Leetspeak is the idea of replacing letters with characters that resemble them. For instance replacing "o"s with "0"s: n00b. This sort of text can also seem like computerized speech, i.e.: if an AI was talking.

There are some ideas for doing this in Chinese. Usually represented by cutting characters in half: "作" becoming "亻乍", "彩" becoming "采彡" or "爆" becomes "火暴," etc.

Are there any other ways to accomplish leetspeech or something that resembles computerized, AI-type of 乱码-styled responses in Chinese?

  • There is a writing style called 火星文. 我 >> 硪, 爱 >> 嗳, 你 >> 祢
    – Krahmal
    Feb 8, 2023 at 6:57

1 Answer 1


Leetspeak, as defined in your question, consists in respelling the words of a language using all of the scripts at its disposal. English has only two: an alphabetic one (a-z) and a numeric one (1-9).

Mainland Chinese has three (hanzi, numbers, Pinyin/Roman letters). Taiwanese Chinese has four (all of the above plus zhuyin/bopomofo symbols).

A leetspeak spelling system for Chinese would presumably involve spelling out each word, or most words, using glyphs from all three (or four) scripts.

My suggestion would be to use both semantic and phonetic principles in this substitution game, and to make phonetic substitutions fuzzy. That would mean using:

  1. Numbers to replace every instance of numerical concepts such as 一, 两, 俩, etc;
  2. Numbers to respell any occurrence of the syllables yi, er, san, si, wu, liu, qi, ba, jiu and shi - whatever their tone, whatever their medials (9 for jiu, ju etc) and/or whatever their manner of articulation (8 for ba, pa, ma & fa);
  3. Normal pinyin (and/or zhuyin/bopomofo) spellings to break up a run of characters or numbers;
  4. Fuzzy pinyin/zhuyin (eg ze/ㄗㄜ for zhe) and/or abbreviated pinyin/zhuyin (eg z/ㄗ for zi) to spell easy-to-guess function words and empty syllables;
  5. Simple characters & character components to graphically re-spell abbreviated pinyin syllables (丫 for y, 亅for j, 乙 for z, etc);
  6. Original characters to replace their zhuyin/bopomofo evolutions (卩 for ㄗ, 刀 for ㄉ, 日 for ㄖ, etc).

For example:

5来ju1个li乙, 这g9ㄗ10用L337拼xie8写刀

Standing in for


To make it even more 乱码-looking, but also far more incomprehensible, you could apply the above rules to character components, rather than characters, and use another romanisation system (such as Wade-Giles or, better yet, Gwoyeu Romatzyh) in addition to or instead of Pinyin.

But mixing up only two or three of the above-mentioned ideas should be enough.

  • I like the zhuyin idea. Adding in some "█" or "�" characters might spice it up nicely too.
    – Mou某
    Feb 5, 2023 at 23:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.