The Skeleton of Kanji — Radicals and the Logic of On and Kun Readings

• Introduction
• How kanji are assembled
• Common radicals and their meanings
• On'yomi and kun'yomi — why one character has two readings
• Practical rules for telling on from kun
• Okurigana — hiragana after the kanji
• Intransitive and transitive — same kanji, different okurigana
• Compounds — on readings joined together
• Number kanji — on and kun at a glance
• Sino-Korean readings as a lever
• Practice catching meaning categories by radical
• An efficient order for studying kanji
• The six principles of kanji formation
• Same phonetic, similar on'yomi — more examples
• Radical names by position
• Types of on'yomi — go-on, kan-on, to-on
• Groups of kanji sharing a radical
• When radicals alone are not enough
• Practical application — analyzing an unseen kanji
• A concrete example of learning by word
• Shinjitai — shape differences from Korean and Chinese kanji
• Stroke order still matters
• Frequently asked questions
• Conclusion
• References

Introduction

Once you clear kana, the next mountain appears: kanji. The set of "regular-use" kanji needed for daily life in Japan is fixed at 2,136 characters, and reading newspapers or work documents without stumbling requires roughly that many. The number alone feels hopeless, but kanji are not random pictures — they are characters assembled by a consistent logic. Understand that logic and the amount you actually have to memorize shrinks considerably.

For native Korean speakers in particular, kanji can be a major advantage. A large portion of Korean vocabulary is Sino-Korean, and those readings often correspond regularly to Japanese on'yomi. Someone who knows the Korean words for school, library, or economy can, on meeting a Japanese word written with the same characters, guess both its meaning and its approximate reading.

This article covers:

1. The construction principle of kanji — radicals and sound components
2. Common radicals and their meanings
3. The difference between on'yomi and kun'yomi, and reading rules
4. Why a single character has multiple readings
5. Compounds and okurigana
6. Using the comparison between Sino-Korean and Japanese on'yomi as a learning lever

How kanji are assembled

Broadly speaking, many kanji pack two kinds of information into one character. One is "roughly what meaning category this belongs to," and the other is "roughly what sound it is read with." The part carrying meaning is the semantic radical; the part carrying sound is the phonetic component.

The vast majority of regular-use kanji are phono-semantic compounds built from "meaning + sound." Consider an example.

All four share the same phonetic component on the right (meaning "blue," read sei), so their on'yomi are mostly in the "sei" family. Only the left radical changes — water, sun, speech, heart — and that splits the meaning. Grasp this and even an unfamiliar kanji invites the hypothesis "left is meaning, right is sound."

Not every kanji divides this cleanly, of course. There are pictographs (mountain, river), indicatives (up, down), and compound-ideographs (rest = person + tree). But the big perspective — "kanji are combinations of parts" — helps with any character.

Common radicals and their meanings

A radical is not only the index key for finding a kanji in a dictionary; it is the core clue for guessing meaning. Learn just a handful of frequent radicals with their meanings and you can place an unknown kanji into a rough category.

For instance, if a new kanji has the water radical, you can guess "probably related to water or liquid"; with the heart radical, "probably related to emotion or mind." That narrowing is the practical payoff of learning radicals.

On'yomi and kun'yomi — why one character has two readings

The most confusing point in learning Japanese kanji is that a single character can have several readings. To understand it, you need the history of how Japan adopted kanji.

Japan originally had its own spoken language (native Yamato words) but no script. So it imported kanji from China, and in doing so two reading modes arose at once.

Take the character for "mountain."

Same character, but when it stands alone to mean "mountain," it is read with the native word yama; when combined with other kanji into a compound, it is read with the Chinese-derived on'yomi san. That is the basic frame for the split.

As another example, take the character for "life/birth," famous for its many readings.

A single character with many readings is an essential feature of Japanese kanji. It looks daunting at first, but there is regularity in "which reading appears in which context," and with familiarity the distinctions sort themselves out.

