It is known that writing style depends on individuals and maintains consistency in each individual. If we can prove that some changes in writing style are caused by a significant transition in writer's state of mind including mental illness, psychological change can be detected from a change of writing style. Previous studies showed that, for a specific novelist, a transition in the writer's mind leads to changes in its writing style. However, little research has been conducted on population studies, because it is difficult to sample many people who have the same type of mind transition. To solve this problem, assuming that women make internal transitions when they become mothers, we constructed a large corpus of their blog entries before and after their delivery. If we select features that depend on the contents of the text such as frequency of words, we will find the changes of writing style based on shifts of their lifestyles rather than their internal transitions. Therefore, we used characteristics that do not depend on the topics of the text such as frequency of non-content words and bigram of parts-of-speech. This study shows that their internal transitions associated with childbirth causes their writing style change.

View full abstract

Through the investigation of Asahi Shimbun, Katakana sofuto was introduced as a soft hat which expresses a soft “trilby hat” against a “mountain cap” in the study. Gradually, sofuto began to appear in Japanese as a modifier and its meaning expanded. Soon, the adjective verb sofutona appeared. Since then, loanwords and their new modification methods have been absorbed by Japanese and occupied part of the meaning of yawarakana, but they have little influence on junan'na.

View full abstract

Using data on illiteracy rate obtained from the survey of the reading and writing ability of the Japanese in 1948, we analyzed the relationship between illiteracy rate and year of birth. Referring to the S-curve theory of language change and the concept of diachronic change analysis using apparent time, logistic regression analysis was conducted with illiteracy rate as the objective variable and year of birth and place of residence (urban or rural) as explanatory variables. The analysis revealed an inverse S-curve relationship between illiteracy rate and year of birth, with the illiteracy rate decreasing with later year of birth (younger age at the time of the survey, with a lower limit of 15 years). The main reason for this is thought to be the increase in the rate of elementary school enrollment over time, which was linked to the introduction of free compulsory education. The analysis also showed that the inverse S-shaped curve of the illiteracy rate in the rural area overlaps with the inverse S-shaped curve in the urban area when the time of six years is shifted parallel to the left along the horizontal axis of the graph (year of birth).

View full abstract

The following four points are discussed in this paper. First, the concept of cross-tabulation table (or cross-tab) is described; a conventional cross-tab example, along with its slightly modified version is presented, to examine the essence of cross-tab analysis. Second, an extended cross-tab, which is a table listing simple tabulation of multiple questions of the same choice group, is presented. The extended cross-tab can be handled in the same manner as the conventional cross-tab. In addition, it has several advantages for data analysis, which will also be explained. Third, the calculation process of reciprocal averaging suitable for using extended cross-tab is explained; this is done by presenting a step-by-step calculation process using a specific cross-tab example, so that the process can be performed using an electronic calculator. Fourth, the characteristics of reciprocal averaging are presented, and the types of its applicable data is explained. Thus, the present paper, as a whole, describes the advantages of reciprocal averaging in the analysis of data given in cross-tab format.

View full abstract