INVITED TALK, UMASS AMHERST DEPARTMENT OF LINGUISTICS           MAY 27, 2004

Tonal changes in Kagoshima Japanese

Haruo Kubozono
Kobe University

This paper discusses lexical tone and tonal changes in Kagoshima Japanese on
the basis of fieldwork involving more than 200 informants. This dialect,
which is spoken in the south of Japan, has a unique but endangered
prosodic system strikingly different from that of many other dialects of
Japanese. Its salient features include: (i) it is based entirely on the
syllable rather than the mora, so that tonal patterns are determined by
counting the number of syllables from the end of the word; (ii) it has two
contrastive tonal patterns, called Patterns A and B, which have a high tone
on the penultimate and the final syllable, respectively; (iii) compound
nouns inherit the tonal pattern of their initial member; (iv) the domain of
tone assignment is not the word but the phrase, so that every phrase
inherits the tonal pattern of its first element. 

This paper provides an analysis of ongoing tonal changes in this dialect.
Native speakers aged 30 or under tend to confuse the two tonal patterns by
pronouncing many words traditionally belonging to Pattern A with Pattern B
and vice versa. A close examination indicates that the changes are not
arbitrary but exhibit a striking correlation with tonal distinctions in
Tokyo Japanese: words pronounced with a falling pitch in Tokyo now tend to
show a falling pitch (Pattern A) in Kagoshima, while those pronounced
without a falling pitch tend to be pronounced without a falling pitch
(Pattern B). Interestingly, a similar bi-directional change is observed in
the pronunciation of compound nouns, thus breaching the traditional compound
rule (but not the phrasal tone rule).