Multisyllabic Hanzi?

Students of Japanese are quite used to characters (漢字) nearly always having multiple pronunciations, ranging from one syllable to five or more. (Example: in Japanese, depending on the context, the character 侍 can be pronounced as or as さむらい.)

That’s one of the areas in which switching from studying Japanese to studying Chinese came as a relief: in Chinese you can be sure each hanzi (Chinese character) has a monosyllabic reading, and 90% of characters have only one reading.

In my studies, I recently discovered that this has not always been the case. My Chinese textbook gives me three examples that were around until 1977, when a character reform had them eradicated.

– 瓩 qiānwǎ (kilowatt); now standardized as 千瓦
– 浬 hǎilǐ (nautical mile); now standardized as 海里
– 呎 yīngchǐ (foot); now standardized as 英尺

Besides their very existence, I found several things about these characters interesting. First, they’re all for units of measure. Maybe at one point people liked the idea of a single character for each unit of measure? Second, it was an interesting evolutionary turn the language was taking. From a student’s perspective, I’m not sure I like it, but it’s interesting. You can clearly see which part in each character represents which syllable. Lastly, it was the government that quashed this fairly recent orthographic innovation in favor of standardization.

Note: You won’t find this info in Wenlin. I got it from 现代汉语 (上海教育出版社, 2004).

2011 Update: The venerable scholar Victor Mair writes about this subject on Language Log: Polysyllabic characters in Chinese writing.


John Pasden

John is a Shanghai-based linguist and entrepreneur, founder of AllSet Learning.


  1. I am somewhat sceptical of your analysis in this case. I would suggest that the second character in these three sets was always understood; in other words, the first character was not two syllables, but one with the other syllable being the unwritten, understood character now identified as standard. (If I am not mistaken, one could find instances of both conditions being written at that time). It is possible, given a use long enough, that the understood part would have been forgotten, and the characters actually assume two syllable pronounciation, but that did not happen. Also, one could argue that this was not a change occurring in the language, but just in the orthography.

  2. There is a series of these for the Imperial measures: 唡: 英两 ounce 哩: 英里 mile 吋: 英寸 inch [口畝]: 英亩 acre

    I’d guess that these originated as written shorthand for ledgers and accounts so that all the columns line up, so to speak. Google search shows them alive and well on Taiwan.

    They remind me of 廿, which people tend to casually read as 二十 (though it has its own pronunciation). Or to a lesser extent with 囍 or that 招财进宝 character.

    JFS: one difficulty with your interpretation is that the characters themselves are innovations: 瓩 clearly has a 瓦 and a 千 combined, so to imagine an original 千瓩 or 瓩瓦 would be redundant.

  3. JFS,

    Well, you can be skeptical, but the examples are not my analysis. They come from an authoratative textbook written by Chinese experts on the Chinese language. The book represents research rather than pure speculation.

    Obviously, my thoughts at the bottom are not so authoratative. When I said, “it was an interesting evolutionary turn the language was taking,” I meant, of course, the written language. I also didn’t mean to imply that I thought these characters would have been the start of a sweeping change across the entire face of the written language. They simply would have been a significant deviation from a well established pattern.

  4. zhwj,

    Thanks for the extra examples. Once again, on a topic that probably just bores most other readers, you provide interesting additional insight…. I appreciate it.

  5. Very interesting. We native speaker would never notice that.

    I’ve learn Japanese for a year. I think they use characters (漢字) just for short. I mean they can use 平假名 instead, but that would be too long. So that’s why their 漢字 have multisyllabic.

  6. The 1977 directive from the standards bureau is available online here; there’s a badly scanned chart at the end that lists a whole bunch of non-standard units (the document itself is interesting – conversational, in a way, and quite lenient on the gradual phasing out of existing packaging material. And with only one reference to building a socialist economy, it’s from a completely different world from today’s policy statements).

    Two others widely used characters were multi-syllabic: 浔 (海寻) for nautical fathom, and a similar construction for non-nautical fathom (I’ve always thought the two were the same length).

    John: No problem. Linguistics / orthography is fascinating stuff – I briefly considered it before deciding on literature, but couldn’t see how I’d go about absorbing all that information for an entrance exam. Much admiration from this quarter for your choice of fields.

  7. John, authoritative research can be authoritative, but it is never so authoratative as to necessarilly squelch scepticism.

    The question arises, was this not the use of jargon by a specialized in-group referencing specialized terms, or was it truly the rise of poly-syllabic pronounciation of terms. If it was poly-syllabic characters, then it use would be rather universal; but it is jargon, then there would be instances of both elements of the compounds be used in written literature, probably by those outside of the in-group.

