Chapter 6
An Emerging Gender Difference in
Japanese Vowel Devoicing
Terumi Imai, Wittenberg University
- Introduction
The Japanese high vowels /i/ and /u/ (the latter phonetically [؉]) tend to be
devoiced when they occur between voiceless consonants or after a voiceless
consonant and before a pause (Sakuma 1929; Martin 1952; Han 1962; McCaw-
ley 1968; Kawakami 1977; Nihon Housou Kyoukai (NHK) 1985; Vance 1987;
Maekawa 1983, 1988; Sugito 1988; Kondo 1994, 1995; Nagano-Madsen 1994,
1995; Tsuchida 1997; Yuen 1997; Varden 1999). This phenomenon has been stud-
ied extensively in terms of its mechanism and phonetic environments in the ¿ eld of
phonetics (Sakuma 1929; Martin 1952; Han 1962; Sawashima 1971; Sawashima
and Niimi 1974; Kawakami 1977; Yoshioka 1981; Yoshioka, Löfqvist, and
Hirose 1982; Maekawa 1983; Beckman and Shoji 1984; Nihon Housou Kyoukai
(NHK) 1985; Sugito 1988; Jun and Beckman 1993; Kondo 1994, 1995; Nagano-
Madsen 1994, 1995; Tsuchida 1997), but this study investigates the social aspects
of vowel devoicing in Tokyo Japanese, particularly the effects of gender and age
and how these factors inÀ uence the rate of vowel devoicing.
I will give an overview of Japanese vowel devoicing and the factors that
play a major role in this phenomenon in Section 2. Next, the data collection
and analysis techniques are discussed in Section 3. Section 4 summarizes the
results of the statistical analyses, and Section 5 concludes the chapter with future
research suggestions.
- Japanese vowel devoicing
2.1 What is vowel devoicing?
Although the most unmarked, or natural, realization of a vowel is a voiced
vowel, when a high vowel occurs between voiceless consonants, it tends to