May 9, 2026 · Learning Hiragana

Is it normal to forget hiragana after learning it?

Yes. Almost everyone does. You're not losing your memory. You're running into a well-documented pattern, and there's a specific fix.

← Back to blog

You knew it yesterday

You studied the chart. You could name every character. You felt solid. Then you woke up the next morning and stared at さ with absolutely nothing coming to mind. Or you could get through the first two rows and then everything after な turned into a blur.

This happens so often that it's one of the most common questions on Japanese learning forums. People genuinely worry something is wrong with them. It's not. What you're experiencing was documented in 1885 by a psychologist named Hermann Ebbinghaus, and it happens to everyone who learns anything new.

The forgetting curve

Ebbinghaus memorized lists of nonsense syllables and tracked how quickly he forgot them. The result: about 70% of new information disappears within 24 hours if you don't actively try to retrieve it. This is the forgetting curve, and hiragana characters are especially vulnerable to it because when you're starting out, they are nonsense syllables to your brain.

か doesn't mean anything to you yet. It's a shape. Your brain has no reason to keep it. So it drops it overnight, the same way it drops a phone number you glanced at once.

The forgetting curve isn't a flaw in your memory. It's your brain being efficient. You encounter thousands of pieces of new visual information every day. Your brain keeps the ones it expects to need again and lets go of the rest. The problem is that looking at a chart once doesn't signal "you'll need this again." You have to send that signal yourself.

Why reviewing the chart doesn't fix it

The instinct when you forget is to go back and look at the chart again. Re-read the rows. Stare at the characters until they feel familiar. This feels like studying, but it doesn't produce durable memory.

There's a difference between recognition and recall. Recognition is seeing か on a chart and thinking "oh right, that's ka." Recall is seeing か with no hints and producing "ka" from memory. Recognition feels like knowing, but it's a weaker form of memory. It's why you can ace a hiragana quiz when the characters are in order and then blank out when you see them in a random word.

Karpicke and Roediger (2008) tested this directly. Students who practiced retrieval, actively pulling answers from memory, retained 80% of material after a week. Students who reviewed the same material by re-reading retained 35%. Same study time. Completely different results.

What actually makes hiragana stick

Two things working together: retrieval practice and spacing.

Retrieval practice means you look at a character and try to name it before checking the answer. That moment of effort, even when you get it wrong, is what strengthens the memory. Your brain flags the information as something it failed to produce and gives it higher priority next time.

Spacing means you don't do all your practice in one sitting. Studying hiragana for five minutes a day over two weeks beats studying for an hour twice. Cepeda et al. (2006) confirmed this across 254 studies: distributed practice consistently outperformed massed practice, regardless of the type of material being learned.

Flashcard apps with spaced repetition handle both of these automatically. Characters you keep forgetting come back sooner. Characters you know get pushed out to longer intervals. You don't have to think about scheduling. You open the app, do five minutes, and the algorithm handles the rest.

The characters everyone forgets

Some hiragana are harder to retain than others, and it's not random. The characters that cause the most confusion share visual features.

The usual suspects: き and さ (both have a horizontal stroke crossing a vertical one), ぬ and め (both are loops with a tail), は and ほ (identical left side, small difference on the right), わ and ね (similar curves, different endings). These pairs interfere with each other in memory because your brain stores them in the same visual category.

Hunt and Worthen (2006) found that memory improves when you focus on how similar items differ, not on what they share. The practical takeaway: studying confusable pairs side by side actually helps, because it forces your brain to encode the differences. Running through the full chart in order doesn't do this because the similar characters are separated by rows of unrelated ones.

When I was learning hiragana, I could get through あ through な without much trouble. The second half of the chart is where everything fell apart. I kept mixing up similar-looking characters because I'd learned them in isolation instead of comparing them. Once I started drilling the confusable pairs together, the mix-ups stopped within a few days.

How long until it sticks for good

With daily five-minute sessions using retrieval practice and spaced repetition, most people can reliably read all 46 hiragana within two weeks. "Reliably" means you can identify any character out of order, in unfamiliar contexts, without hesitating.

After that initial two weeks, the characters still need occasional reinforcement. If you stop practicing entirely, the forgetting curve doesn't disappear. It just gets shallower. Characters you haven't seen in a month might get fuzzy. Characters you review even once a week stay solid.

The good news is that once you move on to reading actual Japanese words, hiragana reinforcement happens naturally. You can't read a Japanese sentence without hiragana. Every word you encounter is practice.

SimplyHiragana is a free app built around retrieval practice and spaced repetition. Every character comes with a visual mnemonic, and the app schedules reviews based on what you're forgetting. Five minutes a day, two weeks, forty-six characters.

If you want to know whether flashcards are the right approach for how you personally learn, there's a free 3-minute quiz at howyoulearn.org. Some people retain better through writing or audio. Knowing your learning style first saves you from grinding through a method that doesn't fit.

References

Helpful links