Problems of projection: Extensions
Problems of projection: Extensions
复制标题
投影问题：扩展
DOI：
10.1075/la.223.01cho
复制
发表时间：
2015
影响因子：
14.9
通讯作者：
Noam Chomsky
中科院分区：
文献类型：
--
作者：
Noam Chomsky
我在这里假设的是极简主义研究计划，我认为该计划的动机很好，因为可学习性、解释性成功以及关于人类语言能力起源的信息非常有限。特别是，我假设乔姆斯基（2013）的结论，包括放弃X-bar理论及其后代的内心性规定，以及将投影（标签）与表达式的构造原则分离。标签算法保持最小搜索，将标签分配给由迭代合并（外部 EM 或内部 IM）构造的表达式 {X,Y}；标签不会产生新类别如果 X,Y 之一是头部，则标签很简单：最小搜索会产生头部作为标签。如果两者都不是头部，则仅当搜索 X 和 Y 产生一致的头部时才可能进行标记，这意味着如果其中一个被提出，那么它现在就处于 Luigi Rizzi 意义上的标准位置。进一步假设在 CI 接口处，对于外部化规则，必须识别（标记）语法对象。由此可见，IM 是连续循环的，导致标准位置，并且被迫确保标记。
I assume here the minimalist research program, which I think is well-motivated on grounds of learnability, explanatory success, and the very limited information on origin of the human language faculty. In particular, I assume the conclusions of Chomsky (2013), including the abandonment of the endocentricity stipulation of X-bar theory and its descendants, and the separation of projection (labeling) from the principles of construction of expressions. A labeling algorithm, keeping to minimal search, assigns labels to expressions {X,Y} constructed by iterated Merge (external EM or internal IM); labeling yields no new category If one of X,Y is a head, labeling is trivial: minimal search yields the head as a label. If neither is a head, labeling is possible only if search of X and Y yields agreeing heads, meaning that if one or the other was raised it is now in its criterial position in Luigi Rizzi’s sense. Assume further that at the CI interface, and for the rules of externalization, syntactic objects must be identified (labeled). It follows that IM is successive-cyclic leading to a criterial position, and is forced to ensure labeling.