Sino - Japanese and a Way to Shape up the Stem

It is well-known that size of the stems in Sino-Japanese (SJ), one of the lexical strata in Japanese, is severely restricted. According to the Richness of the Base hypothesis (Prince & Smolensky 1993), such size restriction phenomena must be treated not by the constraints/rules held in the input, but by interaction of a set of constraints applicable to the output. In other words, while the output stem size in SJ is quite restricted, the input stem size, in principle, can be very large. In this paper, we are concerned with the proper characterization of the size restriction phenomena in SJ. Sino-Japanese morphemes, which form a large part of the modern Japanese lexicon, have their origin in Classical Chinese. They were borrowed into Japanese around the 7 and 8 centuries when Japanese started to use Chinese characters (Oono 1994, Kato et al 1989). It has been noted in the literature that morphemes in the SJ stratum have a strict templatic restriction as summarized in (1).