Noncompositional

Though it is currently trendy for Japanese writers to use foreign words in their prose, this was not always the case. In fact, during the Meiji period, Japanese scholars faced the problem of translating a large amount of foreign (Western) literature into Japanese. In order to accomplish this task, they coined many hundreds of new words from Chinese morphemes. In some cases they took previously-existing compounds and assigned them new meanings. Such is the case for 経済 ‘economy’ and 社会 ’society.’ The former is an abbreviation of 経世済民 (or 経世済俗), which means ‘govern the world and save the people.’ The latter, 社会, referred to a religious ceremony held in the spring and autumn during which sacrifices were made to god(s) of the Earth (this is according to this site, assuming I made correct sense of the Chinese). Examples of what I believe are totally new terms include 主義 ‘ideology,’ literally ‘main meaning/significance’ and 抽象 ‘abstraction,’ literally ‘extract image.’)

In their modern meanings, these compounds were born in Japan and then imported by the rest of the sinosphere, including Korea, China, and (occasionally) Vietnam. When I first learned that so many intermediate-level Sino-Japanese words were actually invented in Japan, I was a bit surprised. How many, exactly? I had thought maybe a few dozen, or a hundred tops. Turns out there are quite a few more, at least according to this list on Chinese Wikipedia. You’ve got words like subjective, library, constitution, system, background, necessity, cash, time, space, and absolute. If the list is correct, these were all coined in Japan. Seriously: words that I would have bet were native to Chinese, like 必要, 時間, 空間, and 現金 are there.

I begin with a story.

Once in China there lived a man who loved to raise monkeys. He loved his monkeys so much that his family left him and he was left with his pets. Thankfully for him, he could speak with these monkeys, and so knew exactly what they wanted; and he always provided. But one day, he realized that his funds had been depleted so much that he could no longer afford to feed them their favorite food (nuts) as much as they wanted. So he told them one morning, “from now on, I will only give each of you three nuts each morning, and four each evening.” The monkeys were not pleased, and pleaded with him to reconsider. After thinking a while, he presented a new plan: “all right, I will give you four each morning, but then only three in the evening.” The monkeys were very pleased with this plan, and did not complain further.

And so today in Chinese, the phrase 朝三暮四 (zhao1 san1 mu4 si4, ‘morning three dusk four’) indicates someone who constantly changes their opinion, or who is unreliable or irresponsible.

Read more »

Until today, I was very confused about the pronunciation of this character: 癌. It is used in Chinese and Japanese to mean ‘(malignant) cancer,’ or ‘carcinoma’ (though, like English cancer, it technically could also be applied to the wider category of either benign or malignant tumors). However, in Japanese it is pronounced gan, and in Mandarin it is pronounced ai. Now, there are certainly cases where Japanese has undergone a lot of sound changes since borrowing a Chinese word, and Mandarin is probably one of the least representative of the sort of Chinese language that the Japanese borrowed from. But at least there is a semi-regular correspondence. gan and ai are not only very different, but they don’t represent any correspondence between Japanese and Mandarin that I’d ever seen before.

Then, today, in Chinese class, I learned that the character 癌 actually used to have another pronunciation in Mandarin: yan. Oh ho! This made perfect sense, as yan/gan is a perfect Mandarin/Japanese pair. But another obvious question presented itself: why the heck did it change?

Apparently, in order to avoid confusion in spoken language between 癌症 ‘cancer(ous symptoms)’ and 炎症 ‘inflammation,’ which were homophonous, ‘cancer’ was changed to ai. This was effective only for Putonghua; other Chinese languages seem not to have altered the character’s reading, likely because the two characters 癌 and 炎 are distinct in those languages. For instance, in Cantonese, 炎 is jim and 癌 is ngaam. And the final question: why ai? Well, I’m not entirely sure, but it seems that in some Chinese languages, 癌 is pronounced ngai. Perhaps whichever organization decided to change the character’s reading took their cue from those languages, and just chopped off the initial velar nasal, which Mandarin did who knows how many centuries ago anyway.

