JLPT N5 Kanji: Your First 100 Characters
The ROI of your first 100 characters
100 kanji sounds like rounding error against the 2,136-character joyo set. It is not. The frequency distribution of written Japanese is a power law, and the first 100 characters sit on the fattest part of the curve. Nothing you learn later — no batch of 100, ever — will pay back what these do.
The coverage curve
Kanji obey Zipf's law the way English words do. Chikamatsu et al. (2000) ran a full year of the Mainichi Shimbun — 23 million kanji tokens, 4,000+ distinct characters — and the cumulative coverage curve has the steep early shape every Zipf distribution has.
Zipf on a log-log axis. A near-straight line is the tell of a power law. Kanji frequency in Japanese news prose has the same shape — which is the entire reason the first 100 characters carry the load they do. Source: Wikimedia Commons.
| Kanji Known | % of Kanji Occurrences | Marginal Gain per 100 Kanji |
|---|---|---|
| 100 | ~45% | 45.0% |
| 300 | ~72% | 13.5% |
| 500 | ~80% | 4.0% |
| 777 | ~90% | 3.6% |
| 1,000 | ~96% | 2.7% |
| 1,600 | ~99% | 0.5% |
| 2,136 (Joyo) | ~99.9% | 0.2% |
Sources: Chikamatsu et al. (2000), Mainichi Shimbun corpus; scriptin/kanji-frequency news dataset; BCCWJ Balanced Corpus of Contemporary Written Japanese, Joyce et al. (2012).
Read the table top to bottom. 100 characters buys you 45% of every kanji token in news text. The next 100 add 13. By character 500, each additional hundred is worth four points. By the time you're grinding rare characters past 1,600, you're paying full price for half a point. The early region is the most asymmetric trade you'll ever get with this language.
JLPT levels vs. school grades
The JLPT and the Japanese school system (教育漢字) carve the same character space along different axes. The JLPT optimizes for adult functional literacy. The school system optimizes for what a seven-year-old can draw without crying.
| Level | Kanji | Cumulative | School Equivalent | Grade Kanji (Cumulative) |
|---|---|---|---|---|
| N5 | ~100 | ~100 | Grade 1--2 | 80 + 160 = 240 |
| N4 | ~170 | ~270 | Grade 2--3 | 240 + 200 = 440 |
| N3 | ~370 | ~640 | Grade 3--4 | 440 + 202 = 642 |
| N2 | ~380 | ~1,020 | Grade 5--6 | 642 + 193 + 191 = 1,026 |
| N1 | ~1,100 | ~2,136 | Junior/Senior High | 2,136 (full Joyo) |
N5 lands roughly on the Grade 1–2 allocation — 240 kyōiku kanji. The overlap is imperfect on purpose. N5 keeps high-frequency adult-life characters like 電, 駅, and 飲 that schools defer because the strokes are gnarly. Grade 1 keeps 石 and 虫, which are easy to draw and almost never show up in a JLPT reading passage. For an adult, frequency wins. Always.
29 of your first 100 are Kangxi radicals
This is the compounding part of the trade. At least 29 N5 kanji are also Kangxi radicals — the 214 classification keys used to index every CJK dictionary printed since 1716:
一 (R1), 二 (R7), 八 (R12), 十 (R24), 口 (R30), 土 (R32), 大 (R37), 女 (R38), 子 (R39), 山 (R46), 川 (R47), 日 (R72), 月 (R74), 木 (R75), 火 (R86), 父 (R88), 生 (R100), 白 (R106), 行 (R144), 西 (R146), 見 (R147), 言 (R149), 車 (R159), 金 (R167), 長 (R168), 門 (R169), 雨 (R173), 食 (R184), 高 (R189).
Learn 水 as a character and you've simultaneously learned radical 85 and its left-side variant 氵 (sanzui) — the component sitting on the left of 海, 泳, 河, 池, 泡, 洗, 清, 湖, and a few hundred more. Each of these 29 characters is a node in a dependency graph. Learn the node, unlock recognition of everything that contains it. This isn't a metaphor. It's how dictionary lookup works, and it's how the visual word form area of your cortex decomposes compound characters into parts.
The compound explosion
The real word yield is combinatorial. 100 characters means a theoretical pair space of 100 × 99 = 9,900 two-kanji compounds. Most of those aren't words. But the actual yield is still striking — N5 kanji form several hundred common compounds. A sample:
| Kanji | High-Yield Compounds |
|---|---|
| 日 | 日本, 今日, 毎日, 日曜日, 休日, 明日, 一日 |
| 人 | 日本人, 一人, 大人, 人口, 外国人 |
| 大 | 大学, 大人, 大きい, 大切, 大変 |
| 中 | 中国, 中学, 日中, 一年中, 中心 |
| 学 | 大学, 学生, 学校, 学年, 入学 |
| 生 | 学生, 先生, 生活, 一生, 生まれる |
日 alone shows up in at least seven common N5 words. 生 pulls double duty — on'yomi in compounds like 学生 (gakusei), kun'yomi in verbs like 生まれる (umareru). Each new N5 character doesn't add one word to your vocabulary. It multiplies the words available from every character you already know.
Optimal ordering: frequency beats grade
A Yamanote Line platform sign at 東京駅 (Tōkyō-eki). 東, 京, and 駅 are all N5 characters an adult traveler reads within thirty minutes of landing at Narita. Source: Wikimedia Commons.
School grade order, JLPT order, or raw frequency order? For an adult, frequency wins. The school curriculum is sequenced for a child's hand and a child's life — 石 (stone) and 虫 (insect) before 電 (electricity) and 駅 (station), because the strokes are simpler and a six-year-old's world has more bugs than train timetables. An adult in Tokyo sees 電車 and 駅 every commute. They see 虫 in a biology textbook, maybe. The JLPT ordering is a frequency filter wearing a test's clothing, and that's why its coverage on the Zipf curve is so efficient.
So: your first 100 characters aren't 4.7% of joyo. They're 45% of every kanji token you'll read, 29 of the 214 keys to the entire CJK indexing system, and several hundred real words once you start combining them. No future batch of 100 will pay back like this.
The Grade 1 deck covers the 80 first-graders. The Kanji Atlas shows how those 80 compose every other Grade 1 character through their parts. Start there.
References
- Chikamatsu, N., Yokoyama, S., Nozaki, H., Long, E., & Fukuda, S. (2000). A Japanese logographic character frequency list for cognitive science research. Behavior Research Methods, Instruments, & Computers, 32(3), 482–500.
- Joyce, T., Horoscek, M., & Nishina, Y. (2012). Kanji coverage for BCCWJ-based corpus word lists. National Institute for Japanese Language and Linguistics.
- Scriptin. Kanji usage frequency. scriptin.github.io/kanji-frequency
- Japanese Complete. Top 777 Kanji by Frequency. japanesecomplete.com/777
- Kyoiku kanji. Wikipedia. en.wikipedia.org/wiki/Kyoiku_kanji