    Concerning scepticism, scepticism is just that, the act of being sceptical; and that is one of the hallmarks of good modern science. There is too much poor research promoting bad hypothesis, especially in the humanities and in the social sciences.

    In modern times much research is conducted in using “Deweyism”, something akin to letting the facts speak for themselves. Facts never speak for themselves, and too often certain assumptions are made that are not valid nor able to withstand scrutiny.

    Concerning this research conclusion, it may be true; I do not know, but I am still sceptical.

  8. JFS,

    The value of skepticism as a practice was never in doubt. I am merely skeptical of the value/basis of this particular application.

    Thanks for the lesson, though.

  9. 现代汉语小词典 (The Concise Dictionary of Modern Chinese, I guess) lists these characters, and for most it says that it can also be read as just the second syllable, ie “mile” instead of “nautical mile”, “inch” instead of “imperial inch”, etc. (瓩 is the exception, for obvious reasons).

  10. Oh, and here’s a question: the three-dot water radical is an obvious one to use for the nautical units, but why a mouth radical for imperial (or “english” as the chinese prefix means) units? It’s not as if there aren’t already enough characters with that radical!

  11. ZHWJ:

    The characters themselves are innovative, but that does not make the combination necessarilly redundant. Taking for example, 英尺, no matter how it is representated, it has represented the English inch. Consequently, the innovative character would have been used in their place, with the pronounciation held intact, then the chi (尺) would have been understood. In other words, it is a shorthand for two other symbols. There is nothing wrong with that, from a theoretical standpoint a symbol can be an entire phrase, a sentence, a paragraph, or an entire book; but the more information it carries, the more cumbersome it becomes. What I am suggesting is that the measure is subsumed in the combination, but is understood by the writers themselves. If sufficient time takes place and the original combination is not associated, then the characters can assume a true poly syllabic pronunciation; unless another evolutionary force takes place. The other force would be conformance to existing practice, a powerful force. In these cases the measure pronounciation, I would suspect, be dropped and in our example only the Ying would be pronounced. That latter part is speculative, it has not been long enough to to see how the evolutionary process would work. What has it been now, maybe around 100 years, three generations of use. No where near enough time to disassoicate the measure character from the innovative character, at least that is why I am sceptical.

  12. Todd,

    If you study all (or most) mouth radicaled characters, they fall in these categories. 1. Action verbs by the mouth (吃-eat, 吹-blow, 呕-vomit, 听-listen-odd but hey listen, 哄-cheat, 哑-mute); 2. Phonetic words (哇-wa1, wow?, 咦-yi2, 呱-gua1, 哈-ha1, ha4); hence 3. Phonetically transliterated foreign words (吨-dun1=ton, 听-ting3=tin/can, 咖喱-ga1-li2=curry, 喹啉-kui2-lin2=quinoline); so why not the fourth category – these non-phonetically translated foreign measuring units, after all, “ton” and “tin” in the third category are foreign measuring units. This indeed poses a problem, though: the fourth category does not fit. Hey, maybe that’s just it, the English units were weird words that did not fit.

  13. Kikko Man Says: May 6, 2005 at 10:20 am

    Would these be the Chinese equivelant of contractions? At least in the written sense.

  14. Gin, a similar thought crossed my mind too. “What radical should we use for these words? Hey, what did I just say…words…we can use a mouth radical to show that they are words!”

    Kikko Man, English contractions reflect a phenomenon in spoken english — two words are fused together into a shorter word in speech. If anything, these multisyllabic hanzi would be more like abbreviations.

    Does chinese have anything similar to English contractions? I’m not sure how the word 别 developed…is it a contraction of 不要? And 甭 for 不用. And some people pronounce “这样” so that it almost sounds like one syllable!

  15. I’ve studied Japanese on my own for a few years now, and I’ve oftened considered comparing the difficulties of learning Chinese and Japanese. Compared to Chinese grammar, Japanese grammar seems harder with its different levels of politeness and many verb forms. The many pronunciations of a single kanji would be a barrier, and the Chinese hanzi would be a lot easier to learn. On the other hand, Chinese uses a lot more characters and is undeniably harder to pronounce. Since Mandarin is my first language, I’d have to ask for the opinions of native English speakers: which is harder to learn from the English-speaking perspective, Chinese or Japanese?
    A reply would be really appreciated!

  16. JFS: I’m not sure that a drop from two syllables back to one would be possible; to avoid ambiguity, units must be distinct, both on the page and in speech.