There are many pitfalls that await a learner of Chinese, if that learner also speaks Japanese or Korean. The reverse is also true. One of those pitfalls has to do with two-character compounds. There are several dozen two-character words in Chinese XY which are present in Japanese (or Korean) as YX. Off the top of my head:

I used to have access to a book that contained a page’s worth of such compounds, but at the moment I don’t even remember what it was called, or if a copy of it is housed in my local library. I’m also not exactly sure why these reversals took place. As far as I can tell, Korean and Japanese share the same order, and some comments I found on various forums seemed to indicate that Vietnamese has the same order as Chinese, at least for some words (like ‘peace’). There are also no cases I’m aware of where the pronunciation doesn’t match the orthography, e.g., writing the compound XY but pronouncing it YX. That would be interesting, though.

That reminds me of a review I recently read of the CGEL in Language (80:1), by Peter Cullicover:

Accepting for the moment that what CGEL says about some phenomenon is all that there is to say, it is somewhat puzzling that as scientists we would have a serious notion of what would be more interesting than the truth. For instance, it would definitely be more interesting to discover that the moon is made almost entirely of green cheese than that it is made of rock and dust, especially given that it looks like it is made of rocks and dust, and the samples that have been brought back are—rocks and dust! It would be more interesting to learn that pigs cannot fly because their wings are made of an invisible substance that is too insubstantial to support their weight, rather than that they simply lack the anatomical and physiological wherewithal in the first place. It would be far more interesting to discover that chimps appear to lack human language because their religious beliefs prohibit the expression of personal thoughts (as opposed to feelings) to other creatures, rather than whatever the true answer is, which is probably some mundane story about neural organization, computational capacity, conceptual structure, and the like. But granting that the less interesting explanations are the right ones, scientists do not give up the good fight and turn to other pursuits. Why should linguists?

I was disappointed that the possibility of data fabrication was not brought up. Surely it would help increase young people’s interest in science if the scientific world engaged on a crusade to find pigs’ invisible wings.

In a previous post I discussed some activities of Japan’s national language research institute, which starting in 2003 started publishing lists of loan words in Japanese, most of them recent loans, that they felt should not be used in government and other popular documents (e.g., newspapers) because they contributed to a lack of comprehension on the part of the general public. Some typical examples of words taken wholesale from English into Japanese are informed consent, community, universal service, and soft landing (as in a slow economic quieting-down).

Over at LanguageLog, Mark Liberman wondered how borrowing a foreign word could lead to evasive language, as I had said. Indeed, many of the concepts that the institute singled out originated in various sectors of the American and/or European business and government world, and so no equivalent is immediately available in Japanese. It thus may seem natural to just take the foreign concept and stick it into your language unomi…uh, I mean…sans accommodation. Though this may be bad technique, surely it is not “evasive.” Or is it?

I admit (and this is hard, being a mini-lexicographer) that I did not mean to imply that such users of gairai-go were evasive or using ii-mawashi. In the case of government documents or in newspapers, the Kokken asserts that the use of gairai-go is due to writers putting priority their own ease of writing rather than their audience’s ease of understanding (外来語の使用状況を見ると，読み手の分かりやすさに対する配慮よりも，書き手の使いやすさを優先しているように見えます。). However, from the point of view of the reader the effect, I believe, may be a sort of evasiveness or equivocation [insert some word that means: 'You keep using that word. I do not think you know what it means']. Because many of these words are used without supplementary definitions in Japanese ([this would also alleviate the problem, I suppose]), those people without much exposure to English (and often, specifically business and government terminology) will be left thinking that the wool might have been pulled over their eyes. After all, if you’re not sure what the difference between サーベイランス surveillance and モニタリング monitoring is, you might be a bit confused if you receive a notice that local law enforcement will be stepping one of them up in your area (the latter is about making sure you don’t miss any changes in a situation, while the former is about making sure you don’t let anything evil/bad escape your attention). So while these two words, if understood by everyone in the conversation, could lend a good amount of precision to the discourse, for someone not familiar with the subtleties, it may as well be doublespeak.

Of course in the end, as with most things like the Kokken’s suggestions, it’s basically a moot point, since the Kokken really just makes suggestions to other government agencies (and newspapers, though I don’t think they have to follow the guidelines). The places where you see the most obvious use of katakana-go is in advertisements, television, and other popular media. In those places it has a sort of chic, or so I’m told. So as long as learning (some sort of) English is liu xing, uh, en vogue, there’s no expecting the use of gairai-go to decrease. Thankfully, the American government is well aware that English is superior to other languages, and so we need not fear our kooteki shorui being contaminated.

