This article aims to cover the Japanese syllabaries for those who are already somewhat familiar with the system. First we begin with moras:
A mora is the smallest unit of rhythm in the Japanese language. It is usually a vowel or consonant + vowel combination. There are roughly around 100 moras. Each mora is given the same amount of time when pronounced; so sounds are produced with even rhythm. Examples: あ、ち、ぬ、し、しゅ、ちゃ、っ、か
As you can see there are some other moras that don't fit the V / VC model. Let's look at those more closely.
Yōon includes sounds such as にゃ、しゅ、みょ. These sounds are palatalized with the exception of しゃしゅしょ which use the same ɕ sound from し. Each yōon constitute one moraic unit.
Ex:
辞書（
塾（
This is the mora っ which represents consonant germination (doubling the length of consonants) of the following mora as seen in:
Ex:
雑誌（
でっかい（
This is the mora ん which represents the moraic nasal as seen in:
Ex:
全然（
本気（
This was a feature of the traditional Tokyo dialect; where the が row was pronounced with a nasal as かﾟ. Most speakers will use the が row without nasalization, however this phenomenon can still be observed today by singers and announcers. Note: NHK announcers use ﾟ to symbolize nasalization of the が row. Its usage doesn't go beyond that.
Ex:
Moving on from moras we are going to take a look a relatively small set of changes to Japanese pronunciation that effect the way you read some mora sequences.
Sino-Japanese (SJ) words underwent changes as a result of the coalescence of vowels that created diphthongs. Many SJ words contain 長音 (
Ex: 高校・カウカウ → コウコウ →
If the morpheme of an SJ word is written with an:
お段 (おこそとのほ…) mora followed by the mora う then pronounce the う as an extended vowel （Ex: 投稿 as
え段 (えけせてねへ…) mora followed by the mora い then pronounce the い as an extended vowel（Ex: 衛星 as
う段 (うくすつぬふ…) mora followed by the mora う then pronounce the う as an extended vowel （Ex: 空港 as
As moras are the basic units of Japanese, syllables are made up of one or moras. Japanese contains two types of syllables:
heavy syllable:
2 moras long
composed of an independent mora 自立拍 (
light syllable
1 mora long
a singular 自立拍 (
特殊拍 include:
長音（
促音（
撥音（
二重母音（
自立拍 is everything else.
Let's try identifying some syllables. Guess how many syllables each word has and which syllables are heavy / light. Click the spoiler to check. Heavy syllables will be bolded.
食べる (たべる)：
高校 (こうこう)：
信じる (しんじる)：
昨日 (きのう)：
使った (つかった)：
名古屋 (なごや)：
いい：
学校 (がっこう)：
切符 (きっぷ)：
外国 (がいこく)：
転校生 (てんこうせい)：
手入れ (ていれ)：
日本 (にほん)：
アメリカ：
中国 (ちゅうごく)：
関西 (かんさい)：
目医者 (めいしゃ)：
政治 (せいじ)：
So how does all of this relate to pitch accent?
Japanese Standard Dialect has a falling pitch contour. Each phrase-element starts with an initial rise in pitch and mantains that pitch until a marked syllable, where at that point the pitch drops and stays low until the next phrase-element. At the sentence level we can observe down-stepping (catathesis) where after each phrase-element the pitch drops ever so slightly. These drops are more prominent following an accented (non-flat) phrase-element (Kubozono 2015).
Words that cannot be distinguished by their reading are distinguished by either a contrast in:
lack of accented syllable (unaccented vs accented) Ex: 気 (
if both are accented then the location of the prominent syllable. Ex: 箸 (
Accent phrase-elements exhibit an initial rise in pitch and then produce a falling contour when a word is accented. phrase-elements that start on an accented word produce a falling contour from the start.
Example: はし
端が：
箸が：
橋が：
Example: あめ
飴が：
雨が：
？が：
Example: きのう（キノー）
帰納は：
機能は：
昨日は：
？？は：
*This not only doesn’t exist but would be impossible since のう is pronounced ノー here and is a heavy syllable. As we know the accent must fall on a specific syllable let's dive into what that means.
It is easier to reason about pitch accent by using syllables instead of mora. For example we can note that: the accent can never be placed on the second mora of a heavy syllable. Thus the accent is always placed on the first mora of any given syllable. This is useful in setting expectations for what you hear and limiting the number of possible options.
Example: 新幹線
Makeup: シン・カン・セン
There are 3 syllables so only 3 places you can place an accent (if there is one).
Options (given the rules we know thus far):
Example: 転校生
Makeup: テン・コー・セー
There are 3 syllables so only 3 places you can place an accent (if there is one).
Options (given the rules we know thus far):
Additionally when an accent phrase-element starts with a heavy syllable containing a 長音 (
Japanese dictionaries typically use mora counting to label pitch accent. A number is placed to tell you what mora the accent is on. 0 is used for unaccented (flat) words.
Ex:
[0] = (unaccented)
銀行
アメリカ
消える
[1] = (accented, 1st mora)
社会
会議
議会
[3] = (accented, 3rd mora)
表現
条件
出られる
For some sets of words it is easier to count from the back. Suffixes that may change the pitch of words are easier to process by counting from the back of the word. Like before 0 is used for unaccented words. -1 is used for the last mora, -2 for the second to last, so on and so forth.
[0] = (unaccented)
[-2] = (accented, 2nd to last mora)
回数
無数
奇数
[-4] = (accented, 4th to last mora)
産業革命
科学革命
Labeling lexical pitch accent by its type is pretty convenient especially for i-adjectives and verbs.
平板
頭高
中高
尾高
起伏
形容詞 (
Note very few 形容詞 (
Ex:
LHH (H) = unaccented; 3 moras long i.e. 希望, 自分, 辞める
LHH = accented on 3rd mora; 3 moras long i.e. 相手, 女, 男
HLLL = accented on 1st mora; 4 moras long i.e. 概念, 魂 , 挨拶
(H) at the end means that the word is [0] as opposed to [-1]
＼ marks the fall of the accent; a lack of ＼ means the word is unaccented. Colloquial usage of ↓ among natives may vary.
世界（セ＼カイ）[1]
真面目（マジメ）[0]
犬（イヌ＼）[2]
納得（ナットク）[0]