    Call them shorthand or jargon, fine. But neither half of the spoken pronunciation can be dropped:

    • In the phonetic borrowings, the characters themselves heavily suggest a two-syllable pronunciation: 加侖 “gallon” becomes 嗧, with the same pronunciation “jiālún”, and the sense-phonetic 千瓦 becomes 瓩 (similarly, 瓼 “centiwatt”, 瓱 “milliwatt”). Any single syllable would be ambiguous.
    • Units that were similar, but not identical, to traditional Chinese market units were given a prefix (英、海), represented in the innovative characters by a distinct component added on to the traditional character (口、氵). When pronounced, the unit must remain, because the prefix is shared among all of the Imperial units or Nautical units; and the prefix must remain, because the unit alone would be confused with the traditional market version.

    The 1977 directive makes just this point in arguing against single-character units, but it comes down on the side of preserving the one-character->one-syllable tradition of written Chinese, whereas the inventors of the characters followed a one-character->one-unit standard.

    What might end up happening would be something like the contractions Todd mentions (别 and 甭), or the ancient Chinese fusions: 诸旃耳盍与邪 (although these written forms probably followed spoken patterns rather than leading them).

  17. Todd – yes; 别 was borrowed to represent a contraction of 不要, and likewise 甭. In the North, and particularly the Northeast, 啥 is used as a contraction of 什么. I’m sure others can think of some more examples.

    俗字 are fun.

  18. ZHWJ: I understand what you are saying, but let me rephrase my scepticism slightly different. Obviously the modified character is polysyllabic, but it is explainable as the combination of two characters, one a modifier and the other a measure component. The modifier was modified and the measure component is understood (unwritten). Most associations are forgotten over time, that is when one is unable to explain the issue, and says instead, “that is just the way it is done”. You have identified possible evolutionary trends, but I suspect that the unwritten measure component will be written back into the phrase. It is has been said that is redundant, which is true; but Chinese is no stranger to redundancy, and as a matter of fact uses redundancy very effectively to reduce ambiquity. And that brings me back to John’s original statement, about the poly-syllabic characters and the speculative evolutionary trend that it may have lead to; that is what I am sceptical about.

  19. YUU: Of course, everyone comes to a foreign language with a different background and capabilities, I will add my views since you requested such information. I did not find the grammar all that difficult, but many of my frineds did find the use of particles rather disconcerting. (I do not wish to imply any intellectual ability on my part, I came to Japanese having studied classical Latin and Greek previously, and so I had already encountered that hump of different grammars). For my part, I think it is a wash between Chinese and Japanese as to which is easier grammatically speaking for an English speaker. Sounds are quite different, though. Japanese has a much smaller selection of sounds, but they are also much easier, in my opinion, for an English speaker to learn. Not only that, but the sound combinations are much easier than Chinese (I have heard Chinese phrases where several sibilants appeared in a row, rather difficult for me). But there is far more information available on how to pronounce Chinese correctly than is available for Japanese. The variance in the spoken language is a lot less for Japanese than it is for Chinese. So it is much easier to understand regional variances (at least in the larger towns) for Japanese than it is for Chinese (Suzhou Putonghua and Changsha Putonghua and Chongqing Putonghua can take some time to acclimate oneself to for an English speaker, at least myself). The written language is rather complex for Japanese, in the since that each kanji may have multiple readings (although Chinese also has multiple readings, it does not appear as complex as Japanese). All in all, I do not think one is more difficult than the other, they each have different hurdles that one must overcome.

  20. Brendan: Is 啥 a contraction? I would have thought it was merely a variant pronunciation of 甚(什), There’s 甚的 as a variant of 什么 in Yuan and Ming drama and fiction (just like 啥的 in some dialects today), and you also find things like 甚时 for 什么时候 (okay, so that one I just pulled out of 《古汉语常用字字典》).

    Taking you up on your suggestion to provide more examples, my personal favorite is 羟, qiǎng, for 氢氧 qīngyǎng, hydroxide. Contraction by design – sort of a reverse 反切.

  21. victor Says: May 6, 2005 at 9:10 pm

    RE: Your Japanese example, ji and samurai. IMHO, there are two basic ways to pronounce a Chinese character in Japanese 音読み: おんよみ on’yomi, called the Chinese or pseudo-Chinese pronunciation 訓読み: くんよみ kun’yomi, called the Japanese or native pronunciation Therefore, ji is on’yomi and samurai is kun’yomi. Again, within kun’yomi there are several ways to pronounce the same Kanji.


