Kanji Radicals (部首): The 214 Kangxi Indexing System
Kanji radicals are the 214 classifying heads that traditional character dictionaries use to file every kanji under exactly one section.12 The word learners meet on WaniKani or in a Heisig deck looks the same, but it means something looser. The two definitions collide whenever a learner opens a different tool.345
Overview
What a radical is, in one sentence
A kanji's 部首 (bushu, "radical") is the single component chosen as that character's dictionary-indexing head. There is exactly one per kanji, no more and no fewer.16 Everything else inside a kanji that is not the radical is a component (or, in James Heisig's terminology, a primitive). The radical is just the one component the dictionary editor picks to file the entry under.375
Wiktionary glosses 部首 as "the categorizing element of a kanji character, used for indexing in a character dictionary."6 The Japanese Wikipedia opens with the same idea: 部首 is the heading-character of the category (部) under which a kanji is filed in a 漢字字典 or 漢和辞典 (kanwa-jiten, "kanji-Japanese dictionary").1
How this article is scoped
This page covers four threads: the strict dictionary-grounded definition, the 214-radical inventory, the 1716 Kangxi Dictionary that fixed that inventory, and the warning about how modern apps redefine the word.
The seven by-position category names (へん, つくり, かんむり, あし, たれ, にょう, かまえ) are previewed only. Other topics have their own dedicated articles in this subcategory: the full radical-versus-component treatment, the top-50 working reference list, and the semantic and phonetic component material. This overview routes to them rather than reproducing them.
What a radical actually is: the strict definition
The "one head per character" rule
部首 is morphologically 部 + 首: 部 (ぶ) means "section, group, department," and 首 (しゅ) means "head, chief."16 Read literally, 部首 is "section head," the head-character of the dictionary section the kanji is filed under.
In a 漢和辞典, every entry is filed under exactly one 部. The character that names that 部 is its 部首.1 The "exactly one" rule is structural, not just an editorial preference. In the Shuowen Jiezi system, in principle a single character belongs to one and only one section.1 Later dictionaries sometimes cross-list a character in multiple sections (重出, "double listing") as a reader convenience, but the canonical radical assignment is still a single choice.1
部首は一文字につき一つだけ決まっています。1
"A kanji is assigned exactly one radical per character."
Outlier Linguistics puts the rule bluntly: a single kanji has only one radical, no matter how many components it has. The choice "is made by the editor of a kanji dictionary, not an inherent part of the nature of kanji."37
Many of the 214 radicals are themselves common kanji (人, 口, 心, 一, 日, 月). When the character is on the list, it heads its own section: its radical is itself.2
What the radical does for you, the reader
The radical does two things, and only the first is structurally guaranteed.
Function (a): paper lookup. The radical is the first key in a 部首索引 (bushu sakuin, "radical index"). The second key is the remaining stroke count after the radical is subtracted. Together they locate the entry in a kanwa-jiten.289
Function (b): a meaning hint, sometimes. Radicals are often, but not always, chosen as the meaning-bearing component of a phono-semantic compound (形声文字 keisei-moji). So knowing the radical can hint at the character's meaning family.103 The water radical 氵 tends to mark liquid-related characters. The hand radical 扌 tends to mark action-related characters.
The hint is a side effect of the filing decision, not the definition. Outlier Linguistics puts the constraint explicitly: "Kanji radicals have a single purpose: indexing kanji in a dictionary. They are not designed to help you learn Japanese kanji, and they are not the building blocks of kanji."3
漢和辞典では、漢字を部首で分類します。1
"In a kanwa-jiten, kanji are classified by their radical."
A deeper treatment of meaning prediction belongs to the semantic-components article. This overview only states that the hint exists and that it is a side effect.
Why "the radical of 明 is 日" but "the radical of 朝 is 月"
Both 明 and 朝 visibly contain both 日 (sun) and 月 (moon), yet the dictionary editor chose a different radical for each. 明 ("bright") is filed under Kangxi radical 72, 日. Wiktionary's radical line records it as "Kangxi radical 72, 日+4, 8 strokes," meaning 日 is the head and the 月 portion supplies the 4 residual strokes used to file the entry.11 朝 ("morning") is filed under Kangxi radical 74, 月, with 8 residual strokes for the rest of the character.12
「明」の部首は日です。11
"The radical of 明 is 日 (the 'sun' radical)."
「朝」の部首は月です。12
"The radical of 朝 is 月 (the 'moon' radical)."
The same component appears in both characters; in one it is the radical, in the other it is just a component. Radical assignment is a filing convention, not a property of the kanji's visible shape.
