Standard language distribution in Linguistic Atlas of Japan (LAJ) and Limestone cave model: geographical and historical interpretation of railway distances

This paper presents a novel technique of showing the railway distance on a real map. Standard Japanese words were grouped into several categories to explain their historical distribution patterns. The influence of the old capital, Kyoto, was significant in the historical development of Japanese, against popular belief that Tokyo speech is the base of the standard language. It has already been shown that the railway distance is effective for explanation of regional differences in dialect. By taking the railway distances from Tokyo and Kyoto as a clue, and by plotting usage values laterally left-to-right on the graph allowed straight-forward visual interpretation of the East-West distribution throughout the Japanese Archipelago. In this paper, we report on the further results of application of two attempts. (1) We used past data on railway distance, comparing the railway distance data in 1980 and in 1910. The railway distance in 1910 showed closer correlation with the standard form usage rate than that of 1980. The correlation coefficient of the railway distance from Kyoto was low. However, it suggested a lot about the correspondence with the actual geographical distribution pattern. (2) The display method was improved. We superimposed a line graph reflecting the standard form usage rate and railway distance on a map of Japan. By visualizing this relationship, it became possible to intuitively understand the correspondence of standardization with traffic conditions and so on. We were also able to interpret the results from (1) above, and we discussed the theoretical possibilities related to railway distance. This analysis made it possible to see two peaks, one at Kyoto and one at Tokyo, for the propagation rate of standard Japanese on the basis of the railway distance. We also explored the possibility of applying the “limestone cave model” to other phenomena. We examine the possibility that geographical location can be used in interpreting the historical diffusion process.