Phonetic transliterations of foreign words are not uncommon in Mandarin, e.g. 沙发 for "sofa", 逻辑 for "logic", etc.

Much less common are what I would call "transliterated characters" or "phono-semantic transliterations", where a loanword is assigned a unique character matching it both in pronunciation and graphic form (the latter usually by means of an appropriate semantic component). In antiquity, characters for foreign plants and animals (e.g. 狮) may have been coined using this type of phonosemantic transliteration, as suggested in the comments.

Nowadays, transliterated characters are only said to occur in the context of naming chemical elements, which are assigned unique phonosemantic characters (形声字) made up of a phonetic component matching their English/Latin name and a semantic component matching their physical typology (gaseous, metallic, etc).

However, I only just noticed that the word for "pump", pronounced beng4 and written with the logographic character 泵 (a 会意字), also fits the description of a "transliterated character" - despite being neither a chemical element nor a 形声字.

I wonder if I have just found the only transliterated character outside of chemistry, if not perhaps the only transliterated loanword in all of Chinese that is written with its own 会意字 logogram. So my questions are:

Outside of chemistry, is there any other character that, just like 泵, is used to transliterate a foreign word both phonetically and semantically?

Even including all neo-characters from chemistry and beyond, is there any other transliterated loanword written with its own 会意字?

  • You can find 泵 in 康熙字典. Its pronunciation is liu, not beng. 康熙字典 gives no meaning, only pronunciation of this character. Obviously, it's a reincarnation of a dead character with a new meaning and pronunciation.
    – joehua
    Jan 8, 2023 at 0:49
  • Thanks - that's more evidence for the peculiar 会意字 nature of this transliteration. On the one hand, you're right to note that 泵 is one of those few 冷僻字 that were deliberately repurposed to write a new morpheme or to transliterate a new loanword. On the other hand, it's striking that finding the right 会意字 character for "pump" should have prevailed over considerations of pronunciation: all other cases (e.g. 啤) are characters chosen because their original phonetic component made for a good 形声字. The character chosen for 泵seems to be the only 会意字.
    – Sanchuan
    Jan 8, 2023 at 7:52
  • 1
    @joehua yes, 泵 is in Kangxi, but only in one of the appendices that deal with characters that have unknown meanings, or readings, or attested usages:【備考】【巳集】【水字部】 【五音篇海】音流。— the only location given is from another dictionary. That makes it rather unlikely that whoever created the modern usage of 泵 had the old character 泵 in mind. It's a coincidence. Jan 10, 2023 at 12:54
  • Thanks for that note, John. I actually just found a source tracing it back to the 清 with a reading of [pìn]:  ㄆ〡ㄣˋ  方言。謂水沖激磯石。  清鈕琇《觚賸‧語字之异》:“粵中語少正音,書多俗字……水之磯激為泵,音聘。” The mystery thickens!
    – Sanchuan
    Jan 10, 2023 at 22:04

2 Answers 2


I agree with your argument with 秀. The meaning of show can be seen as an extension of its original meanings, it's 会意字 and share the same sound. In my experience, before I was told its new meaning is transliterated from English, I thought it was native. The same experience goes for 卡.

And your comment with characters with 口. I can't persuade you further in that regard, but I stand corrected for its use in chemistry. In fact, 口 is used as a 形旁 in chemical names, like 嘌、呤、嘧、啶, not a marker for transliteration. They are 形声字, 口 indicates that it forms a 闭环, which is commonly seen in organic compounds.

The answer is no. (The following bold characters are also transliterated.)

If including characters for chemicals, there are some that are both 会意 and 形声, for example 氢、、氮、溴、碳. Apart from element names, many terms for chemical compounds are 会意字, e.g. 烷、烯、炔、羟、羰、羧、巯、膦、、锍. To understand how they 会意 one needs adequate knowledge for chemistry. I give one example here. Oxygen is historically called 養, hydrogen 輕. 羟 is a compound of oxygen and hydrogen, and the character put together part of each character and use that part as a semantic reference to the corresponding element.

are 会意字 and transliterated. For the latter two, we can probably find more in its like. To begin with, find the name of any transliterated element name. Then we check for the name of their compounds. The name of its compounds are likely to have the same stem in the source language. The Chinese counterparts are usually newly-coined or revitalized 会意字, with one part, like in the example 申 is used as a semantic and phonetic reference to 砷, while the other part indicates with which it's compounded, like 月 means it contains nitrogen. Since transliteration in chemistry usually only take one syllable, chances are that the coined 会意字 for the compound also transliterate the source word. (To distinguish the Chinese element name and its compound name, we alter the tone of the latter. However, in grade school, the correct tone is mostly not used and confusion is easily resolved in context.)

Among the above 烯、炔、硫 also take into account the pronunciation of the source word to some extent.

I'm not sure if your question limits time periods. If not, some Buddhist terms also qualify. For example, from Sanskrit kalpa. It's 会意字 and transliteration, and semantically an extension of its original meaning. Probably more in this category. Since the loans happened quite long ago and the sound of spoken Chinese has changed significantly since then, it could be hard to identify at first glance.

Related words include 圐圙 from Mongolian хүрээ. Though it's transliterated into two syllables, the two characters are 会意字 and have exactly the same meaning. The sound is assigned according to the source language.

In daily life, 咖啡 is transliteration, and the radical 口 indicates it's something to eat or drink. The same goes for 啤(酒). But 咖、啡、啤 are not newly-coined but revitalized for transliterated words.

