Exrolorations into the Top-Down Reconstruction of Logical Structure

在表明自下而上的重建现象方法存在经验和理论缺陷之后，研究者提出了一种自上而下的替代方法。第二章讨论了约束理论应用的时机问题，认为乔姆斯基认为约束条件最好作为一种适用于逻辑形式的解释规则来分析。第三章回顾了基于相位的重构理论，证明了必须做出一些辅助假设来维持这种方法。第四章从最新的研究中进一步证实了研究者长期以来试图从Chonisky的介入条件中推导出对句子主语提取的一般禁止，根据该条件，一个介入特征会阻碍对其他特征的检查。第五章研究了各种语言使役句的自顶向下推导。第六章指出了印地语-乌尔都语中向右置乱的理论中存在的几个经验和理论问题，其中向右置乱涉及到通过向左置乱提取成分的谓词的残余运动。第7章介绍了一种自上而下的重建方法，提出被移动的不是先前引入的成分的整个副本，而是其后续派生所必需的语法特征。“更具体地说，音韵、形式和语义特征是在一个动作依赖的最左边位置引入的，然后只创建形式和语义特征的副本。”这一理论的可行性得到了几个重建事实的证明。

After showing that bottom-up approaches to reconstruction phenomena have both empirical and theoretical flaws, the investigator has proposed a top-down alternative. Chapter 2 discusses the issue of timing of the application of the Binding Theory, claiming that Chomsky's view that the Binding Conditions are best analyzed as an interpretive rule applied at Logical Form. Reviewing the phase-based theory of reconstruction, Chapter 3 demonstrates the fact that some ancillary assumptions have to be made to maintain this approach. Chapter 4 adds some further confirmation from the latest researches to the investigator's perennial attempt to derive the general ban on extraction from the subject of a sentence from Chonisky's intervening condition, according to which an intervening feature blocks checking of other features. Chapter 5 is devoted to the top-down derivation of causative sentences in various languages. Chapter 6 points out several empirical and theoretical problems with the theory in which rightward scrambling in Hindi-Urdu is analyzed as involving remnant movement of a predicate from which a constituent has been extracted via leftward scrambling. Chapter 7 introduces a top-down approach to reconstruction, proposing that what is moved is not the whole copy of the previously introduced constituent, but its grammatical features neccessary to the subsequent derivation. 'To be more concrete, phonological, formal and semantic features are introduced in the leftmost position of a movement dependency, followed by creation of the copies of only formal and semantic features. The viability of this theory is attested by several reconstruction facts.