Abstract
The aim of this paper is to reconsider Harada's (1976) arguments concerning the morphological structure of a subject honorific form o- ... -ni naru in Japanese. One of the arguments against the nominal analysis where the root of the subject honorific form is nominal comes from the fact that the sentences with the subject honorific form cannot be deleted, relativized, or clefted. The current analysis of these phenomena in Japanese reveals that these data can be handled without any stipulations and indicates that Harada's original argument that the root of the subject honorific form is verbal should be correct. I also review the morphological structure that Harada (1976) proposes, and point out two problems with it.