Practical rules for telling on from kun

Not perfect, but the following rules of thumb are very useful.

1. When a kanji stands alone (or with okurigana), it is usually kun. e.g., water (mizu), to eat (taberu).
2. When two or more kanji form a compound, they are usually on. e.g., water supply (suidou), meal (shokuji).
3. Names of people and places have many exceptions. Proper nouns often have special readings outside the rules above.

Compare the same character for "water" under the two rules.

The rules are tendencies and have exceptions. That is why kanji are best learned not as isolated characters but as words, memorizing their readings together.

Okurigana — hiragana after the kanji

When you write a verb or adjective with a kun'yomi, the hiragana that follow the kanji are called okurigana. Okurigana are an important device for distinguishing inflected forms and readings of the same kanji.

Even with the same character for "up," adding garu makes it "rise," while adding geru makes it "raise" — a different reading and meaning. The okurigana determine a word form and reading that the kanji part alone cannot reveal. So when memorizing verb and adjective kanji, always learn the okurigana as part of the bundle.

Intransitive and transitive — same kanji, different okurigana

The most important case where okurigana splits meaning is the intransitive-transitive pair. Japanese has many pairs that share the same kanji and differ only in okurigana.

Intransitive verbs express "it becomes so by itself," transitive verbs "someone makes it so." The difference between "the door opens" (intransitive) and "to open the door" (transitive) lies precisely in the okurigana aku versus akeru. The kanji part is the same, but the following hiragana determines both the verb's nature and its reading.

It is efficient to memorize such pairs together. They often correspond to intransitive-transitive pairs in your own language, which makes them relatively intuitive. The okurigana forms follow consistent patterns, so grouping by pattern helps memory.

Confusing intransitive and transitive when writing or speaking sounds off, so it is good to learn both forms together from the moment you memorize the kanji. Remembering that they share the same kanji lets you store the two words as a pair rather than separately, which is efficient.

Compounds — on readings joined together

Two or more kanji joined into a single word are called compounds. As the rule above suggests, compounds are mostly read on + on.

The strength of compounds is that, knowing each character's on'yomi, you can read and guess the meaning of combinations you have never seen. Knowing "electricity" (den) and "speech" (wa) lets you assemble "telephone" (denwa); knowing "self" (ji), "move" (dou), and "vehicle" (sha) lets you assemble "automobile" (jidousha). Seeing kanji as "parts" of words accelerates vocabulary learning.

That said, some compounds have special readings (jukujikun) that cannot be split character by character — today (kyou), adult (otona), skillful (jouzu) — and these must be memorized whole.

Number kanji — on and kun at a glance

Nothing shows the difference between on'yomi and kun'yomi more intuitively than number kanji. Japanese numbers split between counting with on'yomi and with native words (kun'yomi), and the sound can also shift depending on the counter that follows.

Interestingly, 4, 7, and 9 have two on'yomi each. Four is shi or yon, seven is shichi or nana, and nine is kyuu or ku, used by situation. In particular, shi sounds like the word for "death" and is felt unlucky, so daily usage tends to prefer yon. Readings splitting for such social and cultural reasons is also a feature of Japanese numbers.

When a counter follows a number, the sound shifts again. One plus the counter for long objects becomes a fused form, and three plus the same counter takes another sound change. These changes have complex rules, so at first it is efficient to memorize the frequent combinations whole, like words.

Sino-Korean readings as a lever

For Korean speakers, the most powerful weapon is the Sino-Korean readings they already know. Korean and Japanese on'yomi both descend from ancient Chinese, so they frequently show regular correspondences.

Here are examples of the tendency between Sino-Korean and Japanese on'yomi.

Patterns emerge. For instance, a final k in Sino-Korean tends to surface as ku or ki in Japanese on'yomi, and a final n or m tends to end in the syllabic n. In "study" (hak/gaku) and "nation" (guk/koku), the final k maps to ku; in "citizen" (min/min) and "three" (sam/san), the final n and m map to the syllabic n.

