I'm Sona Orbit, the solo developer of Fudebo. The first six sections cover learning methods you can use with pen and paper. Section seven introduces the app I'm making. Refer to this site for its features. The numbers in this article come from the cited papers. I put this together in November 2026.
1. Why rereading a kana chart doesn't stick
After looking at a kana chart again and again, the characters start to feel familiar. It's easy to mistake that feeling for knowing them, but it may not last.
In an experiment by Roediger and Karpicke (2006), college students studied the same short passage in four different ways. On a test 5 minutes later, the group that kept rereading scored highest. A week later, the result flipped: the group that read once and took three tests remembered 61%, while the group that read four times remembered 40%.
The same problem comes up with kana. The more familiar the chart feels, the easier it is to overestimate what you know. To find out, cover the chart and try writing the characters.
2. Try to remember before checking
Trying to recall an answer is practice in itself. Psychologists call this retrieval practice. Pulling something out of memory makes it easier to recall next time.
- A meta-analysis of 61 studies (Rowland, 2014) found the same pattern: testing helped people remember more than rereading. Seeing the correct answer right after trying to recall it helped even more.
- Another meta-analysis, covering 86 studies and more than 17,000 participants (Bertsch et al., 2007), found that people remembered answers they produced themselves better than answers they simply read. This is called the generation effect.
For kana, the routine is simple. Look only at the reading, such as “ki.” Write the kana, then check your answer. It's okay if you can't remember it. Trying first and then checking helps more than looking at the answer right away.
Sources: Rowland (2014), Psychological Bulletin; Bertsch et al. (2007), Memory & Cognition
3. Write by hand, then write without a guide
When learning a new writing system, handwriting works better than just looking at the characters or typing them.
- Wiley and Rapp (2021) had 42 adults who had never studied Arabic learn 20 Arabic letters. One group wrote by hand, one typed, and one only looked at the letters. The handwriting group reached the target in the fewest training sessions on average. The benefits also carried over to tasks they hadn't practiced, such as saying letter names and reading words.
- In an experiment by Longcamp et al. (2008), adults recognized new characters they had learned by hand better than ones they had learned by typing. The advantage lasted several weeks.
- Naka (1998) had Japanese elementary school students learn unfamiliar letters. Writing helped them remember more than just looking, but tracing made that advantage smaller.
Tracing is a starting point. After tracing once or twice, put the example away and write on your own.
What about stroke order? In the same study by Naka, children who wrote with a different stroke order still remembered the letters well. That suggests much of the memory benefit comes from writing itself. Using the correct stroke order helps you write neatly, follow Japanese writing conventions, and read handwritten characters later. Both matter, but you don't need perfect stroke order to remember a character.

Sources: Wiley & Rapp (2021), Psychological Science; Longcamp et al. (2008), Journal of Cognitive Neuroscience; Naka (1998), Memory & Cognition
4. Write them again a day or two later
For the same amount of practice, spreading it across several days helps you remember more than doing it all at once.
Cepeda et al. (2006) reviewed data from more than 14,000 participants. People who spread out their practice remembered 47% on average, compared with 37% for people who did it all in one session. The longer you want to remember something, the longer the gap between practice sessions should be.
How much should you practice? In a study of 533 students, Rawson and Dunlosky (2011) recommended practicing until you can recall the answer correctly 3 times, then reviewing it 3 more times on spaced-out days. Those later reviews did more than extra practice at the start.
Here's one way to do it: test yourself tomorrow on kana you missed today. Test yourself again on kana you got right after 2 days, 4 days, and a week. Each time, write from memory before looking.
Sources: Cepeda et al. (2006), Psychological Bulletin; Rawson & Dunlosky (2011), Journal of Experimental Psychology: General
5. Practice similar-looking kana together
These kana are easy to mix up:
| Similar kana | Readings | What to look for |
|---|---|---|
| ぬ/め | nu/me | ぬ ends with a small loop |
| ね/れ/わ | ne/re/wa | At the lower right, ね loops, れ flicks outward, and わ curves inward |
| る/ろ | ru/ro | る ends with a small loop |
| シ/ツ | shi/tsu | シ has two more horizontal marks on the left, and its final stroke goes up; ツ has two more vertical marks at the top, and its final stroke goes down |
| ソ/ン | so/n | ソ has a mark at the top and a final stroke that goes down; ン has a mark on the left and a final stroke that goes up |
Quiz yourself on these groups together instead of practicing each character separately. That makes the differences easier to spot. In an experiment by Kornell and Bjork (2008), students who alternated between paintings by different artists identified new paintings correctly 61% of the time. Students who saw one artist's work before moving to the next got 35% correct. A meta-analysis of 59 studies (Brunmair & Richter, 2019) also found that alternating practice helped more when the categories looked more alike.