[edited to fix some markup and minor content]

The recent news that the national language-monitoring agency of Iran has put forth a long list of loan words that are to be replaced by native words or (sometimes newly-coined?) compounds. Posts from Language Log, Gwynn Dujardin, and Language Hat cover the issue, mostly amused at the attempt of a central regulatory agency to control the uncontrollable.

Back when I was in Japan in 2003-4, the National Institute for Japanese Langauge (国立国語研究所, or 国研 Kokken) released its second list of suggested rewordings of gairai-go (外来語言い換え提案), or loan words mostly from western languages. However, unlike the efforts of some other national bodies, Kokken (or rather, the Gairaigo committee) does not wish to purge the Japanese language of evil foreign influences (yet! mwa ha ha), but instead encourage understanding and discourage evasive language. They point out that often the use of gairaigo is more about increasing ease for the writer or speaker (who can just import a foreign concept without explaining it), as opposed to increasing understanding for the reader or listener. Their suggestions are also, well, suggestions, rather than written-in-stone law. Their documents in particular single out government-issued documents, newspapers, and other texts with a high level of exposure to the public. They don’t really care what people use in their own homes, but if people don’t understand what their government is saying, maybe something isn’t going right.

Read more »

On a recent trip to Chinatown in San Francisco, I noticed that although most American banks have transparent names in Chinese (e.g., Bank of America is 美洲銀行 ‘America(n continent) Bank’, WaMu is 華盛頓互惠銀行 ‘Washington Reciprocate Bank’). However, Citibank is called 花旗銀行, literally ‘flower flag bank.’ Could it be an attempt to approximate the sound of citi? In Mandarin it would be hua1 qi2, and in Cantonese faa1 kei4. Okay, so it’s not the sound. Maybe some previous name of Citibank had something to do with flags or flowers, or had a name that sounded more like the Chinese characters? Well, no. So what is the answer?

Read more »

The street name Broadway is often rendered 百老匯 in Chinese. In particular, Broadway in NYC is so-named, and street signs in Oakland near the Chinatown area have signs with this name (as I recall). I like this name, for several reasons. First, the first two characters (bai3 lao3 in Mandarin) are reminiscent of the word 老百姓 (lao3 bai3 xing4), which means ‘commoners,’ (though it literally means ‘old hundred names’); this fits with (the sorts of streets that are named) Broadway as a sort of common place for normal people to gather or travel on. The third character, hui4, has a meaning of ‘converge’, ‘flow,’ or ‘gather together’. Again, all broadly fitting with Broadway.

Of course, in order to be a good rendition the pronunciation has to fit reasonably well. Now, [bai laʊ xueɪ] is really only acceptable if you’re aware of the limitations of Mandarin phonology, as well as the fact that most famous place names (at least in America) were not given Chinese names by speakers of Mandarin, but by speakers of other Chinese languages that preserved useful things like syllable-final oral stops. I suspected that it was a speaker or group of speakers of a southern Chinese language like Cantonese that first used words like 百老匯, but the Chinese Wikipedia article says that it was likely a speaker of Shanghainese or Ningpo that first coined the term, along with the word for Washington, 華盛頓 (Mandarin hua2 sheng4 dun4 [xwa ʂɣŋ duən]). According to the indices on a Wu Dictionary website, Broadway is pronounced (sans tone) [paʔ lɔ uɛ], and Washington is [uo zã təŋ]. For the curious, the standard Cantonese would be [baak lou (w)ui] and [wa siŋ dœn] (hopefully I didn’t mangle that; I just started trying to learn Cantonese).

It definitely seems that (modern) Cantonese sounds much closer than either Mandarin or Shanghainese for Washington, but the coda k makes its Broadway a bit anomalous. The more source-language-loyal conventions currently used(*) in Mandarin for foreign names might get something closer, like bu luo de wei, but it would create a rather long name, which (IMO) looks and sounds rather ungraceful, and doesn’t have the punch of something like 百老匯.

(*) I actually have only a very limited grasp of how sounds (in English) are usually written out with characters in Mandarin, and have only intuited that there is some sort of standard that may in truth only be limited to official discourse.