Abstract
It is often said that for Japanese learners to acquire Japanese accent, they can only memorize each word individually. However, it is unlikely that Japanese native speakers do so. Thus, this study clarifies a tendency in the accent pattern and its rules using a word-list from the former JLPT. The results from the analysis are as follows: 1) Most two-mora + two-mora Chinese origin words belong to flat. 2) Most two-mora + one-mora and one-mora + two-mora Chinese origin words belong to flat or HLL. Based on the results, it is assumed that teaching these accentuation rules explicitly is effective.
