Skip to main content

What Is Kanji? A Complete Beginner's Introduction

What is kanji? Kanji (漢字, kanji) is the logographic script Japanese inherited from China. Each character writes a morpheme: a unit of meaning that also carries one or more Japanese readings.12 For a beginner who already knows kana, the right mental model is not "a picture of an idea." It is "a written morpheme that the rest of the sentence will pronounce for you."

Overview​

Modern Japanese is written in three scripts at once: kanji (logographic), hiragana (phonetic, for native words and grammar), and katakana (phonetic, for loanwords and emphasis).23 Of the three, only kanji encodes meaning directly; the kana scripts spell sound. For the full breakdown of how the three combine inside a single sentence, see the mixed-script article.

This article defines what kanji is, separates it from the related but different idea of an "ideograph," and frames the practical unit of beginner study. It deliberately stops short of history, the six 六書 (rikusho) categories, the on-yomi / kun-yomi machinery, stroke order, and lookup workflows. Each of those has its own dedicated article in the sidebar.

Where kanji sits in the Japanese writing system​

The three scripts split the work of a Japanese sentence by function, not by appearance.23 Kanji carry the content morphemes (noun stems, verb roots, adjective roots). Hiragana carry the grammatical glue (particles, verb and adjective endings, and native words without a familiar kanji). Katakana carry loanwords, foreign names, scientific terminology, onomatopoeia, and visual emphasis.

The division is functional. Each script does work the other two cannot do efficiently: kanji compress meaning into a single glyph, while kana spell sound and inflection.23

The three-script system is the default, not a flourish

Running modern Japanese prose, including newspapers, government documents, novels, and casual messaging, uses the three scripts together. Pure-kana or pure-kanji text is marked as special: children's books, classical pastiche, or official forms.2

How this article is scoped​

Kanji is a large topic. To keep this introduction focused, this article covers only the conceptual core: what a kanji is, how it differs from kana, why readings travel with words, and how kanji sits inside the wider CJK family. History, the six 六書 categories, the on-yomi / kun-yomi system, stroke order, lookup, and handwriting each have their own articles linked below.

Logographic, not ideographic: what kanji actually encodes​

Phonetic scripts vs logographic scripts​

A phonetic script writes sound. Each kana symbol represents one mora (one short sound unit) and has no meaning by itself. The kana か is the sound ka and nothing else.3

A logographic script writes morphemes. Each kanji represents a unit of meaning that also carries a sound, or a small set of sounds, in the language.12 The kanji 火 writes the morpheme {fire}. It is pronounced ひ (hi) or カ (ka) depending on which word it appears in.

火ひ4
"fire"

The contrast is structural, not stylistic. The kana か and the kanji 火 can share a sound, but only 火 carries the meaning {fire}.1

What "logographic" means in linguistics

"Logographic" means "encoding a word or morpheme of a specific language." It does not mean "encoding pure meaning independent of language." That second sense is what "ideographic" claims. The next section unpacks this distinction.51

Why "ideographic" is almost-but-not-quite right​

Older Western reference works and much beginner-aimed English material call kanji "ideographic" or "ideograms." Japanese linguistic convention disagrees: 漢字 are classified as 表語文字 (hyōgo moji, logographic), not 表意文字 (hyōi moji, ideographic).51

The Japanese Wikipedia entry on 漢字 puts it directly: 「漢字は表語文字であり、それぞれの書記素は言語の意味要素を指す。」 ("Kanji are a logographic script; each grapheme points to a meaning element of the language.")1 The same article notes that kanji no longer directly indicate ideas in the strict ideographic sense. Understanding them depends on knowing the language they are written in.1

The practical consequence is simple. A kanji is not a picture of an idea. It is the written form of a Japanese morpheme that has both a meaning and at least one Japanese reading.12

One character, one morpheme, often several readings​

A single kanji typically maps to more than one reading. The same Chinese-origin grapheme was borrowed into Japanese at different historical periods and was also assigned to pre-existing native Japanese morphemes.2

