A romantic and persistent idea holds that Chinese characters are semasiographic, that they convey meaning directly, as pictures or symbols, without representing any particular spoken language. It is a lovely idea, and mostly wrong. Getting the linguistics right matters, because it changes how you should learn characters.

What semasiographic means

A semasiographic system conveys meaning without representing the sounds of a language: think of a no-smoking symbol, which means the same regardless of what language you speak. True semasiography is rare for full writing systems, because conveying complex thought without grammar and words is extraordinarily limited.

The myth is that Chinese works this way, that each character is a little picture of an idea, readable independent of any spoken language. It is an appealing story, and it does not hold up.

Chinese characters are logographic

Chinese writing is logographic: each character represents a specific spoken word or morpheme, a unit of the Chinese language, with both a sound and a meaning. 馬 does not mean “horse-ness in the abstract”; it writes the Chinese word mǎ, meaning horse. The character is bound to the spoken word, not floating free of language.

Crucially, the vast majority of characters are phono-semantic compounds, built from a component that hints at meaning and one that hints at sound, see which part of a character hints at its sound. A sound component is the opposite of meaning-without-language: it ties the character to pronunciation. So most characters are explicitly linguistic, not pictographic.

The small genuinely picture-like set

To be fair, a minority of characters come close to conveying meaning visually:

  • Pictographs: 日 (sun), 月 (moon), 山 (mountain), 人 (person), 木 (tree), drawn from the things they depict.
  • Simple ideographs: 一, 二, 三 (one, two, three), 上, 下 (up, down).

These are evocative and worth knowing, and they are where the picture-language story comes from. But they are a small fraction. Build your understanding of characters on them, and you will be lost the moment you meet the phono-semantic majority.

Why this matters for learning

If characters were pictures, you would learn them by recognising images. Because they are sound-and-meaning units built from components, you learn them best as exactly that: meaningful parts, a sound clue, a written form to produce. Trying to treat every character as a picture fails; treating each as a linguistic unit with phonetic and semantic structure works, the approach behind which part of a character holds its meaning and etymology breakdown plus writing.

And as always, understanding the structure is not the same as being able to write it; that still takes from-memory production, see the case for a dedicated Hanzi writing app.

Where Hanzi Write Practice fits

Hanzi Write Practice trains characters as the linguistic units they are: you produce each from memory on a grid, connecting its written form to its sound (pinyin) and meaning, with stroke order and spaced repetition. That matches how characters actually work, sound-and-meaning units built from components, rather than the picture-language myth.

Admire the pictographs. Just remember they are the exception, and learn the rest as what they are: writing tied to a spoken language, not pictures of ideas.

Join early access and learn characters as the language units they are.