Reference sheet

Tenten and maru

The twenty-five marked hiragana — each one beside the character it is built from, with stroke order — and the seventy-three words they unlock. Made to print, and meant to sit beside the stroke chart for the original forty-six.

Taught in lesson 5 · stroke data from KanjiVG, © Ulrich Apel, CC BY-SA 3.0 · readings and glosses from Jisho.org (JMdict)

1.The two marks

• Dakuten, the “tenten” — two short strokes, as in か → が. It voices the consonant: the mouth holds still and the vocal cords start. Taken by the か, さ, た and は rows.
• Handakuten, the “maru” — a small circle, as in は → ぱ. It turns h into p. Taken by the は row alone.

The two marks on their own, enlarged: ゛ ゜ — on a real character they are a fraction of this size, which is exactly why they get skipped.

Both are written last, after the base character is finished, in its top right corner. A tenten adds two strokes, a maru adds one. Nothing else about the hand movement changes.

2.The grid

Twenty base characters, twenty-five marked ones — the は row appears twice because it is the only row that takes both marks.

Read the sounds, not the letters: じ is ji (from shi, not from si), ぢ is also ji, and づ is zu, identical to ず. Everything else in the grid is the plain consonant plus the vowel at the top of its column.

3.Stroke order

Numbers mark where each stroke begins. The caption names the character each one grew out of, so it is always visible how many strokes are old and how many are the mark.

4.The three irregulars

And one rule about position: the maru row (ぱ ぴ ぷ ぺ ぽ) practically never begins a word and never falls between two plain vowels. It appears after ん or after a small っ — さんぽ, きっぷ, えんぴつ, てんぷら.

5.The words

Seventy-three words: two kana, at least one of them marked, both drawn from the characters taught so far, and every one tagged common by JMdict. No word here appears on the two-kana word list — that sheet holds the words with no marks in them, and together the two have no overlap at all.

The final group, まる, is longer than two characters and has to be: no common word puts a maru character anywhere but after ん or っ, so the ぱ row has no two-kana vocabulary to teach. One word for each of the five.

6.What is not here

• ぢ and づ have no word card, because there was no common word to give them. That absence is the fact worth memorising.
• Words rejected for not being common enough, though they are perfectly real: ざる (colander), かば (hippopotamus), やぎ (goat), はなぢ (nosebleed). If one turns up in class, it is not wrong — it just did not clear the bar this list uses.
• The group headings on the two-kana word list needed marks that list was not allowed to use — からだ, どうぶつ. They are readable now.

Readings, glosses and commonness tags from Jisho.org, serving JMdict/EDICT. Arabic checked against Madinah Arabic and Almaany. Voicing and the は-row history after Vance, The Sounds of Japanese (Cambridge, 2008). For the original forty-six characters, see the hiragana stroke chart.