  22. IMHO, 啥 and 甭 can be better thought of as regional variant (dialectical) expressions than as contractions, though no big difference here. Zhwj’s examples of 甚 and 甚时 are actually still in use in Shanxi dialects. These are easily taken as contractions because the pronuciations are similar to an abridgement of the longer phrases (as in the “这样”example cited by Todd) but my consideration that these are dialects is based on (my own) observation that the same regions which use 啥 for 什么 would likely also use 娃 for 孩子 which obviously is not a contraction, and other regions which use 啥子 would also use 娃子 or 娃儿. This shows a pattern sonsistent with dialect phrases rather than contractions. An example of non-contraction is the word 孬 (nao1) for 坏/不好/调皮, used in some northern dialects.

    The scientific words such as 羟 were “invented” rationally out of needs to translate foreign invented terms, rather than “grown” variations of the Chinese language.

    There is a somewhat unique example of an unofficial contraction or multisyllabic word invention. In handwriting, many use a shorthand for “问题” where first they simplified 题 by replacing the 页 radical with a roman letter (really!!!) “T”, then further on they made 问 a radical by dragging a tail out of it and add the same “T” above that tail. Yet another variation is simply a 门 radical with “T” inside. These new single words represent 问题 and, as far as I know, are still pronounced wen-ti.

  23. 甚事 in a Shanxi dialect becomes 嘛事儿 in a Tianjin dialect which becomes 啥事儿 in a Dongbei or Shandong or Shaanxi dialect.

  24. Yuu,

    Having studied Japanese for 4 years and Chinese for 6+, I definitely have an opinion on this.

    Japanese is hard for its grammar. The word order, conjugations, and levels of politeness are a lot to get a handle on at first. The pronunciation, however, is rather easy.

    Mandarin Chinese is hard for its pronunciation. Tones in particular are very hard to master, and some of the consonants take some time to master as well. The grammar, however, is rather easy.

    Both languages are very hard to read — Japanese because each character can have so many possible readings, and Chinese because it’s ALL characters (no “kana breaks” for relief).

    Japanese is easier to communicate as a beginner because the sounds of the language are relatively easy and “foreigner intonation” will not cause communication breakdown. That’s different from Chinese, where a beginner’s bad tones can totally kill communication.

    On the other hand, I think it’s harder to sound like a native speaker in Japanese, because the finer points of natural intonation can be extremely subtle. In China, there are so many natives with funny-sounding Mandarin that a foreigner’s very slight accent stands out less.

  25. Brendan,

    I was wondering when you’d make your appearance in this discussion… 🙂

  26. Gin,

    不好 = 孬?? Haha, that is awesome! Character of the week!

    My book mentions 巯 (qiu2, “organic chemical base SO”), from 氢 (qing1, H) and 硫 (liu2, S).

    I see in addition to 甭 there’s also 甮 (ning4/feng4). Maybe that’s a 吴语 variation? My book also mentions a character for a 吴语 word meaning 不曾 which combines 勿 and 曾 (read fen1), but I wasn’t able to type it out or find it and copy and paste it. I did find 朆, which is read fen1, but the character printed in my book has the 勿 component on the left side, not the right. Weird.

  27. Not really weird, as quite a few characters have alternate forms — for example there is a alternate form of 够 in which the left and right sides are reversed in this way. My dictionary lists 勿+曾 as the headword, with 曾+勿 as a variant. The former is the form standardised by 《现代汉语通用字表》, so the fact that you could only find the latter in the character set is perhaps a bit strange.

    PS. I think of 咋 and 啥 tending to co-occur, in any case they both sound so 俗 that you have to love them! Gin: did you mean 娃娃, rather than 娃?

  28. Todd,

    What was weird was that the one printed in the book was not the one I could find in the extended character set. That the one I was able to find is not part of the 现代汉语通用字表 makes it even weirder.

  29. Todd,

    I don’t mean 娃娃 which is so common. In places where the dialects are so 俗 or “土” that 咋 and 啥 are used, you often can hear 娃 exclusively, for 孩子,儿女,or 小孩. Example: “你那娃咋没出去挣个啥钱呢?” Take a listen, the usage might exist where you are.

    In Shaanxi, and maybe Gansu too, “别” is not represented by 甭 but instead by a character consisting of a 不 on top of a 要, pronounced bao.

    People don’t notice this any longer, but 歪 is of the same construct and probably was developed the same way.

  30. Kikko Man Says: May 9, 2005 at 9:39 am

    Isn’t the very essence of Chinese this sort of contraction pattern weve been discussing? Aren’t all Chinese words, other than the basic radicals just, recombinations of two or more other characters for meaning + meaning or meaning + sound?