Using 会意字 to transliterate foreign words is rare, mainly because, I think, is that foreign words are usually more than one variable and the morphemes of Chinese is predominantly monosyllable. 泵 is one of those rare cases where the source word is monosyllable. There are so many dead Chinese characters. To find a 会意字 that creates a semantic match and assign a new sound to it, just like 泵, is not difficult. Another monosyllable source word example is 卡 (会意字). In this case we didn't look for a 会意字 but instead found a similar sound. ka3 and qia3 differs in their 等, the palatalization of qia3 as a result. For people from some regions they are free variants. I would say it's not only phonetic but also somewhat semantic. Something is qia3-ed and creates a thin seam, which is where a card comes in. Or when you forget your keys, maybe you can qia3 you card into the door seam to swipe open the door.

Transliteration of some chemical elements solves this issue by only using the first syllable of the source words.

Being translated both phonetically and semantically at the word level is a very common way to translate a foreign word, e.g. 霓虹、普罗大众、盖世太保、苦力、风信子、图腾、甜不辣、休克、俱乐部、心地、引擎、媒体、霸凌. There are just so many. It creates a phonetic match and its component characters contribute to the meaning of the whole word as the semantic components contribute to the corresponding character.

  • Yes, finding a dead character for transliteration purposes shouldn't be too hard. And, yes, finding one that can act as a 会意字 is harder. But for it to be "rare" we need to find more than a single exception. So far, 泵 seems the only exception. I say this because 1. the 口字旁 in the characters you mentioned has never been described as a semantic component, but as a marker to indicate that the characters are being used for their phonetic value only (exactly the reverse of a 会意字); and 2. I struggle to see the semantic, pictographic or ideographic side of 卡 when used as a transliteration of "card".
    – Sanchuan
    Jan 8, 2023 at 8:21
  • @SanchuanSan For 1., what you said mainly applies to chemicals. The examples I gave are for daily usages. Several sets of characters have historically competed for acceptance (e.g., 嗑肥、考非、加非 for coffee, 皮酒 for beer), and finally they arrive at the current one. One very important reason for their success is the semantic component. The existence of the semantic relation fits the general rule of Chinese word component and thus more acceptable.
    – lilysirius
    Jan 8, 2023 at 8:57
  • 2. qia3 is to stuck/to get stuck in a narrow seam, which often gives a very thin shape. As a result it relates to the shape of a card. To qia3 (or for some regions qia1) A into B, either A is thin enough, as we inseart our card into the card slot, or we squeeze A into a flat shape, like how the Knight Bus managed to get through the buildings and the traffic.
    – lilysirius
    Jan 8, 2023 at 9:00
  • @Sanchuan It's rare in everyday life but not so in chemistry. See my update.
    – lilysirius
    Jan 8, 2023 at 9:31
  • I agree that the two-character transliterations you mentioned are indeed used as phono-semantic transliterations, just like 泵. However, none of them make up a single "transliterated character" when - crucially - they could have done (just like multisyllabic chemical elements were deliberately abbreviated into single mono-morphemic words and transliterated into single phono-semantic characters). This means that multisyllabic transliterations cannot answer the question "Is 泵 the only transliterated loanword with its own logogram in all of Chinese?".
    – Sanchuan
    Jan 8, 2023 at 10:19

a related example may be 砼, but even more complicated in terms of the etymology and phonology.

泵 is wholly made up for the word pump, defined to sound like it, but the composition of the character is 石+水 which purely express the meaning, in Chinese 会意字, without a part indicating what it sounds.

this is an essential difference compared to characters of chemical elements 氢氦锂铍硼... where all of them are 形声字, where one part indicates what it sounds, and the semantic part partly shows what kind of element it is. also, some of them existed in ancient time, some are newly made up for the extension of the periodic table. although, most of them sound like the name of the chemistry element.

砼 means concrete but doesn't sound like "concrete", instead it followed the right part 仝, which looks like 人工 "made by human" (but etymologically, not so), plus 石, gives "stone made by human". it could be regarded as a 会意字.

if not restricted to 会意字 (that's how 泵 differs from chemical element characters), many names of plants, birds, fish, insects 花鸟鱼虫 will quite likely be some loan words from their origin place (near ancient China), plus a semantic part. they were just created maybe many years earlier than those for chemical elements, but with the same method.

so, if what you find is a character that specifically satisfies both "sounds like the foreign word" and "is a 会意字 wholly based on meaning of it".....

i haven't find one yet. for now this is just intended to be a clarification of the question :)

edit: from @lilysirius's answer i find the word that qualifies both criteria: 圐圙, 囗 四方 八面 all components express the meaning, none of them is used for pronunciation.

秀 might be a coincidence that the extension is frictionless, though originally not "created for that and by that". but for the original question "is used to transliterate a foreign word both phonetically and semantically", it works.

but others mentioned don't quite agree with my definition... for example 卡 is 会意字 itself when it means "stuck", but the glyph doesn't express "card", instead only the pronunciation is borrowed. it's more like a 假借字 for "card". 氯 is both 会意 and 形声 but doesn't sound like chlorine. 氯氢氧 from 绿轻养 work quite similarly.

  • Thanks for this clarification. 砼 is a very interesting case, though you're right to note there's no evidence of any phonetic transliteration there. A better fit would be ancient characters originally created to name foreign animals and plants, though it's hard to prove if their semantic component was a transliteration at the time (and from which language). Thank you for pointing out that chemical elements are all 形声字, whereas 泵 is a 会意字 - that's a useful distinction to make here and I'll acknowledge as much in the question.
    – Sanchuan
    Jan 7, 2023 at 20:21
  • Concrete is called beton in some languages in Europe.
    – alephalpha
    Jan 16, 2023 at 12:02

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.