Send message to Author

Paper short abstract:

The mid-western Satsuma dialect prefers closed syllables. The syllable-final consonants are distinctive but, in some cases, semi-neutralized. The analysis of the semi-neutralization extracts two contrastive categories. They let us study vague distinctions that have been difficult to treat so far.

Paper long abstract:

The Satsuma and Ōsumi dialects of Japanese prefer closed syllables when compared to other, major Japanese dialects. The closed syllables in Satsuma and Ōsumi end in (1a) a stop, (1b) nasal, (1c) palatal fricative, or (1d) palatal approximant. Since these syllable-final consonants are distinctive, they can be defined as forming the syllable-final contrastive categories (2a) {Q}, (2b) {N}, (2c) {ç}, and (2d) {j}.

Closed syllables in some words, however, can end in either of the two or more consonants listed in (1). This fact shows that these syllable-final consonants are semi-neutralized in particular words.

Analyzing the syllable-final semi-neutralization between (1a-d), we can further define the following five syllable-final contrastive categories.

(3)

a. {Q,N}: realized as a stop or nasal

b. {ç,j}: realized as a palatal fricative or palatal approximant

c. {N,j}: realized as a nasal or palatal approximant

d. {Q,ç}: realized as a stop or palatal fricative

e. {Q,N,ç}: realized as a stop, nasal, or palatal fricative(/approximant)

Satsuma and Ōsumi are famous for their preference for closed syllables, and the closed syllables have been studied from both phonological and historical perspectives. Previous studies, however, paid no attention to syllable-final semi-neutralization between (1a-d) because its occurrence is restricted both among Japanese dialects and within each particular dialect. Furthermore, it is unclear whether contrastive categories as in (2-3) can be defined as phonemes. This is because traditional phonology has trouble in treating different contrastive categories that share the same realization.

With this issue in mind, this paper tackles the issue of whether (2) and (3) can be defined as phonemes. Through examining the distribution of and distinction between consonants in the mid-western Satsuma dialect, a variation of Satsuma and Ōsumi, I will argue for the existence of semi-neutralized but contrastive phonemic categories.

The syllable-final contrastive categories in (2) are not phonemes because their distinction results from the syllable-final neutralization between consonant phonemes. For the same reason, (3a-c) are not phonemes as well. However, (3d-e) can be defined as phonemes because they are distinguishable from all other consonant phonemes.