The mechanics of this, namely the on-yomi / kun-yomi system and its historical strata (呉音, 漢音, 唐音, 慣用音), belong in the dedicated readings article. The takeaway for now is simple: when you see a new kanji, do not assume there is "the" reading. Ask which word the kanji appears in. The word fixes the reading.2

The kanji ↔ vocabulary ↔ reading triangle​

Why memorizing kanji in isolation fails​

The same kanji surfaces with different readings in different words. The character is stable; the reading travels with the word.2

Take 水 ({water}). Three common words use the same character with three different readings:

The kanji 水 is unchanged across all three. The reading shifts every time, fixed by the word it sits inside.674

水みずを飲のみます。4
"I drink water."

水曜日すいようびに会あいましょう。6
"Let's meet on Wednesday."

水着みずぎを持もってきました。7
"I brought my swimsuit."

What to memorize together​

The reusable unit of study is the triple (kanji, vocabulary word, reading in that word), not the lone kanji.2 A list of "all the readings of 水" is information without a home. The word 水曜日 is information that sticks, because the reading is anchored by the company the kanji keeps.

Learn the word, the reading follows

When you meet a new kanji, do not try to memorize its full reading list. Memorize the first word you saw it in, with its reading. Add other words later as you meet them. The character will accumulate readings the way a verb accumulates conjugations: one real example at a time.2

Why this matters for beginner study order​

Because readings live in words, every mainstream modern Japanese textbook introduces kanji inside vocabulary items, not as a separate column of isolated flashcards.2 The character-first approach means drilling a kanji's "list of readings" before learning any words that use them. It is the older, slower, and less reliable route, and beginner methodology has been moving away from it for decades.2 Detailed strategy comparison belongs in the script-order article linked below.

Kanji and grammar: where okurigana comes in​

The stem-plus-tail pattern​

Okurigana (送り仮名) are kana suffixes that follow a kanji stem and complete a Japanese word, especially inflected verbs and adjectives.8 The kanji is the meaning anchor. The kana tail carries the grammar.

Look at 食 ({eat}) across three conjugations:

食たべる9
"to eat (non-past)"

朝あさごはんを食たべた。9
"I ate breakfast."

今日きょうは何なにも食たべない。9
"I'm not eating anything today."

The kanji 食 holds still in all three. The hiragana tail does every bit of the conjugation work.8

Why a logographic script needs a phonetic helper​

A Chinese character, taken alone, is a meaning-bearing unit but a fixed shape. Modern Japanese verbs and adjectives must conjugate for tense, polarity, politeness, and mood, so the writing system needs somewhere to mark those changes.8

Hiragana okurigana carry that grammatical load. The same kanji 食 stays put across 食べる, 食べた, 食べない, 食べさせる, and 食べられる. The conjugation lives entirely in the kana tail.8 This is also why Classical Chinese, written in pure kanji, cannot encode modern Japanese on its own: it has nowhere to put the conjugation.810

Where the rules actually live​

The Japanese Ministry of Education's 送り仮名の付け方 (okurigana no tsukekata, "how to attach okurigana") notification (1973, revised 1981) lays out seven general rules and two rules for difficult cases.8 The full mechanics live in the dedicated okurigana article linked below.

Kanji, hanzi, and hanja: same origin, three roads​

Shared root​

Chinese characters reached Japan from the Asian mainland. The earliest dated material evidence in Japan is the gold seal 漢委奴国王印, conventionally dated to 57 CE. Sustained Japanese literacy in Chinese writing does not appear until roughly the 4th to 6th centuries CE.110

Transmission via the Korean peninsula, notably scholars from Baekje, is part of the traditional account. Some specifics, such as a Baekje origin for man'yōgana, are disputed in mainstream Japanese scholarship.10 Full history is deferred to the kanji-history article linked below.

