The theme of this paper is to express the Japanese language into written words. However, this paper discusses this theme from two perspectives: phonographic transcription and non-phonographic transcription. This paper covers four different eras: the Jyodai Era(ancient times), the Tomatsukensho Era(from the end of the 11th century to the end of the 12th century), the Chusei Era(medieval times), and the Meiji Era. In the Jyodai Era, the Japanese words used were only Kanji(Chinese characters). Kanji was originally used to write Chinese. Thus, adequate consideration is necessary in the use of Kanji in writing Japanese. What was fulfilled in the Jyodai Era was "phonographic transcription" that resulted from distinguishing voiced consonant and unvoiced sound of Kanji, which was originally not used for the Japanese language. The Tomatsukensyo Era was the period that followed the era that brought significant change in phonography in which there was a shift from the period using "phonographic transcription" to the period using "non-phonographic transcription. Regarding Kaigo(Chinese phonetics) of the Chusei Era, the author of this paper argues that a different perspective may be possible by introducing phonographic transcription and non-phonographic transcription In the Meiji Era, phonographic transcription was commonly used. This paper indicates that the observation and analysis of each word would offer an opportunity to find conditions that have not been fully studied until now. Although this paper focuses on viewing the phonology of words, the author designates the validity of adopting the two perspectives of phonographic transcription and non-phonographic transcription in lexical analysis.