Why WaniKani radicals have weird names
New WaniKani users keep hitting the same speed bump in the first few weeks: the radicals have names like "leaf", "fins", "explosion", and "drop". Search those names anywhere outside WaniKani and nothing comes up. Open a Japanese dictionary and you get totally different names. So which is right?
Short answer: both are right; they serve different purposes. This guide is the longer explanation.
Two kinds of "radical"
The word "radical" in kanji learning means two different things depending on who is using it.
The traditional sense. In a Japanese dictionary, each kanji has a single official "radical" (ι¨ι¦, bushu) - the component used to look it up. There are 214 traditional radicals in the Kangxi system, inherited from Chinese. Each has a name in Japanese (e.g. θε kusakanmuri for the grass-top shape) and an associated meaning category. Useful if you are looking up a kanji by stroke index in a paper dictionary; less useful in 2026.
The mnemonic sense. WaniKani treats radicals as building blocks for mnemonics: recurring shapes that show up across kanji, named whatever helps the brain remember the kanji they appear in. The names are deliberately concrete and visual. "Fins" looks like fins. "Leaf" looks like a leaf. The point is that ζ (hold) contains "hand" and "ground" and "ten", and the mnemonic "I'm holding the ground with my hand for ten seconds" sticks better than "this kanji's bushu is ζ".
Why WaniKani chose mnemonic names
Tofugu's explicit reasoning, summarised from their blog posts over the years:
- Memorability beats etymology. "Slide" is easier to picture than γδΈΏγ. The vivid concrete name produces a stickier mnemonic. The etymology can come later, if at all.
- One name per shape, even when a shape has multiple historical roots. A traditional radical book treats variants as different radicals; WaniKani collapses them when they look the same on the screen.
- Splits where traditional radicals merge. The other direction: WaniKani sometimes treats a single traditional radical as multiple visual components if breaking it apart makes the mnemonic cleaner.
- Made-up shapes when convenient. Some "radicals" on WaniKani are not standard components at all - they are convenient visual groupings the team invented to make a particular kanji's mnemonic land. "Eggplant", "tofu", "viking" are not real radicals anywhere outside WaniKani.
When the names stop mattering
Around level 15-20 most learners stop caring about the radical names entirely. The mnemonics have done their job - the kanji themselves are now atomic units in your brain - and you no longer need to walk the chain "drop + leaf = some kanji". You just recognise the kanji.
At that point the WaniKani-specific names become harmless: a leftover scaffold from the early levels. They do not collide with anything in real Japanese, because in real Japanese nobody looks at a kanji and thinks "ah, that's the leaf radical plus the ten radical". Kanji are read as wholes.
When the names DO get in the way
If you want to look up a kanji in a Japanese dictionary. Paper dictionaries use the traditional radicals and stroke counts. If you are doing serious lookup work (calligraphy, kanji scholarship, certain JLPT prep books), you will need to learn the traditional radical system separately. For most learners this is "later, if ever".
If you study with materials from elsewhere. Heisig (Remembering the Kanji), Marshall's book, jpdb, Kanken prep books all use different systems. You will see the same shape called "grass" in one, "kusakanmuri" in another, and "leaf" only on WaniKani. Pick a primary system, treat the others as translations.
Should you memorise the real radicals?
No, not as a separate project. The traditional radical system is useful for two specific things: paper- dictionary lookups, and Kanken-style writing exams. If you are doing neither, the time is better spent on grammar or reading practice.
If you ever do need them, picking them up takes a weekend, not weeks. You already know the shapes from WaniKani; only the names are new.