The 214 radicals: an inventory, not a learning list
How the 214 are organized
The inventory is exactly 214 entries.2813 They are sorted strictly by stroke count, from 1-stroke through 17-stroke. The median is 5 strokes, and the average is slightly below 5.7.2 Radical 1 is 一 ("one"), one stroke. Radical 214 is 龠 ("flute"), seventeen strokes.214
部首は画数順に並んでいて、一が最初です。2
"The radicals are arranged by stroke count, with 一 at the head of the list."
The list is the spine of every traditional radical-indexed CJK dictionary. The standard reference Japanese kanwa-jiten is the Dai Kan-Wa Jiten (Tetsuji Morohashi, 1955–1960; over 50,000 character entries). It files its entries against the 214 Kangxi radicals, subdivided by the total number of remaining strokes in the character.9
Radical variants by position
Many of the 214 take a visually compressed form when they sit in a fixed structural slot of a character. Common cases from the by-frequency reference include 水 → 氵 (left side), 心 → 忄 and ⺗ (left side and bottom), 火 → 灬 (bottom), 人 → 亻 (left side), and 手 → 扌 (left side).10 These compressed forms are listed in the 214 as variants of their parent radical, not as additional entries. That is why the inventory stays at 214.210
The seven positional category names (へん, つくり, かんむり, あし, たれ, にょう, かまえ) name where the variant sits in the character grid. This overview previews the names only. The full taxonomy, with examples for each category, lives in the dedicated article on radicals by position.
How many you actually meet
The List of Kanji Radicals by Frequency reports the coverage curve for the 2,136 jōyō kanji.10 The top six radicals (口, 水/氵, 木, 人/亻, 手/扌, 心/忄/⺗) occur in one quarter of jōyō kanji.10
The top twenty cover half of jōyō kanji, and the top fifty-two cover three quarters.10
上位の六つの部首だけで、常用漢字の四分の一を占めます。10
"Just the top six radicals account for a quarter of the jōyō kanji."
In everyday Japanese, you meet a heavily skewed subset of the 214. The long tail of rare radicals contributes the remainder. The working top-N list for a learner belongs to the dedicated top-50 radicals reference. Here, only the shape of the distribution matters.
Where the 214 came from: the 1716 Kangxi Dictionary
From 540 to 214: Shuowen Jiezi, Zihui, Kangxi
The 214-radical inventory has three predecessors, each one a step toward the modern indexing model.
The Shuowen Jiezi (説文解字, c. 100 CE), compiled by Xu Shen during the Eastern Han dynasty, was the first dictionary to organize its entries into sections according to shared components called radicals. Xu Shen sorted the Chinese lexicon into 540 sections. Their section headers are generally referred to as "radicals."15 This is the origin of radical-based filing, at a 540-radical scale.
The Zihui (字彙, 1615), edited by the Ming-dynasty scholar Mei Yingzuo, introduced the 214-radical system and replaced the older 540-radical Shuowen Jiezi inventory.13 Mei also established the radical-and-stroke sorting method. In that method, graphs belonging to a single radical are arranged by the number of residual strokes.13 This is the modern shape of radical indexing.
The Kangxi Dictionary (康熙字典, 1716), compiled at the order of the Kangxi Emperor from 1710, adopted the Zihui 214 wholesale. The particular set of 214 radicals was first used in the Zihui and is now largely known as the Kangxi radicals because of the Kangxi Dictionary's distribution and authority.8 The 1716 Kangxi Dictionary soon became the standard dictionary of Chinese characters.13
Why the system survived the move to Japan
The Kangxi 214 became the universal reference for CJK lexicography. The Kangxi Dictionary remains a reference standard for Unicode character encoding and modern lexicography across the region.8
Japanese kanwa-jiten inherited the inventory directly. The definitive Japanese 漢和辞典, the Dai Kan-Wa Jiten, files its characters against the 214 Kangxi radicals, subdivided by remaining stroke count.9 The same indexing model carries into digital tools. Jisho.org is built on KANJIDIC2 and Radkfile (EDRDG kanji files), which index against the 214-radical system.16
部首制度は『説文解字』に始まり、『字彙』で二百十四になりました。1315
"The radical system began with the Shuowen Jiezi, and was reduced to 214 in the Zihui."