Where the three diverged​

The same root grew in three different directions. Three things in particular set kanji apart from modern hanzi and hanja:

Two different simplification programs. Postwar Japan promulgated the 当用漢字表 (Tōyō kanji list) on 16 November 1946. The corresponding 新字体 (shinjitai) simplified forms followed in 1949; the Tōyō list was replaced by the 常用漢字 (Jōyō kanji) in 1981 and expanded to 2,136 characters in 2010.11 Mainland China's reform began with the 1956 Chinese Character Simplification Scheme and continued through the 1950s. A second round in 1977 was rescinded in 1986.12 The two reforms produced overlapping but distinct results. For example, the character for "country" is 国 in both modern Japan and mainland China. It was simplified from the traditional 國, which remains standard in Taiwan, Hong Kong, and Macau.1112

Japan-only characters: kokuji (国字). A small set of kanji were invented in Japan rather than borrowed from China. These are also called 和製漢字 (wasei kanji).13 Confirmed examples:

働はたらく13
"to work"

峠とうげ13
"mountain pass"

込こむ13
"to be crowded"

The jōyō list contains roughly nine kokuji; the full inventory across all registers reaches into the hundreds.13

Korean hanja decline. In South Korea, hanja are now very rarely used in everyday writing. They survive mainly in academic, legal, and disambiguating contexts. The decline accelerated sharply in the 1960s and 1970s under educational policy. A 1968 ban on teaching hanja in public schools was reversed in 1992, but the generational gap remained.14 In North Korea, hanja have not officially been used since June 1949.14

The net effect is clear: kanji remains structural to written Japanese, hanzi remains the entire written system for Chinese, and hanja has receded to a marginal role in modern Korean.21415

What this means for learners with Chinese or Korean background​

Prior knowledge of Chinese characters is a real shortcut for recognizing kanji shapes and rough meanings, but it is a trap for readings. A kanji's Mandarin pronunciation is not its Japanese reading. Even the on-readings, which are themselves Sinitic borrowings, are mostly Middle Chinese strata mediated through centuries of Japanese phonology and do not align with modern Mandarin.2

Prior hanja exposure gives a small recognition boost for the shapes and meanings of common characters. But day-to-day modern Korean uses hanja so rarely that the head start is smaller than the hanzi case.14

Same shape does not mean same reading

A Mandarin speaker reading 学生 will pronounce it xuésheng. A Japanese reader pronounces the same two characters がくせい (gakusei). The shapes are nearly identical. The sounds are unrelated. Treat shape recognition as the gift and reading prediction as the work still to do.2

Good to know​

The "kanji is a picture" trap​

Beginner material often describes kanji as stylized pictures of the things they mean. That framing fits only a small minority of the modern character inventory and misses the dominant pattern.

True pictographs (象形, shōkei) account for roughly 18 percent of the 1,026 educational (kyōiku) kanji and around 13 percent of the full 2,136-character jōyō list under Shirakawa's Jōyōjikai classification.16 The dominant category is phono-semantic compounds (形声, keisei). In these characters, one component hints at meaning and another hints at sound. They account for about 49 percent of kyōiku and 61 percent of jōyō kanji.16 The remaining slices are semantic composites (会意), indicatives (指事), and loangraphs.16

The honest summary is that pictographs are a small minority, roughly one in eight jōyō characters. The mechanics of the six categories belong in the dedicated 六書 article linked below.

"How many kanji are there?" is the wrong first question​

Readers often ask it. The short answer is that there is no single number. The 常用漢字 list fixes 2,136 characters for general-purpose modern writing, but the wider inventory reaches into the tens of thousands once historical, literary, and personal-name characters are counted.11 The question that actually matters for a beginner ("how many do I need to read Japanese comfortably?") has its own dedicated article linked below.

