Chinese characters vs. Japanese: what is actually different

If you only take one thing from this page: look for the round, loose shapes. Japanese sentences are written with characters and kana together, and kana are visibly lighter on the page than the characters around them. Chinese has nothing of the kind. Every symbol in a Chinese sentence is a full character filling a full square.

There are two kana sets and both are worth recognizing. Hiragana is rounded and continuous: あ の は に ん. Katakana is angular and clipped, used mostly for foreign words: コ ー ヒ ー, which is kōhī, coffee. If either set appears anywhere in a passage, the passage is Japanese. No further checking is needed.

The test runs in only one direction, and this is where people get burned. Text with no kana is probably Chinese, but short Japanese labels are frequently pure characters. 非常口 is an emergency exit sign in every building in Japan and contains no kana at all. A Chinese reader parses those three characters as extremely plus mouth, which is not an exit.

The best-known single giveaway is の, the possessive particle. It is so recognizable that Chinese advertising sometimes borrows it purely as decoration, to make a product look Japanese.

Korean is the easiest of the three to rule out, because modern Korean is written in an alphabet that looks like nothing else on earth. Hangul was designed from scratch in the 1440s under King Sejong, and its letters are built from a small set of geometric strokes: circles, short bars, and right angles, stacked into square blocks. 사랑 is sarang, love. If you see circles and clean straight lines assembled into syllable blocks, you are looking at Korean, and no Chinese dictionary will help you.

The complication is historical. Korea wrote in Chinese characters for roughly fifteen centuries before hangul, and kept using them alongside it well into the twentieth. Those characters are called hanja, and they are still taught, still printed in some newspapers and legal texts, and still used to disambiguate names. So a Korean document can contain Chinese characters. What it will almost never do is contain only Chinese characters, the way a Chinese one does.

Which leaves the question underneath all of this. A great many people search for an "Asian symbol" for a word, and there is no such thing to find. Asia writes in dozens of scripts, most of them unrelated to Chinese: hangul, Devanagari across northern India, Thai, Arabic through the west of the continent. The phrase persists because one script did travel, borrowed by Japan, Korea, and Vietnam over more than a thousand years, and in English it came to stand in for a whole continent. If someone hands you an "Asian symbol" for love or strength or courage, what they are handing you is a Chinese character, and it is worth knowing which one before it goes on your skin.

The part of Japanese that looks least Chinese was made out of Chinese characters. Before either kana set existed, Japanese was written by taking characters purely for their sound and ignoring their meaning, a practice called man'yōgana.

Hiragana is what happened when scribes wrote those sound-characters fast, in cursive, from around 800 CE onward. The hiragana あ is a cursive 安, written until it flowed. Katakana came out of ninth century Buddhist monasteries in Nara, where monks annotating Chinese scriptures needed marks small enough to fit between the lines, so they clipped a fragment out of each character and used that. カ is the left half of 加. The name katakana says so: fragmentary kana.

So the honest picture is not two writing systems meeting. It is one writing system that reached Japan, was used to write a completely unrelated language, and grew two new organs to do it.

Almost every English page on this subject treats the question as two systems against each other. There are three. Taiwan, Hong Kong, and Macau kept the traditional forms and simplified nothing. Mainland China and Singapore use simplified forms. Japan uses shinjitai, its own postwar forms, which are neither.

Where all three diverge you get characters like these. The reading columns show what the same shape is called on each side of the sea.

A real name makes it concrete. The director of Seven Samurai signs his surname 黒沢 in Japan. A Taiwanese poster prints it 黑澤. A mainland Chinese poster prints it 黑泽. Three spellings, one man, both characters different every time.

Several characters people actually choose for tattoos sit in this three-way split. War is 戰 traditional, 战 in mainland China, 戦 in Japan. Happiness runs 樂, 乐, 楽. Listening runs 聽, 听, 聴. And the dragon is 龍 traditional and 龙 simplified, while Japan's standard form is 竜, with the old 龍 surviving mainly in personal names.

Japan went first. The Ministry of Education published a list of 1,850 characters for official use on 16 November 1946, and the table that fixed their new simplified shapes followed in 1949. Mainland China published its Chinese Character Simplification Scheme in 1956, with the fuller general list arriving in 1964 and a final revision in 1986. Two independent committees, a decade apart, working on the same inherited script.

Neither committee invented much. Both went shopping in the same warehouse: centuries of handwritten shorthand, the abbreviated forms that clerks, printers, and letter writers had used informally for as long as anyone had been writing by hand, plus the shapes that cursive brushwork had already worn smooth. The Japanese term for the prewar stock is ryakuji, abbreviated characters. The Chinese equivalent is 俗字, popular characters, and the tone of both words is the same: this is what people actually wrote when nobody official was watching.

That single fact explains both the agreements and the disagreements. When only one good shorthand existed for a character, both committees found it and the two reforms match:

And when a character had two or three competing shorthands in circulation, the two committees reached for different ones, which is how 澤 became 泽 in Beijing and 沢 in Tokyo. Nobody was copying anybody, and nobody was reacting against anybody. They were both picking from the same shelf.

The other half of the problem is not shape but sense. A compound that crossed to Japan in the seventh century has had thirteen centuries to drift, and some of it drifted a long way. These are the ones that catch people:

None of these are mistakes by either language. They are two vocabularies that started from the same stock and were used by different people for a very long time. But they are the reason a phrase pulled off a Japanese site and used as Chinese can be grammatically fine and socially catastrophic.

The standard story stops in the fifth century with characters traveling east. It left out the return trip, which is more recent and larger than most readers expect.

When Meiji Japan set about translating European science, law, and philosophy, its scholars built the new vocabulary the obvious way: out of Chinese characters. They coined thousands of compounds. After 1895 those compounds flowed back into Chinese along with the students who went to Tokyo to study, and they are now so ordinary that almost no Chinese speaker knows they are imports. Telephone 电话, philosophy 哲学, bank 银行, democracy 民主, republic, society, economy: all assembled in Japan.

Japan even invented characters outright, a category called kokuji, and at least one of them made the crossing. 腺, the character for a gland, was built in Japan for Western anatomy and is now standard in Chinese medical writing. A Chinese doctor uses a Japanese character every working day without noticing.

If a character is going on your skin, a sign, or a logo, three checks settle it and none of them require you to read anything.

Every character entry on this site records which forms exist and flags the ones with a distinct Japanese shape. You can also see a character rendered in classical scripts before committing to a shape, and browse the full set of character entries with the sources each meaning rests on.

Sharing a script is not sharing a language. Japanese and Chinese are unrelated: the grammar, the sound systems, and the native vocabulary have nothing to do with each other, and the relationship between them is closer to English and Vietnamese both using the Latin alphabet than to English and German. A Chinese reader who can guess the topic of a Japanese headline is doing the same thing an English reader does with a Spanish menu, and the confidence it produces is not reliable.

The three-form split is also a minority case. Of the roughly two thousand characters in everyday Japanese use, most are written exactly as they are in traditional Chinese, and a large share are written exactly as they are in simplified Chinese too. The divergences matter out of proportion to their number because they cluster in common characters, which is precisely why they turn up in tattoos and logos.

Korean and Vietnamese belong in this story and are not in it. Both wrote with Chinese characters for centuries. Vietnam moved to a Latin alphabet, and South Korea now uses characters rarely, mostly for disambiguation. That is a separate page.