Two Types of Doubly-filled COMP in Middle English:

Cases of Wh-words and Wh-phrases

Abstract

It has been observed that in contrast to Present-day English, Middle English was not sensitive to Doubly-filled COMP Filter, permitting sequences such as a wh-element followed by a complementizer that. Based on the result of the corpus research conducted by the author, this paper first presents a critical review of the analysis proposed by Nawata(1999), who treats the wh-element of Doubly-filled COMP in a uniform way; the result of the corpus research suggests that we should introduce a distinction between two different types of the wh-element for an adequate treatment of Doubly-filled COMP in Middle English: wh-words and wh-phrases. Following the hypotheses developed by Iwasaki(2021), the present paper posits a different syntactic structure for the sequences of “a wh-word--that” and “a wh-phrase--that” and claims that while earlier stages of Middle English only had the former structure, its later stages had both structures. It is shown that the proposed analysis provides a natural account for the fact that different particular types of Doubly-filled COMP in Middle English disappeared at different timing. This proposal implies that the existence of Doubly-filled COMP in Middle English is closely related to the absence of that-trace effects in English at that time.