A new kanji for tapioca
« previous post | next post »
Remixed meme:
The concocted kanji is 姏 (Unicode U+59CF) with the punctuation marks /。superimposed upon it.
I haven't met anyone who knows the underlying character, so it might as well have been made up by the memer.
It is composed of nǚ 女 ("female"), Kangxi radical #38, as presumed semantophore on the left and gān 甘 ("sweet"), Kangxi radical #99, as presumed phonophore on the right, but it is not pronounced "gān" or anything like that. Rather, according to Hànyǔ dà zìdiǎn 漢語大字典 (Unabridged dictionary of Sinoglyphs), 2.1035a, the rare character 姏 is pronounced mán. Phonologically, that does not seem possible with the 甘 phonophore.
Morphologically, the underlying character is equally mystifying. The devisers of the character probably meant for it to mean "sweet woman", but, according to the Hànyǔ dà zìdiǎn 漢語大字典 (Unabridged dictionary of Sinoglyphs), loc. cit., it means "old woman".
The meme format is explained here and here.
Can you "just invent a new Kanji"? Yes — although governments frown upon it for private persons.
Phonological notes for specialists
South Coblin:
The graph is late, dating from Six Dynasties times. One has to be skeptical about its structure and the way it was constructed when it is that late.
At least one Qing writer has attempted to explain it: 《通俗編》- sub 婦女 – “姏婆，婦之老者；能以言悅人，故字从甘，其音讀若鉗.”
So he doesn’t accept the Guangyun pronunciation. This same explanation is also included in the 《字彙》(1615). So the Qing guy must have copied his entry from here.
John Carlyle:
The character has a pronunciation of /mam/ in the Qieyun System, so 甘 /kam/ was likely intended to be phonophoric. I don't think the initial is resolvable, though. Syllables with the shape MVM are marginal to begin with in Chinese, so it was probably only used for the rhyme.
Etymologically, I suspect this word is related to 姥 /muX/ 'matron'. Likely a fusion of it with something else that ended in -m, but I am not sure what.
Chris Button:
It looks like it is U+59C. Hopefully it displays here: 姏
As for the pronunciation, I suspect that the 甘 series was uvular rather than velar and that this character goes back to a voiced uvular ʁ- onset that then rounded to m- in the labial environment conditioned by the bilabial coda. For the phonology, compare how 惟, which also had a ʁ- onset in Old Chinese, gives an m- reflex in Fuzhou Min. Incidentally, Pulleyblank had ʁ- for 卯 in his 1983 ganzhi formulation, which I suspect was to account for the simultaneous evidence for an m- onset and a rhotic r- onset that clashes with evidence in the Tai forms that preclude the reconstruction of a cluster such as mr-.
By the way, the Morohashi entry (at least in the abridged “kou kanwa” instead of “dai kanwa” version that I have here) also includes a qián reflex apparently coming from the Zihui, but the Guangyun and Jiyun both support the mán reflex.
Axel Schuessler:
There are many graphs that look like phonetic compounds but are not: 好, which incidentally is also a phonetic compound in the clan name zǐ. So the famous Fu Hao 婦好 should really be read Fu Zi.
Selected readings
- "A new variant of a common Chinese character" (10/14/19)
- "Hong Kong protesters messing with the characters, part 2" (9/1/19) — with dozens of relevant Language Log posts listed in the "Readings" there; new characters invented during the Hong Kong protests are just one type of hanzi / kanji / hanja proliferation; we have recorded dozens of others on Language Log over the years
- bubble / boba tea
[Thanks to Nathan Hopson and Ben Zimmer]
Kevin Yeung said,
April 22, 2023 @ 10:46 pm
To my mind, the made-up character represents a woman drinking a cup of bubble tea with a tapioca boba at the bottom using a straw.
Jonathan Smith said,
April 22, 2023 @ 11:27 pm
Not for 'tapioca' per se it would seem but rather for 'bubble tea' ~ '(tapioca) milk tea' or watchamcalit– thus the "(drinking) straw" and "tapioca ball ~ boba" radicals /。
Wikipedia informs that there is now a Japanese verb tapiru タピる 'drink bubble tea', so I assume the kanji in question has at least the reading タピ alongside タピオカティー :D
Also to give credit where credit is due, "boba" is a Taiwan thing first FWIW
Victor Mair said,
April 23, 2023 @ 5:19 am
@Kevin Yeung and Jonathan Smith:
Bingo boba!
Chris Button said,
April 23, 2023 @ 10:38 am
The point about bilabial onset and coda dissimilation is a good one. So we either have a non-standard coinage or we have the m- onset coming from something else.
The proposal that 婦好 should be read as Fu Zi seems to be mostly accepted now. However, I'm not sure if we should be so quick to dismiss 好 as not having a phonetic component in its "hao" reading, especially if the Shuowen analysis of 从女子 should really be 从女子聲. It is possible that 好 "hao" originally had 㐬 (inverted 子 with dots of parturition fluid that were often removed when abbreviated) as phonetic (note 毓) but then became conflated with 好 "zi".
