A Study of the Formation of the Western Dialects of Japanese Using the Computer Program for Analysis of the Linguistic Data in the Grammar Atlas of Japan (GAJ)

《日本语法地图集》（GAJ）由国立日语语言与语言学研究所编纂出版，收录了大量日语方言的语言资料，重点关注语法项目。GAJ中的每一张地图都旨在展示在每个地区观察到的语法形式，而不是呈现编辑对语法形式分布的解释，GAJ可以作为日语方言语法形式的数据源。事实上，由于GAJ中的数据太多，变体的区分过于详细，很难将其作为数据源使用。为了解决这个问题，我们对GAJ中的数据进行了一些修改，并编写了一个程序，使我们能够自动绘制分布图。我们对这个程序制作的地图进行了详细的比较，发现了一些对考虑日本西部方言形成很重要的事实。

The Grammar Atlas of Japan (GAJ), compiled and published by the National Institute for Japanese Language and Linguistics, contains a large number of linguistic data of Japanese dialects, focusing on the grammatical items. Every maps in GAJ are intended to show the forms observed in each area, rather than to present the editors' interpretations on the distribution of the grammatical forms, and GAJ can be used as a data source of the grammatical forms of Japanese dialects. In fact, it is difficult to use it as a data source because there are too many data in GAJ and the distinctions of variants are too detailed. In order to solve this problem, we added some modifications to the data in GAJ and made a program which enables us to make distribution maps automatically. We made detailed comparisons among maps made by this program and found some facts which are important to consider the formation of the Western dialects of Japanese.