Paper short abstract:

This paper is to investigates tonal variation of Kagoshima Japanese in terms of generational difference of tone production and related social factors, and to discuss the possibility that this variation represents an ongoing change of a systematic shift of accentuation.

Paper long abstract:

Hayata (1999) argues that, while Tokyo Japanese (henceforth TJ) is classified as a ‘Word accent’ (or pitch-accent) dialect, some regional dialects of Kyushu Island employ tone patterns, or ‘Word tone’, to decide tonal contour of words. KJ belongs to this category and has two types of tone pattern. Type A has a pitch prominence (or a high tone) on the penultimate syllable of word (e.g. kaede, ‘maple’, LHL), and Type B does on the final syllable (e.g. momiji, ‘autumn leaves’, LLH). As to pitch movement, these tone types correspond to the pitch-fall (or accented) pattern and the flat (or unaccented) pattern of TJ respectively. However, there is a sharp discrepancy of auditory image of many lexical items between the traditional KJ tone contour and that of TJ, and to resolve this discrepancy, young KJ speakers tend to select the same accentuation types as TJ which can be called as ‘innovative’ KJ. This tonal change seems to have been brought about by dialect contact between KJ and TJ, possibly through mass media broadcast.

By analysing data collected from 30 speakers (10 elderly and 20 young), the following two findings were obtained.

First, the elderly speakers often produce wrong tones in TJ style by mistakenly utilizing TJ tone contours which sounds ‘typical TJ’ to them. This fact suggests that their tone production is heavily influenced by ‘word tone’, not ‘pitch accent’. In contrast, most young speakers are a highly competent bidialectal, producing almost perfect TJ tones with correct pitch-accents. Second, social meaning is a promoting factor of innovative tones. Young people preferring to the youth or pop culture in Tokyo area are the main user of the innovative flat tone, and this seems to be closely related to their media practice. Media broadcast contributes to form language ideology in a local community. Individual differences of tone production could be ascribed to local ideology including their attitude to language varieties.

From these findings, I argue that the accentuation of young KJ speakers is shifting towards the pitch-accent type system like TJ and the ongoing change is motivated by these social factors.