There is a limit: in that meta-analysis, learning vocabulary one set at a time worked better than alternating. Use alternating practice to tell similar-looking kana apart. Study words in the usual way.
Sources: Kornell & Bjork (2008), Psychological Science; Brunmair & Richter (2019), Psychological Bulletin
6. A 10-minute daily routine with pen and paper
Get a grid notebook and a kana chart. Look at the chart only when checking your answers.
| Step | What to do | Why |
|---|---|---|
| 1. New kana (3) | Look at the chart and trace each one twice. Cover the chart and write it once more | Tracing gets you started; writing without the chart is the practice |
| 2. Review (about 10) | Look only at the reading. Write from memory, then check your answer | Try to recall before looking |
| 3. Kana you missed | Write each one twice in the correct stroke order. Mark it and test yourself tomorrow | Bring back missed characters sooner |
| 4. Similar-looking kana | If you missed one today, quiz yourself on its whole group in turns | Compare them side by side to see the differences |
Keep the number of new kana small. Around 3 a day is enough. Review matters more than new material. If your review list grows, skip new kana for a day. At that pace, you can write through all 46 basic hiragana in about two to three weeks.
7. How I built Fudebo around these methods
The hardest part of doing this with pen and paper is making your own prompts, checking your answers, and keeping track of when to review each character. I built Fudebo to handle those tasks. Its App Store name is "Fudebo: Write Kana from Memory." It focuses on handwriting:
- The first time you open it, there are 4 intro lessons. Each new row starts with a short lesson, and each new kana has a lesson card with a stroke-order animation, pronunciation, writing tips, the character's origin, look-alikes to watch for, and example words. Then you trace it, write it over a faint guide, and finally write it from memory in an empty box.
- After that, a prompt shows only the reading or romaji. You write with your finger from memory, and the app checks the stroke count, order, direction, and shape.
- If you get it wrong, the wrong stroke turns red and the correct stroke order plays. You trace it once, then write it on your own. Kana you miss go on tomorrow's page; kana you get right come back at longer and longer intervals.
- Your daily amount is counted in writes. You can set it from 10 to 200; the default is 40, about 10 minutes. You can also change how prompts are given, how much help you get, how many times you write a new kana from memory, how strict the checking is, and the grid.
- To write more, open Unlimited writing in Practice: a kana counts as learned after you write it correctly 3 times in a row without help. Look-alike showdown, Dictation, Hiragana ↔ Katakana, and the 60-second challenge are in Pro.
Fudebo works on iPhone with iOS 17 or later, and on iPad too. It's available on the App Store. The free version teaches the 46 basic hiragana and lets you use Unlimited writing with the hiragana you've learned.
8. Frequently asked questions
Should I learn hiragana or katakana first?
Start with hiragana. It's used most often in Japanese. Katakana is mainly used for loanwords and names. Once you're comfortable writing the 46 basic hiragana, start katakana.
Can I practice by typing on my phone?
You can use typing as extra practice, but handwriting is better for remembering character shapes. In the experiment by Wiley and Rapp (2021), the handwriting group reached the target in the fewest training sessions, followed by the typing group. The group that only looked at the letters took the longest.
Do I have to use the correct stroke order?
I recommend it. It helps you write neatly and follow Japanese writing conventions. But research suggests the memory benefit comes mainly from writing. A different stroke order can still help, so don't worry too much about it when you're starting out.
How long should I practice each day?
About 10 minutes a day is enough. Short daily sessions with a review the next day help you remember more than practicing for an hour all at once.
Is Fudebo free?
It's a free download on the App Store. The free version teaches the 46 basic hiragana. Katakana, dakuten, handakuten, yoon, and 66 basic words are in Pro.
