Abstract
In this article, I propose a historical model of Japanese vocabulary structure using 13 classics from the eighth century to the twelfth century. I developed a statistical method for analyzing vocabulary structure, and analyzed the vocabulary structures of three eras -the Nara era(710-784), the Heian era(794-1185) and the Kamakura era(1185-1333). Because the table to analyze vocabulary structure consists of frequencies and lexeme use ranges, every vocabulary structure of the three eras is a two-dimensional structure at first. I call this the "surface structure". However, when we focus on the distribution of stable data, a one-dimensional structure becomes apparent in the numerical distribution form of the lexemes. I call this the "deep structure". I propose a historical model of Japanese vocabulary structure by arranging the vocabulary structures in chronological order, and I interpret this historical model of Japanese vocabulary structure as a multilayered structure.
