192 Midori Yonezawa Morris
I conducted a perception experiment to examine how Tokyo and Kinki
people judge a speaker as a local or a non-local person based on devoicing
variation and pitch accent pattern. This chapter reports the results of their
judgments, focusing on voicing variants. It also reports how social factors
such as gender and age affect judgments.
- Previous studies
2.1 Vowel devoicing in the Tokyo dialect
In previous studies, different aspects of devoicing in the Tokyo dialect
have been extensively reported. Those include physiological characteristics
(Yoshioka 1981), phonology, and variability (Imai 2004; Han 1962; Sugito
1969). Devoicing is avoided when a devoiceable vowel is in an accented
(high-pitched) mora (Vance 1987; and others), in a mora that carries intona-
tion (Vance 1987), in a successive devoicing environment (Kondo 1999; and
others), in an unaccented high-pitched mora (Han 1962), and at a morpheme
boundary (Tsuchida 1997; and others). Han (1962) suggested some features
that affect frequency of devoicing, and Sugito (1969, 1988) also reported
variation in devoicing. Imai (2004) reported environments and features that
promote devoicing in production based on a large amount of production
data. Maekawa (1983) states that vowel devoicing may be required as a norm
in the society, and it is prescribed in dictionaries and in the training of
announcers and teachers of Japanese. Yuen (1997) and Imai (2004) point
out, however, that devoicing also has a non-standard character because vow-
els are devoiced more frequently by men and in casual and rapid speech.
2.2 Vowel devoicing in the Kinki dialect
Generally it is believed that vowel devoicing does not occur in the Kinki
dialect or it is described as “less frequent in Kinki” (Tsujimura 1996). Data
from previous studies, however, show that devoicing does occur there and
with more than minimal frequency. In a large sound database (Tahara et
al. 1998), 33 out of 40 tokens of devoiceable vowels were devoiced. Nakai
(1991) reported that sentences ending /u/ in Kinki were regularly devoiced
by elementary school children, and Sugito (1969, 1988) shows variation in
devoicing in Kinki as well as in Tokyo. Finally, Fujimoto (2004) found that