This correspondence is not an absolute rule and has many exceptions, but as a starting point for guessing an unseen on'yomi it is extremely useful. A learner who knows Sino-Korean readings effectively begins with about half of Japanese on'yomi already in their head.

There is a trap to note, though. Sometimes the same kanji compound carries a different meaning in Korean and Japanese. For example, the compound written 工夫 means "study" in Korean usage but means "ingenuity, contrivance" (kufuu) in Japanese. Do not assume that identical kanji shapes guarantee identical meanings.

Practice catching meaning categories by radical

To feel the power of radicals, apply them yourself. The exercise: look at characters sharing a radical and guess "what is the common meaning of these characters."

The characters for snow, cloud, electricity/lightning, and thunder all carrying the rain radical reveals at a glance that they are weather phenomena occurring in the sky. Grouping meaning categories with the radical as a clue greatly reduces the burden of memorizing each character separately. Make a habit, whenever you meet a new character, of recalling "what is its radical, and what did other characters with that radical mean."

An efficient order for studying kanji

Finally, an order for turning these principles into actual study.

The key is not to view kanji as "2,136 isolated pictures." Catch meaning with radicals, guess sound with phonetic components, lever Sino-Korean readings, and learn readings inside words — and the real burden becomes far smaller than the number suggests.

The six principles of kanji formation

The traditional system classifying how kanji were created is called the six categories (rikusho). Not every kanji fits cleanly into one category, but it is useful for the big picture.

What learners meet most often is the phono-semantic compound. Because the vast majority of regular-use kanji are of this type, the hypothesis "left is meaning, right is sound" lands so frequently. Pictographs and compound ideographs are fewer but common among basic kanji, so their shape and meaning are easy to connect early on.

Compound ideographs are especially fun to learn. A person leaning on a tree means "rest," two trees make "woods," and three make "dense forest." Weaving the parts of a character into a little story makes it stick.

Same phonetic, similar on'yomi — more examples

We saw a group sharing the sei phonetic. This principle applies broadly to other phonetics. Knowing a phonetic lets you guess the on'yomi of an unseen phono-semantic compound.

As the last "master" group shows, the phonetic's sound can shift slightly per character (shu to chuu, juu). The phonetic is only a starting point for guessing, not an absolute rule. Still, it is a powerful clue for narrowing on'yomi candidates when you meet an unknown character.

Radical names by position

A radical is named differently depending on where it sits inside a character. The same radical can change shape when its position changes, so knowing the position concept helps when looking words up or analyzing characters.

For example, the radical for "heart" becomes the side form on the left of a character (as in "nature") but stays as the full form at the bottom (as in "think"). They are two appearances of the same radical.

Types of on'yomi — go-on, kan-on, to-on

One reason a single kanji has multiple on'yomi is that Japan adopted Chinese pronunciations from different regions and eras several times over its history. These are divided into go-on, kan-on, and to-on.

This is exactly why the character for "go/walk" reads as different on'yomi depending on the word. Learners need not memorize this taxonomy, but knowing that "there is a historical reason a character has several on'yomi" reduces vague resistance. In practice, just memorize which on'yomi a given word uses along with that word.

Groups of kanji sharing a radical

Earlier we saw characters that share a phonetic (clear, clear-sky, request, feeling). Now let us group characters that share a semantic radical. A shared radical means a shared broad meaning category, so memorizing several characters around one radical is efficient.

Grouping this way applies the effort of learning one radical to several characters at once. When you meet a new kanji, you can guess its meaning by recalling "other characters with this radical."

When radicals alone are not enough

Radicals are not omnipotent. Knowing these limits helps you study accurately.

1. When the radical is detached from the meaning. Some characters, through historical change, have a blurry link between radical and current meaning. The radical is only a clue, not a verdict.
2. When the phonetic does not pin down the sound. The phonetic is also only a tendency. The same phonetic may diverge in on'yomi, or the sound may have drifted over time.
3. When decomposition is meaningless, as with jukujikun. Readings that cannot be split character by character, like "today" and "adult," must be memorized whole.

