexical content and implicature. We consider four typologically diverse languages using both theoretical and experimental diagnostics. In addition we compare the set of alternatives required for quantity- vs. manner-related effects. The empirical findings will put us in a strong position to formulate a novel account of scalar alternative sets and exhaustification. Our working hypothesis is that the make-up of the alternative set involved in deriving a given meaning component correlates with a distinct psycholinguistic profile identifiable in the interpretation and derivation of the relevant meaning component.Specifically, the working hypothesis postulates three categories of meaning components: A) lexical content, which is independent of the alternative set, B) strong implicatures derived from only subconstituent and numeral alternatives, and C) weak implicatures based on lexical scales such as the 'some'-'all'-scale. This main hypothesis will be tested by addressing eight open empirical questions: 1) Can all cases of Free Choice disjunction receive a pragmatic explanation?, 2) Can Free Choice inferences be primed by other kinds of implicatures and if so, which ones? 3) Do variable force connectives in Warlpiri and Japanese on the one hand, and 4) variable force modals in St'at'imcets on the other hand, exhibit the psycholinguistic profile of other implicatures? 5) When are free choice effects with indefinites and 6) variable force of connectives acquired by children? 7) Which phenomena can exhibit non-local implicatures and implicature cancellation when embedded under non-monotone logical operators? 8) Do manner effects and scalar implicatures which involve the same alternatives exhibit the same processsing profile? In pursuing these eight empirical goals, we will establish important new evidence in the controversy about lexical vs. implicature-based meaning, enabling us to revise our working hypothesis on the basis of diverse and broad observations.