Abstract
The polysemy of ri-form used in Old Japanese （OJ） has been analyzed from various points of view（ Nomura 1994; Nishina 2009; Oka 2013）. This paper argues ri-form only has the ability to describe strongly homogeneous phases in the phase structures of verbs, and this ability creates the polysemy. Furthermore, this paper reveals that two subcategories of homogeneity, strong and weak, exist. This distinction is useful to analyze the difference between ri-form and shiteiru-form used in Modern Japanese.
