For which parts of speech does a pitch accent contrast between word-final H(-H) (unaccented, 平板 heiban) and H(-L) (nucleus on last syllable, 尾高 odaka) exist?

Someone tells me "mainly nouns, but also adjectives". Is this correct, or are there other word classes for which this contrast exists as well? If so, which?

Finally, I also welcome advice on how to ease the memorization process. I'm thinking of the following method: Flat-accented words will be left untouched in my vocabulary decks; those with a word-final accent should be subvocalized with an additional suffix. For nouns, I will simply subvocalize a final -は, as in [あかさ(は)]{LHHLLL} (for a hypothetical noun あかさꜜ). For other parts of speech, I might need to adjust this method, depending on which suffixes the given part of speech allows.

  • As for the final paragraph, I can make a separate question out of it or move it to Meta; please anyone let me know if you recommend this. Commented May 19 at 8:31
  • It may be only me but I have no idea what contrast you are talking about. Do you have any examples?
    – aguijonazo
    Commented May 19 at 10:33
  • @aguijonazo It's the oft-discussed contrast between 端(は) and 橋(は). Commented May 19 at 11:13
  • 1
    They are just two of three words that happen to be different only in pitch accent. What's so special about those two that doesn't apply to the difference between 箸 and either one of them? In Kansai-ben, they have three distinct pitch patterns even without a particle.
    – aguijonazo
    Commented May 19 at 12:32
  • @aguijonazo The accent of 箸 is clear from saying it aloud in isolation. The value of the question is didactic: For which parts of speech (at least: nouns and adjectives) or word classes does one need to worry about adding a dummy mora afterwards when learning the word's accent pattern via reading it aloud? Commented May 19 at 14:11

1 Answer 1


Aside from nouns, na adjectives such as 好き may potentially be accented on their final mora (e.g. [好きな]{LHL}, OJAD).

Verbs and i-adjectives show limited accentuation options. Timothy Vance’s descriptions of the accentuation patterns of various inflected forms in The Sounds of Japanese (2008) appear consistent with a generalization that the forms of verbs and i-adjectives never have an underlying accent on the final mora, although I didn’t see an explicit statement to this effect.

Per Vance, there are a few situations where a verb or adjective form may receive an accent on the final mora when followed by another word, but there is no contrast here between underlyingly accented and unaccented lexemes. Vance describes it in terms of adding an accent to the last syllable of an unaccented verb or adjective form in the following circumstances:

  • The gerund of an unaccented verb gains an accent on the last syllable when followed by /wa/ or /mo/, as in /maketeꜜwa/ and /maketeꜜmo/. (Vance, pages 168-169)

  • When the dictionary form of an adjective is followed by です (forming the polite nonpast affirmative) or by the pronoun/nominalizer の, an accent is placed on the final syllable of the dictionary form. For nearly all adjectives, the final syllable of this form contains a diphthong (/ai/, /oi/, /ui/) or the long vowel /ii/, in which case the accent falls on the second-to-last mora: /aꜜi/, /oꜜi/, /uꜜi/, /iꜜi/. (In Tokyo Japanese, it's normally impossible for an accent to be placed on the second mora of a long vowel or of a diphthong.) But 遠いです and 遠いの are pronounced as /tooiꜜdesu/, /tooiꜜno/ with an accent on the final mora, suggesting 遠い is here syllabified as /too.i/. (Vance, pages 178-179) Interestingly, OJAD displays the alternative accentuations /tooꜜidesu/ and /toꜜoidesu/ for 遠いです; I don't know any more about this.

  • See also In what cases do 平板 verbs change their pitch? and Pitch accent of nominalizers

One exception is accented verbs like 吹く where the second-to-last mora has a devoiced vowel; some speakers reportedly shift the accent onto the following, final mora. As far as I know, there is no verb where shifting the accent from the second-to-last mora to the last mora is obligatory.

I'm only an occasional learner so cannot offer much practical advice.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .