Formal Semantic and Pragmatic Approaches to Binding Theory

虽然代词、专有名称和明确的描述是一些最基本的语言表达，但直到最近30年，规范它们行为的规则才被正式详细地描述出来。例如，在“他喜欢导演的母亲”中，代词“他”和“导演”不能指同一个人(条件C)；但在“他的母亲喜欢导演”中没有这样的限制。虽然最近的语言学已经开发出复杂的规则来描述这些限制，但它们的存在从来没有得到令人满意的解释。本项目试图开发一种新的帐户，在该帐户中，这些限制是计算意义的程序的副产品。基本的观察是，经济学的一般原则禁止明确的描述包含的材料多于挑出其预期所指所需的内容--这解释了为什么说“约翰的金发碧眼的父亲”是奇怪的，因为“金发碧眼”在语义上是空闲的(因为一个人通常只有一个父亲)。在简单的情况下，这一经济原则的描述与其语言外的语境有关。然而，有人认为，意义是通过在句子中发现的代词、专有名称和描述的外延来字面上扩展语言外的上下文来计算的。当同样的经济原则应用到这个扩展的上下文中时，可以解释约束的经典例子(特别是条件C)，以及一些到目前为止还没有描述过的新数据。因此，该项目将增加现有理论的深度，并对相关事实提供更精细的描述，将通过比较来自英语、罗曼史和斯拉夫语的数据进行研究。

Although pronouns, proper names and definite descriptions are some of the most basic of linguistic expressions, it is only in the last thirty years that the rules that govern their behavior have been described in formal detail. For example, in "He likes the director's mother" the pronoun "he" and the description "the director" cannot refer to the same individual (Condition C); but no such constraint holds in "His mother likes the director". While sophisticated rules have been developed in recent linguistics to describe these constraints, their very existence has never been satisfactorily explained. The present project seeks to develop a novel account in which these constraints are the by-product of the very procedure by which meaning is computed. The basic observation is that a general principle of economy prohibits definite descriptions from containing more material than is necessary to single out their intended referent - which explains why it is odd to speak of "John's blond father", since "blond" is semantically idle (since one generally has but one father). In simple cases, this economy principle relates a description to its extra-linguistic context. It is suggested, however, that meaning is computed by literally extending the extra-linguistic context with the denotation of the pronouns, proper names and descriptions that are found within a sentence. When the same economy principle is applied to this extended context, the classical examples of the constraints (especially Condition C) can be explained, as well as some new data that had not hitherto been described. The project will thus add to the depth of existing theories and offer a more fine-grained description of the relevant facts, which will be studied by comparing data from English and from Romance and Slavic languages.