Definition:
Vowel devoicing is a phonological process in which a normally voiced vowel is produced without vocal fold vibration — effectively whispered or silent — typically in specific phonological environments. The phenomenon is best known in Japanese, where the high vowels /i/ and /u/ are regularly devoiced between voiceless consonants (such as /k/, /s/, /t/, /h/, /p/) or in word-final position after a voiceless consonant.
Also known as: 母音の無声化 (boin no museika)
In-Depth Explanation
All vowels are inherently voiced — they are produced with vibrating vocal folds. Vowel devoicing overrides this default: the vocal folds stop vibrating during the vowel, producing a breathy whisper or complete deletion. The result is that the vowel becomes nearly inaudible while the surrounding consonant framework remains.
Japanese Vowel Devoicing
Japanese is the most-studied case. The rule affects high vowels /i/ and /u/ when they appear between two voiceless consonants, or after a voiceless consonant at the end of an utterance:
- です (desu) → [desɯ̥] — the final /u/ is devoiced, sounding like “des”
- すき (suki) → [sɯ̥ki] — the /u/ between /s/ and /k/ is devoiced
- きた (kita) → [ki̥ta] — the /i/ between /k/ and /t/ is devoiced
- ください (kudasai) → [kɯ̥dasai] — initial /u/ after /k/ and before /d/
- した (shita) → [ɕi̥ta] — /i/ between /ɕ/ and /t/
The phenomenon is gradient rather than categorical — speakers devoice more in fast speech, formal contexts, and Tokyo/standard Japanese. Regional dialects (particularly western and Kansai dialects) devoice less frequently.
Other Languages
Vowel devoicing is not unique to Japanese:
- French: High vowels can devoice in similar environments (petite → [pəti̥t])
- Korean: Vowels between voiceless stops may partially devoice
- Turkish: High vowels /i, ɯ, u, y/ devoice between voiceless consonants
- Quebec French: High vowels undergo devoicing and deletion between voiceless consonants
Why It Matters for Learners
L2 learners of Japanese who fail to devoice sound noticeably foreign. Pronouncing desu with a fully voiced [u] or suki with a clear [u] marks non-native production immediately. Conversely, understanding devoicing helps with listening comprehension — learners who expect to hear every vowel clearly will struggle with natural-speed Japanese.
Devoicing interacts with Japanese pitch accent — a devoiced syllable cannot carry a pitch accent, which shifts to an adjacent syllable.
Key Researchers
- Maekawa Kikuo — extensive corpus studies of Japanese devoicing patterns
- Vance, Timothy — The Sounds of Japanese (2008), comprehensive treatment