  31. John: Good point about all the different accents, I think that’s one bonus in learning Chinese.

    Yuu: I haven’t studied Japanese beyond the basic I’ve-had-and-exchange-student level, but my best friend learnt Japanese and lived in Japan for a long time, I’ve learnt Chinese for about 3-4 years. Personally I think the spoken aspect of Chinese is much easier, in terms of the subtlety of Japanese tones. I found Chinese tones relatively easy to learn, and the grammar was a breeze compared to a) my friend learning Japanese and b) my experiences with other languages. Pinyin is an awesome system, because it means you can write down what you hear for future reference.

    However, I think there are bonuses to the Japanese written language because of some phonetic aspects. I really struggle with written Chinese, and comparing with my friend’s Japanese learning, it seems much easier for an English speaker to learn the Japanese style… yet of course, we have different experiences in language learning so its not exactly a good comparison I suppose.

  32. Kikko Man,

    Not really, in the following sense. The majority (95~99.9%?) of Chinese characters were “carved in stone” by the first emperor of Qin dynasty 2200 years ago. Sure most characters are constructed by contracting a radical and another sound or meaning marker, but it is not like the Chinese are doing this kind of word building (in the written form) all the time. The examples cited above are thus rarities, some of them anyway (in contrast to word building in English, which in my view is easier and more frequently done, e.g., feminaziism, bioterrorist, positron, cybervandal, political anticorrectnessism). On the other hand, phrase building (regrouping characters to express a new-fashioned or imported meaning) do occur in Chinese all the time, though.

  33. Kikko Man Says: May 10, 2005 at 8:31 am

    The fact that Qin Shi Huangdi “carved them in stone” does not negate the fact that the characters he just whittled down and codified were originaly created from this contraction-esque process, right? Qin Shi Huang just got rid of the old unused ones and or the ones he just didn’t like. The fact is the forms of many or most words had been around along time before him.

    Other question? What is the English equivilent of Chinese shortening of phrases by just refering to say the 1st or 3rd character or the 2nd and 4th? For example guoji maoyi becomes guomao or when shouji youpiao becomes jiyou. Are these also contractions? Peace!

  34. Kikko Man, you are right. I shouldn’t have said “not really.” All I wanted to get across is that the character building has largely stopped with the few exceptions in this thread. Students of Chinese should not expect character contractions or abbreviations actively. Phrase regrouping can be expected actively, such as your “guomao” example.

    I can think of some examples of what you refer to, similar to “guomao.” In English, you’d have nanotech from nanometer technology, dompol from domestic policy, IntAdm for International-student Admission, and HoJo for Howard Johnson — but I don’t know what this type of abbreviation is called, either.

  35. Kikko Man,

    I always thought of those shortened Chinese forms as akin to English’s acronyms (like F.B.I.). I guess thinking of it as a form of contraction makes some sense too, though. I thought it was more similar to acronyms because the characters chosen usually represent the essential meaningful parts of the whole. Contractions are often just for ease of pronunciation, regardless of how well the original parts are represented in the new form. (Example: ‘rents for “parents” pays no heed to the possible ambiguity introduced since “rent” is already a word.)

    I’m sure there are Chinese terms for this linguistic phenomenon, but I haven’t studied it yet. (I probably will this week, actually!)

    If Chinese worked like English…

    Acronym: 北京大学 -> 北.大.
    Contraction: 北京大学 -> 北’大’


  36. Gin: I wonder if maybe character building hasn’t really stopped. One can’t add a new variant to the GB or Unicode set as simply as was possible when publishing meant carving a set of wood blocks. There are still people out there who, when their son is born deficient in fire, tack a fire radical on to one of the characters in his name. It just can’t be entered into the computer, is all.

    And in the centuries since the Emperor standardized the system (and particularly since the Han when they assumed their present form) there have been literally thousands of new characters created (鲁迅 in 《门外文谈》 says “自然,后来还该有不断的增补…新字夹在熟字中…直到现在,中国还在生出新字来”).

    What is it with Emperors leaving their marks on the writing system, anyway – apart from banning characters, there was Wu Zetian’s attempt to write herself into history, so to speak, leaving us with such beauties as 瞾(曌)、埊、圀.

  37. this let me remeber an image i saw long time ago, which was created by westerner, used the Chinese characters insted of English words for their brand name since they had trade with Qing dynasty.
    The sentence reads from right to left like most ancient ppl did, so the “BOVET CO.” was written in 口威 播( 2 characters, sorry i cant find the character 口 with 威).

Leave a Reply