Subject Extraction and Clause Size

Abstract

It has been noted in the literature that subject extraction shows peculiarities. This paper discusses the argument that when subject extraction takes place, the embedded clause out of which the subject is moved will always be a smaller-than-full clause, and addresses why such a clause matters for subject extraction. Assuming the Strong Minimalist Thesis (SMT) to be the basic hypothesis in quest of an explanatory theory of language, I claim that subject extraction can be explained by the structure building operation Merge and the labeling algorithm called Label, both of which are elements of principled explanation that follow from SMT, answering why clause size counts for extraction of the subject. I argue that the proposed analysis provides the simplest, hence principled, account of language. I also show that it brings three consequences, which endorse the role of Merge in language.