Background¶

Japanese pitch accent¶

Japanese language is a pitch-accented language. While English accents are based on the strength each syllable, Japanese accents are based on the pitch (height) of each mora. Accent is important in spoken Japanese language, as it is in other languages.

There are large regional differences in Japanese accents. Of these, this module deals with the accents of the contemporary Tokyo dialect, which is one of the most influential Japanese dialects in the country.

On the Tokyo Japanese accent¶

Raise and lower (accent nucleus)¶

In the contemporary Tokyo dialect, there are two kinds of events that invoke the change of pitch.

[: raise the pitch here
]: lower the pitch here (a.k.a. accent nucleus)

Not all accent patterns are possible. Usually, at most one ] can appear in a word. Also, [ can appear only at the beginning. Thus, an n-mora word basically has one of the n+1 possible accent patterns as follows.

*[****
*]****
*[*]***
*[**]**
*[***]*
*[****]

Accent type: representation of accent nuclei by digits¶

Accent patterns are often expressed by the digits which indicate the positions of the accent nuclei. This digit is often called an accent type. Some accent dictionaries including UniDic use this notation.

Accent type = 0 means that there are no accent nuclei (no fall ]).
Accent type = n (n > 0) means that the accent nucleus appears after the n-th mora.

Heiban, Atamadaka, Nakadaka, and Odaka types¶

Accent type 0 is also called the Heiban (平板; flat) type.
Accent type 1 is also called the Atamadaka (頭高; head high) type.
Accent type 2 ~ n-1 are also called the Nakadaka (中高; middle high) types.
Accent type n is also called the Odaka (尾高; tail high) type.

For examples,

0 (Heiban)
- 野球ya[kyuu, パソコンpa[sokon, 鉛筆e[npitsu, 緑茶ryo[kucha, りんごri[ngo, 渋谷shi[buya
1 (Atamadaka)
- サッカーsa]Qkaa, ジュースju]usu, 猫ne]ko, メロンme]ron, 金魚ki]ngyo, 新橋shi]mbashi
2 ~ n-1 (Nakadaka)
- バドミントンba[domi]nton, 折り紙o[ri]gami, カブトムシka[buto]mushi, 冷蔵庫re[ezo]oko, 池袋i[kebu]kuro
n (Odaka)
- 足a[shi], 紙ka[mi], 花ha[na], 海苔no[ri], 米ko[me], 光hi[kari], 犬i[nu], 馬u[ma]

Difference between Heiban and Odaka¶

Although there may seem to be no difference in the pitch pattern within the words, Heiban and Odaka types have differences in the pitch of the subsequent particles e.g. “ga が” which is the NOM case marker for subjects.

hana ga はなが
- 鼻が ha[na -ga: nose is
- 花が ha[na] -ga: flower is
hikari ga ひかりが
- ひかりが hi[kari -ga: Shinkansen Hikari is
- 光が hi[kari] -ga: light is
hashi ga はしが
- 橋が ha[shi] -ga: bridge is
- 端が ha[shi -ga: edge is
- 箸が ha]shi -ga: chopsticks are
ha ga はが
- 葉が ha[ -ga: leaf is
- 歯が ha] -ga: tooth is

Note that the GEN case markar “no の” is often unaffected by the preceding downforce. (This law also has some exceptions.)

hana no はなの
- 鼻の ha[na -no: nose’s
- 花の ha[na -no: flower’s
hikari no ひかりの
- ひかりの hi[kari -no: Shinkansen Hikari’s
- 光の hi[kari -no: light’s
hashi no はしの
- 橋の ha[shi -no: bridge’s
- 端の ha[shi -no: edge’s
- 箸の ha]shi -no: chopsticks’
ha no はの
- 葉の ha[ -no: leaf’s
- 歯の ha] -no: tooth’s

Compounds¶

The accents of compound words are a little complicated. For example,

東京to[okyoo
東京都to[okyo]oto

東京都知事to[okyootochi]ji

東京都知事選挙to[okyootochijise]nkyo

世界se]kai
遺産i[san
世界遺産se[kaii]san

機械ki]kai
学習ga[kushuu
機械学習ki[kaiga]kushuu

Logo of tdmelodic¶

There are many pairs of words that have same sound patterns except the accentuation. For example,

fuji
- 富士 (Mt. Fuji) is pronounced as fu ] ji.
- 藤 (wisteria) is pronounced as fu [ ji.
sake
- 鮭 (salmon) is pronounced as sa ] ke.
- 酒 (alcoholic beverage) is pronounced as sa [ ke
hashi
- 端 (edge, corner) is pronounced as ha [ shi.
- 橋 (bridge) is pronounced as ha [ shi ].
- 箸 (chopstics) is pronounced as ha ] shi.

As you can see, the logo is a combination of two “fuji”-s with different accent patterns.