摘要
唯递归假说认为人类自然语言所共有的狭义句法只包括递归计算,它是人脑的先天特有属性。研究者纷纷从不同视角支持或反对该假说,但由于他们对"递归"的界定不明确、不统一,因而结论之间缺乏可比性和连贯性。本文较全面地概括了生成语法文献中对递归界定3个方面的争议,尤其是明确了一些反复使用却又纠缠不清的术语,如"递归""合并""嵌套""标注""迭代/重复""并列"等之间的联系和区别,并结合文献中一语习得和神经语言学研究结论提出更合理的界定思路。
The Recursion-Only hypothesis holds that the syntax in the narrow sense only includes recursion,an innate property of human brain,unique to human and natural languages.The literature has presented theoretical and experimental methods from various perspectives to support or oppose this hypothesis,but their definitions of recursion are not common or clear,resulting in the incomparability and inconsistency of the conclusions.The purpose of this paper is to summarize the different arguments on the definition of recursion in the generative literature and to identify and infer which argument may be more appropriate,especially in the light of the experimental findings of previous first language acquisition studies and neurolinguistics studies.Moreover,the relationship between terms like recursion,merge,embedding,label,iteration/repetition and coordination,which are frequently used but often misunderstood in the literature,is clarified in the paper.
作者
杨彩梅
董昕
YANG Cai-mei;DONG Xin
出处
《解放军外国语学院学报》
CSSCI
北大核心
2022年第1期1-10,160,共11页
Journal of PLA University of Foreign Languages