ZwriRat0 said,
April 23, 2023 @ 8:34 pm
Kanji is for Japanese, hanzi is for Chinese, you bring up Chinese pronunciations, so which one is it?
Josh R. said,
April 23, 2023 @ 8:42 pm
I'm with Kevin Yeung in that the use of the "woman" radical is meant as a reference to the popularity of tapioca drinks among women in Japan.
JOHN S ROHSENOW said,
April 24, 2023 @ 3:19 am
"Kanji is for Japanese, hanzi is for Chinese; you bring up Chinese pronunciations, so which one is it?"
This is another discussion, but I have long noticed that (even?) among American
sinologists trained in the 1960s and shortly thereafter, they/we often use the term "kanji"
when talking/speaking about Chinese "Hanzi". For some reason except for food words from Cantonese which came in with the Chinese immigration to the USA in the 19th C,
(e.g. chow mien, chop suey, bak choi, etc.) other Chinese terms are not as familiar to educated American English speakers as many Japanese terms are: e.g, "Kabuki, Noh, shamisan, koto", etc., but "Peking (sic) opera", not 'Jingju'; "moon guitar", not 'yuechin"; "Chinese zither", not "guzheng", etc. –Whether this is due to the earlier influence of Ernest Fenallosa, or the US occupation of Japan after WW II I am not sure.
Peter Grubtal said,
April 24, 2023 @ 9:13 am
@JOHN S ROHSENOW
For mother-tongue English speakers there's no problem vocalising "kanji", whereas "hanzi": I haven't a clue what's correct. I suspect even sinologists, subconsciously or no, go for the version which makes sense in English orthography.
Pinyin is for the Chinese (if their government decrees it) and sinologists. For the rest of us, they shouldn't have tampered with the English language. And that includes Beijing(sic), which I can't find in my atlases.
wanda said,
April 24, 2023 @ 10:27 am
@ZwriRat0, JOHN S ROHSENOW:
The invented character seems to have been invented in Japan, so kanji is appropriate. Then Prof. Mair opines on the origin of the Chinese character that the Japanese kanji was derived from. At that point, he or his correspondents call it a character, graph, or Sinoglyph, all of which are appropriate. They do not call 姏 a kanji.
Victor Mair said,
April 25, 2023 @ 5:20 am
@Wanda:
You put it precisely!
Scythd said,
April 27, 2023 @ 10:42 pm
Germans: Why not? We have "einlogen", "choken", "skypen", etc.
Victor Mair said,
April 27, 2023 @ 11:22 pm
@Scythd
Those may be invented words, but they are not wholly new invented elements of the writing system.
Swadloon said,
April 28, 2023 @ 5:22 am
I'm kinda surprised. I visited this article because I got curiously attracted by the word "tapioca", a native meal in Brazil (inherited from indigenous people) that's totally different from what's described here, since it's not even a drink. Both seem to be made from manioc, what makes their origin the same, though.
Chas Belov said,
April 29, 2023 @ 1:26 am
I've heard, but don't know for a fact, that Boba is Taiwanese slang for women's breasts, so it would be logical that the radical was a woman.
I tried futzing around with combining characters in Unicode, to no success. There is a ring symbol as a combining character but it only renders above (U+030A) or below (U+0325) the hanzi. I also couldn't get a slash (U+0338) to render as the combining character. I was able to get a small slash (U+0337) to render, but it renders in the center of the whole character, not in the center of the phonophore. (By the way, phonophore is a new word to me.)
Chris Button said,
April 29, 2023 @ 12:27 pm
The point about bilabial onset and coda dissimilation is a good one. So we either have a non-standard coinage or we have the m- onset coming from something else.
Further to my comment here (https://languagelog.ldc.upenn.edu/nll/?p=58574#comment-1605022), it looks like we have ʁ- in 姏 shifting to m- before the labial -m coda in a similar way to q- shifting to p- 灋/法 before the labial -p coda. Li Fang-kuei notably makes the tentative suggestion that kw- was the source of p-in 灋/法.
Chris Button said,
April 29, 2023 @ 1:57 pm
Rather than positing some kind of dialectal variation, I wonder if 劫 was prevented from evolving like 法 (their OC reconstructions would have been the same) because of its association with 劫波 "kalpa"
Chris Button said,
April 29, 2023 @ 9:24 pm
… which would support a medial -r- that could then block the rounding
astrange said,
May 1, 2023 @ 6:45 am
> I'm kinda surprised. I visited this article because I got curiously attracted by the word "tapioca", a native meal in Brazil (inherited from indigenous people) that's totally different from what's described here, since it's not even a drink.
Bubble tea is made out of tea + condensed milk + quite a lot of sugar syrup + "bubbles" which are boiled tapioca spheres (originally made from sago, but tapioca/cassava/manioc was cheaper.)
In Japan, "tapioca" means this drink and nothing else, afaik. I don't believe any other countries call it that.
Also, Japan stereotypically has very high quality food but for the longest time all the native brands of this stuff were cheap and nasty. It's only recently actual HK/Taiwan tea shops have opened with good recipes for it.