So use radicals and phonetics as "tools to cut the load and aid guessing," but make a habit of final confirmation with a dictionary and real words.

Practical application — analyzing an unseen kanji

Pulling the principles together, here is the thought process for meeting an unknown kanji, shown by example. Suppose you see the character for "hold" for the first time.

Once this flow becomes second nature, an unknown kanji no longer leaves you stuck. Catch the meaning direction with the radical, the sound candidate with the phonetic, and confirm finally with a dictionary or word. At first you walk the steps consciously, but with familiarity an almost unconscious sense of "decomposing to read" a kanji develops.

A concrete example of learning by word

Now that the principle is clear, let us demonstrate learning a kanji by word, using one character. Branch out from the character for "life."

Memorizing seven or eight words around one character naturally teaches all its main on'yomi (sei, shou) and kun'yomi (i-, u-, nama). The real-world utility is incomparably higher than memorizing the isolated "life" alone. Each time you meet a kanji, asking "how many words containing it can I learn" makes vocabulary spread quickly from stem to branch.

Shinjitai — shape differences from Korean and Chinese kanji

A surprisingly frequent snag for Korean speakers learning kanji is the difference in character shape. Kanji used in Korea are generally close to the traditional forms, while Japan adopted simplified forms (shinjitai) for some kanji in the mid-twentieth century.

As the table shows, even the same character is reduced in strokes or simplified in parts in Japanese shinjitai. Meaning and on'yomi are largely the same, so the Sino-Korean lever still holds. Only the "shape" must be relearned. Fortunately shinjitai are generally simpler, so the burden is small. Just map "the familiar 國 is written shortened as 国."

For reference, mainland China's simplified characters are yet another system, simplified even more boldly than Japanese shinjitai. The same character can split into three shapes across Korea (close to traditional), Japan (shinjitai), and China (simplified), so if you are learning Japanese, learn the shapes by the Japanese shinjitai standard.

Stroke order still matters

Even as occasions to write kanji by hand decline, knowing stroke order remains valuable — for the same reasons it mattered for kana.

The basic principles of kanji stroke order match those of kana and Korean-used kanji: top to bottom, left to right, horizontal first (usually), and enclose the outside first then fill the inside. For example, the character for "ten" is written horizontal first, vertical second. Even complex kanji come out naturally when you apply these principles to each part.

If you are only taking the JLPT, you need not dig deep into stroke order, but if you want a deep grasp of kanji structure or will write by hand, it is worth learning at least the basic principles.

Frequently asked questions

Here are questions learners often ask at the kanji-introduction stage.

1. Must I memorize every on'yomi and kun'yomi of every kanji? No. Memorizing the frequent readings as words is enough. You need not memorize every reading listed in a dictionary.
2. Should I study radicals separately? Just learning the meanings of 30–50 common radicals helps a lot. You need not memorize all radicals.
3. If I know Sino-Korean readings, do I get on'yomi for free? About half true. Regular correspondences favor guessing, but exceptions and false friends mean you still need to confirm.
4. Must I memorize shinjitai separately? If you are learning Japanese, learn by the Japanese shinjitai standard. They are generally simpler than Korean-used kanji.

Conclusion

Kanji is the longest and most persistent mountain in Japanese, yet also the most rewarding. Understand the logic of radicals and phonetics and exploit the strength of Sino-Korean readings, and kanji becomes a game of assembly and inference rather than random memorization.

In the next article, we will connect this kanji knowledge to real-world use and JLPT success: kanji counts per level, spaced repetition, sorting out confusable characters, and using the right tools.

References

• JLPT official site
• Agency for Cultural Affairs — regular-use kanji table
• Tofugu — Learn Kanji
• WaniKani — Radicals and Kanji
• Jisho — kanji dictionary
• Wikipedia — Kanji
• Wikipedia — Japanese readings of kanji
• Wikipedia — Radical (Chinese characters)