A quantitative study of Tokyo Japanese

The present investigation is the first attempt to describe Japanese as is spoken in Tokyo within the variationist framework, focusing on the four linguistic variables: (ng), (ci), (/d u), and pitch accent placement of 3-mora native Japanese words. Chapter Two presents the methods of this dissertation. The progress which has been made in the last two decades in the area of data collection and analysis shows that informants should be selected in such a way that the sample may provide both gain in depth and representativeness of linguistic behavior of the speech community. I argue that the analyses of substantial amount of recordings of natural speech show the systematic aspect of linguistic variation. Chapter Three examines the variation (ng). It is the process of change; the informants of the younger age levels favor the newer variant (g) while those of the older age levels disfavor it. I show that the variable exhibits a case of change from above. Besides the social factors, some linguistic internal parameters are correlated with the variation; by means of multivariate analysis, it is shown that the variation is constrained by general phonetic factors and grammatical parameters. All these effects are subsumed under the notion of salience . Chapter Four is a discussion of the variable (ci). The variant (i) is strongly associated with older generation, male, working class speakers from shitamachi , a sub-area of Tokyo. Chapter Five presents the results of the analyses of the variable (/d u). It is shown that there is only one factor that constrains this variable: age. Those over 70 favor the (/d i) variant. Chapter Six discusses the change in progress of pitch accent placement. I argue that the change is spreading out to the younger informants. The findings of the present investigation are an addition to the body of research which has been conducted within the variationist framework; they confirm the existence of inherent variation and orderly heterogeneity