Why modern apps use "radical" to mean something else
What WaniKani and Heisig call a "radical"
WaniKani's publisher, Tofugu, describes radicals as "'building blocks' for creating kanji" and compares them to "the letters of the alphabet."17 On WaniKani, the learner sees several "radicals" per kanji and uses them as named pieces inside a mnemonic story.417
Heisig's Remembering the Kanji calls reusable kanji pieces primitives, not radicals. Wikipedia summarizes the system this way: primitives "may be traditional radicals, other kanji themselves, or a collection of strokes not normally identified as independent entities." A single kanji is broken into several primitives linked by a mnemonic scene.5 The primitive set is therefore a superset of the traditional radical list, not a synonym for it.5
The two systems differ in terminology: Heisig keeps "primitive," while WaniKani keeps "radical." But both deliberately allow multiple meaning-named pieces per kanji. Under either system, a single kanji has many "radicals" or "primitives," directly contradicting the dictionary-grounded one-per-character rule above.1345
Why they redefined it
The looser definition is a teaching convenience. Tofugu argues that traditional radical names are too generic to anchor a vivid mental image, so WaniKani substitutes invented names chosen to be "specific and easy to imagine in your mind's eye."17 The named pieces then drive mnemonics that compress the kanji-learning timeline from years into months.17
Heisig's teaching case is the same: invented mnemonic names for each primitive let the learner construct a memorable scene. As Wikipedia summarizes the method, "a kanji's written form and its keyword are associated by imagining a scene or story connecting the meaning of the given kanji with the meanings of all the primitives used to write that kanji."5
The cost is a terminology collision. The strict-radical school argues that calling visual chunks "radicals" leaves the learner with a word that does not match what any 漢和辞典, or any digital tool built on KANJIDIC, uses it to mean.3716
Which definition this site uses
J-Compass uses 部首 / "radical" in its strict, dictionary-grounded sense throughout: one per kanji, the indexing head.163 For the looser learner-facing pieces with mnemonic names, J-Compass uses "component" (or, when faithfully reporting Heisig's framing, "primitive").5
WaniKaniの「ラジカル」は伝統的な部首ではなく、覚えやすい部品のことです。4
"WaniKani's 'radicals' are not the traditional radicals; they are mnemonic-friendly components."
When reading another learner's notes, a forum post, or a textbook, check which definition is in play before quoting "the radical of X is Y."37 The same English word can mean an indexing head in a kanwa-jiten or a mnemonic chunk on a flashcard app. The full treatment of the conflict lives in the dedicated radicals-versus-components article in this subcategory.
Good to know
Treating the radical as a meaning guarantee
The radical hints at the meaning family for many kanji because radicals are often the meaning-bearing element of phono-semantic compounds. But the hint is not a rule. The radical of 明 is 日, yet the kanji means "bright," not "sun-related." The radical is a filing-system choice, not a meaning rule.103
Outlier Linguistics puts the constraint clearly: kanji radicals have a single purpose, indexing kanji in a dictionary.3 A fuller treatment of when the meaning hint holds, and when it does not, belongs to the dedicated semantic-components article.
Reading the radical as a stroke-order anchor
Knowing a character's radical does not tell you the writing order of its strokes. Stroke order is governed by its own ruleset (top-to-bottom, left-to-right, enclosing-strokes-first, and so on). The radical name does not predict it. The full ruleset has its own dedicated article on kanji stroke counting and stroke order.
Reading 部首 as "section head"
部 (ぶ) is the same morpheme as in 部分 (bubun, "part / section"), an N3-level word.618 首 (しゅ) is the same morpheme as in 首相 (shushō, "prime minister"), literally "head minister."619 Reading 部首 as "section head" is the cleanest one-line argument for the strict-definition rule. There is one head per section, therefore one radical per kanji.16
One head, one section, one kanji
The Kangxi index is built on three matching "ones": one radical per section (部), one section per kanji entry, and one head-character per section (首).128 If you lose any of the three, the indexing breaks. If you keep them, the strict definition follows.
"Radical" in English-language learning material is ambiguous
Outside academic linguistics and 漢和辞典 reference work, "radical" in casual English usually means the WaniKani- or Heisig-style component with a mnemonic name, not 部首.37 Outlier Linguistics flags the collision explicitly, and Tofugu's own framing acknowledges that WaniKani's radical names diverge from the traditional ones.417 When another learner says "the radical of X is Y," ask which definition is in play before quoting it back.
Component selectors versus canonical radicals in digital dictionaries
Most digital dictionaries (Jisho, Yomitan, and several iOS kanji apps) let you pick several visual pieces to filter the character set. That selection interface uses the looser components model.16 The result list, however, still files each character under its single canonical radical. Jisho relies on KANJIDIC2 and Radkfile, which index against the 214 Kangxi radicals.16
The two models coexist in the same toolchain. The radical-lookup workflow itself belongs to the dedicated article on how to look up a kanji you don't know.
See also
- What Is Kanji? A Complete Beginner's Introduction
- The History of Kanji: From Oracle Bones to the Jōyō List
- The Jōyō Kanji List (常用漢字): The 2,136-Character Set Explained
- How to Learn Kanji: A Strategic Overview of Heisig, WaniKani, and Kanji-in-Context