Gyō and Dan: Reading the Kana Chart as Consonants and Vowels

The Japanese kana chart, the gojūon, is built from 10 gyō (行) and 5 dan (段). A gyō is a group of kana that share a consonant; a dan is a group that share a vowel. The か行 (ka-gyō) is かきくけこ, ka ki ku ke ko. The あ段 (a-dan) is あかさたなはまやらわ, a ka sa ta na ha ma ya ra wa.

Japanese children learn these two words at school, and you will meet them too: in textbooks, in dictionaries, and in grammar explanations. This post explains what they mean, what you see when you set the chart beside romaji, where the pattern breaks, and where the terms come back later. How to memorize the chart itself is covered in How to Memorize the Kana Chart.

Don’t learn them by direction

English explanations often call gyō “rows” and dan “columns”. That works for a chart laid out horizontally. But Japanese charts are just as often printed vertically, read from the right, and there the あいうえお of the あ行 runs top to bottom. The same gyō can be vertical or horizontal depending on the chart.

So I think it is better to learn the two words by what they contain, not by which way they run.

• Gyō (行): kana with the same consonant. Everything in the か行 begins with k.
• Dan (段): kana with the same vowel. Everything in the い段 ends in i.

Another pair that gets mixed up is あ行 and あ段, since both start with あ. The あ行 is the five vowels あいうえお. The あ段 is the ten kana that end in a: あかさたなはまやらわ.

What romaji shows

In romaji, the structure jumps out.

	a-dan	i-dan	u-dan	e-dan	o-dan
a-gyō	a	i	u	e	o
ka-gyō	ka	ki	ku	ke	ko
na-gyō	na	ni	nu	ne	no
ma-gyō	ma	mi	mu	me	mo

Across, every kana in the ka-gyō starts with k. Down, every kana in the a-dan ends in a. か and き look nothing alike, yet in romaji you can see they share a k. That is the whole idea of the chart: almost every kana is one consonant plus one vowel, so instead of 46 separate shapes you are learning 5 vowels and 10 rows.

Where the pattern breaks

Romaji also shows where the pattern fails.

In Hepburn romaji, し is shi, ち is chi, つ is tsu and ふ is fu. Their consonants are spelled differently from the rest of their gyō, because they are pronounced a little differently. The Kunrei system writes them si, ti, tu and hu instead, following the chart exactly. Both systems are explained in Hepburn vs Kunrei Romaji.

For a learner, it is enough to know these four as the kana whose sound shifts. Beyond that, the や行 and わ行 have empty cells, and ん belongs to no gyō or dan at all. The reasons are in Why the 50-Sound Kana Chart Has Only 46 Hiragana.

Dakuon and yōon, seen the same way

Dakuten change the gyō and leave the dan alone. Add ゛ and the か行 becomes the が行, the さ行 becomes the ざ行, the た行 the だ行, and the は行 the ば行. か turns into が, but both are still in the a-dan.

Yōon, the combined sounds, are all built from the i-dan. きゃ, きゅ and きょ are き plus a small ゃ, ゅ or ょ. しゃ, ちゃ and にゃ work the same way: the first kana always comes from the i-dan. And きゃ is kya, きゅ is kyu, きょ is kyo, so the vowel at the end comes from や, ゆ or よ. More in Dakuon, Handakuon and Yōon.

Where gyō and dan come back

If you go on to study Japanese grammar, the terms return.

You may know verbs like 書く (kaku, to write) as “u-verbs” or “godan verbs”. Godan (五段) means “five dan”. Look at what happens to 書く: 書かない (kakanai), 書きます (kakimasu), 書く (kaku), 書けば (kakeba), 書こう (kakō). The ending walks through か, き, く, け and こ, every dan of the か行. That is why Japanese grammar calls these verbs 五段活用, five-dan conjugation.

Verbs like 起きる (okiru) and 食べる (taberu), often called “ru-verbs”, are called 上一段 (kami-ichidan) and 下一段 (shimo-ichidan) in Japanese. Ichidan means “one dan”: the sound before る stays in the i-dan (き) or the e-dan (べ). Once you know gyō and dan, these names stop being arbitrary labels.

Japanese dictionaries are also ordered by the gojūon: あ, い, う, え, お, then か, き, and so on. Knowing the chart by heart makes looking things up faster.

Comparing them in StudyX Japanese Letters Table

StudyX Japanese Letters Table is a kana chart app where every cell plays its sound. The a row is on the right and the wa row on the left, the same order as a chart written vertically.

The buttons at the bottom switch the chart between hiragana あ, katakana ア, upper-case romaji A and lower-case romaji a. The layout stays the same and only the characters change, so the かきくけこ column in hiragana becomes ka, ki, ku, ke, ko in romaji, in the same place. The romaji is Hepburn, so し shows as shi and つ as tsu. You can see the pattern and the exceptions side by side.

The ゃ button turns on the yōon chart. The a-dan cells become きゃ, しゃ and so on, the u-dan cells きゅ, and the o-dan cells きょ, while the i-dan and e-dan cells keep their original kana. The chart shows the yōon lining up with their vowels.

In the developer interview I said, “I always want to design the experience so the child can discover something on their own.” Pressing the buttons and watching how the chart changes is one way to do that, and it works for adult learners as well.

Tap a kana and you hear it. There are no points or rewards. The app runs in landscape on both iPhone and iPad. On iPad there is a writing area under the chart for a finger or Apple Pencil. It is a one-time purchase with no ads and no in-app purchases. The on-screen labels are in Japanese. For a first session, see Getting Started with StudyX Japanese Letters Table.