Script & grammar

Why kanji stops being hard around character 300 Copy 2

Most learners quit kanji in the first eight weeks right before the radicals start doing the work for them. Here is what changes at the turn, and how to reach it in half the time.

S
Sukhada Sensei
Founder of Asahii Institue

9 min read

Every batch has the same shape. Weeks one to three are energetic. Week four, hiragana clicks. Week six, katakana clicks. Week eight, kanji arrives and about a third of the room quietly decides they are not a kanji person.

They are wrong, but not unreasonably so. At eighty characters, kanji genuinely is memorisation. There is no system yet, no pattern to lean on — just shapes. The learner is doing the hardest possible version of the work, and getting the least possible return for it.

The eight-week wall

The wall is not difficulty. It is absence of leverage. Below roughly 150 characters, every new kanji costs the same as the last one. You memorise 三, then you memorise 山, and neither one helped you learn the other. Effort scales linearly, motivation does not.

So students do what anyone would do with a flat cost curve and no visible progress: they switch to grammar, where the returns feel immediate. Six months later they can hold a conversation and cannot read a train station sign.

注


The mistake is the order, not the effort

Learners who front-load vocabulary and postpone kanji almost always hit N4 fast and then stall for a year. Learners who take kanji seriously from week eight are slower until N4 and considerably faster after it.

What actually changes at 300

Somewhere between 250 and 350 characters, the cost curve bends. Not because your memory improved — because you have finally accumulated enough components that new characters stop being novel shapes and start being recombinations of shapes you already own.

This is the single most under-communicated fact in Japanese teaching. The difficulty of kanji is front-loaded, and almost nobody tells beginners that.


"Kanji is not two thousand things to memorise. It is roughly two hundred things to memorise, and then a lot of reading."

~ Sukhada Sensei, on the first day of every N4 batch

Radicals are a compression format

A radical is a reusable component with a stable meaning. Once you know 木 means tree, three characters come nearly free:

  • 木ki — tree

  • 林hayashi — grove

  • 森mori — forest

  • 板ita — board

One radical, four characters. The fourth adds 反 as a phonetic hint — which is the next layer of the system.

There are around 214 traditional radicals; in practice about 80 of them do most of the work at N5–N3. Learn those 80 deliberately and you are not learning kanji one at a time anymore — you are learning a notation.

  • Meaning radicals tell you the semantic field. 氵 means the character has something to do with water — 海, river 河, wash 洗.

  • Phonetic components tell you the likely on'yomi reading. 青 (sei) shows up in 清, 晴, 請 — all read sei.

  • Position matters. The same component means different things left, right, top and bottom. This is learnable in an afternoon and saves months.

How we sequence it in class

Asahii does not teach kanji in JLPT frequency order, which is how most textbooks do it. Frequency order is optimised for exam coverage, not for how a human brain builds a system. We teach it in component order, then reconcile with the exam list.

  1. Weeks 1–7: no kanji at all beyond recognition. Hiragana and katakana get full attention, because a shaky phonetic base makes everything after it slower.

  2. Weeks 8–14: 80 high-yield radicals, taught as meaning units with stroke logic. Roughly 15 minutes per class, every class.

  3. Weeks 15–20: the N5 set of 120 characters, introduced as radical combinations rather than as new shapes.

  4. Weeks 21+: reading volume. Once the system is in place, exposure does more than drilling — and this is the point where most self-study plans have already collapsed.

A realistic timeline by level

Numbers people ask us for, with the caveat that a 90-minute class three times a week is assumed:



N5

Foundations~120 kanji. Radical system introduced. Reading is still slow and deliberate.

4–5 months


N4

The turn~400 kanji. The cost curve bends here. Most students notice it themselves.

4–5 months


N3

Compounding~700 kanji. New characters mostly decode from components you already have.

5–6 months


N2

Reading, not studying~1000 kanji. Acquisition shifts from class time to volume of text.

6–8 months

What to do this week

If you are stuck at the wall right now, the intervention is small and specific:

  • Stop learning characters. Spend one week on radicals only — 80 of them, meaning and position.

  • Re-approach the fifty characters you already know and decompose each one. You will recognise most of the parts.

  • Only then continue forward. The next fifty will take noticeably less time than the last fifty did.

That is the whole method. It is not clever and it is not fast in week one — it is just correctly ordered, which turns out to matter more than either. If you want it done with structure and someone correcting your stroke order in real time, that is what a batch at Asahii is for.

-WHY ASAHII

Start with a conversation, not a sign-up form.

Tell us where you’re starting from student or company and we’ll take it from there.

-WHY ASAHII

Start with a conversation, not a sign-up form.

Tell us where you’re starting from student or company and we’ll take it from there.

-WHY ASAHII

Start with a conversation, not a sign-up form.

Tell us where you’re starting from student or company and we’ll take it from there.