Readings live in words, not in characters​

A kanji often has half a dozen attested readings across the lexicon. But in any given word, the reading is one specific value, predictable from the company the kanji keeps (Sino-Japanese compound, native compound, or standalone).2 The takeaway for an early-stage learner is to treat (kanji, word, reading) as one unit of memorization. Learn 水 inside 水, 水曜日, and 水着 as three vocabulary items. The readings みず, すい, and みず attach themselves naturally.674

漢字 literally means "Han characters"​

The word itself records the writing system's external origin. 漢 names the Han Chinese; 字 means "character." The compound was borrowed directly into Japanese as the on-reading kanji of the Chinese term hànzì (漢字).215 Kanji are explicitly "the characters from the Han," not a native Japanese invention. Kokuji are the small set of exceptions noted above.213

仮名 ("borrowed name") versus 真名 ("true name")​

The very names of the scripts encode an old cultural hierarchy. 仮名 (kana) is 仮 (kari, "borrowed, provisional, false") plus 名 (na, "name").3 The contrasting term 真名 (mana, "true name") was used for kanji proper, "true" in the sense of carrying real meaning rather than being borrowed merely as a phonetic label.3

The pair is the cleanest one-line explanation for why kanji feels heavier than kana in Japanese text, even before a learner can read any kanji. Mana (kanji) was for serious, meaning-bearing writing. Kana was for sound-only, supplementary writing. The distinction was historically gendered in Heian usage as well.3

Handwriting and reading are separable skills​

Reading modern Japanese fluently does not require handwriting kanji. IMEs and digital input have decoupled production from recognition, and a learner can become a strong reader while typing exclusively.2 Whether to invest in handwriting is a separate pedagogical decision with arguments on both sides. That decision is deferred to the dedicated handwriting article linked below.2

See also​

References​

Footnotes​

  1. 『漢字』. ウィキペディア日本語版. https://ja.wikipedia.org/wiki/%E6%BC%A2%E5%AD%97 ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9

  2. "Kanji". Wikipedia. https://en.wikipedia.org/wiki/Kanji ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18 ↩19 ↩20 ↩21 ↩22

  3. "Kana". Wikipedia. https://en.wikipedia.org/wiki/Kana ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7

  4. "水". Jisho.org (drawing on JMdict / EDICT, Electronic Dictionary Research and Development Group). https://jisho.org/word/%E6%B0%B4 ↩ ↩2 ↩3 ↩4

  5. 堀田隆一 (Hotta Ryūichi). 「#3856. 漢字は表意文字か表語文字か」. hellog 〜英語史ブログ〜, 慶應義塾大学. https://user.keio.ac.jp/~rhotta/hellog/2019-11-17-1.html ↩ ↩2

  6. "水曜日". Wiktionary. https://en.wiktionary.org/wiki/%E6%B0%B4%E6%9B%9C%E6%97%A5 ↩ ↩2 ↩3

  7. "水着". Wiktionary. https://en.wiktionary.org/wiki/%E6%B0%B4%E7%9D%80 ↩ ↩2 ↩3

  8. "Okurigana". Wikipedia. https://en.wikipedia.org/wiki/Okurigana; citing 文部省 (Ministry of Education) notification on 送り仮名の付け方, 1973, amended 1981. ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  9. "食べる". Wiktionary. https://en.wiktionary.org/wiki/%E9%A3%9F%E3%81%B9%E3%82%8B ↩ ↩2 ↩3

  10. "Japanese writing system". Wikipedia. https://en.wikipedia.org/wiki/Japanese_writing_system ↩ ↩2 ↩3

  11. "Tōyō kanji". Wikipedia. https://en.wikipedia.org/wiki/T%C5%8Dy%C5%8D_kanji ↩ ↩2 ↩3

  12. "Simplified Chinese characters". Wikipedia. https://en.wikipedia.org/wiki/Simplified_Chinese_characters ↩ ↩2

  13. "Kokuji". Wikipedia. https://en.wikipedia.org/wiki/Kokuji ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  14. "Hanja". Wikipedia. https://en.wikipedia.org/wiki/Hanja ↩ ↩2 ↩3 ↩4

  15. "Hanzi". Wikipedia (entry redirects to Chinese characters). https://en.wikipedia.org/wiki/Chinese_characters ↩ ↩2

  16. "Formation Types of Educational and Joyo Kanji". Kanji Portraits. https://kanjiportraits.wordpress.com/2021/11/19/formation-types-of-educational-and-joyo-kanji/; methodology attributed by the post to 白川静 (Shirakawa Shizuka), 常用字解, 平凡社, 2004. ↩ ↩2 ↩3