Of honey, bee, mead, and Old Sinitic reconstructions
« previous post | next post »
Pamela Kyle Crossley wonders:
Why, when mi– ma– words for “honey” are so widespread across Eurasia, do English speakers say “honey” instead of some modern form of medhu or meli (except when referring to mead, of course)? Turns out all the Germanic languages left the medhu theme early on, and instead went with variation of *hunaga, which they might originally have cut off from hunigcamb. It sort of suggests that these Germans first encountered honey as imported in combs or frames, not as if they were extracting it from the bees themselves.
Now this proto-Germanic *hunaga is reported by Eva Crane to be derived from an IE root *kenəkó, “which means golden-yellow.”
Does it remind you of any similar words that mean “golden yellow,” or at least “yellow”?
Second, I wonder this: Did the Germanics really give up on mi- except for “mead”? What of the word “bee” itself, which has a murky history (from Old Germanic bion). Is it in fact the more venerable mi-?
At first glance, a PIE root for Germanic "bee" does not seem to be terribly secure:
Old English beo "bee," from Proto-Germanic *bion (source also of Old Norse by, Old High German bia, Middle Dutch bie), from PIE root *bhei- "bee."
A closer look indicates that "bee" cognates might extend beyond Germanic languages, but not very far:
From Middle English bee, from Old English bēo, from Proto-Germanic *bijō (compare West Frisian and Dutch bij, Upper German Beie, Danish and Swedish bi), from Proto-Indo-European *bʰi- (compare Old Irish bech (“bee”), Welsh bydaf (“beehive”), Latin fūcus (“drone”), Latvian bite (“bee”), Russian пчела́ (pčelá, “bee”)).
Judging from this, "bee" belongs to the northwest sector of the Indo-European world. See J. P. Mallory and D. Q. Adams, Encyclopedia of Indo-European Culture (London and Chicago: Fitzroy Dearborn, 1997), p. 57b.
Thus a supposed PIE root for "bee" does not seem to be that assured, since all we have before the oldest Germanic form is Pokorny's conjectural bhei- 116
And what about "honey"?
Middle English hony, from Old English hunig "honey," from Proto-Germanic *hunang- (source also of Old Norse hunang, Swedish honung, Old Saxon honeg, Old Frisian hunig, Middle Dutch honich, Dutch honig, Old High German honang, German Honig "honey"), of uncertain origin. Perhaps from a PIE *k(e)neko- denoting yellow, golden, or brownish colors (compare Sanskrit kancan- "golden," Welsh canecon "gold," Greek knēkos "yellowish"), or perhaps from a substratum word. Finnish hunaja is a Germanic loan-word.
From Middle English hony, honi, from Old English huniġ, from Proto-Germanic *hunagą (compare West Frisian hunich, German Honig), from earlier *hunangą (compare Swedish honung), from pre-Germanic *kn̥h₂onkós, from Proto-Indo-European *kn̥h₂ónks (gen. *kn̥h₂kós) (compare Middle Welsh canecon (“gold”), Latin (pl.) canicae (“bran”), Tocharian B kronkśe (“bee”), Albanian qengjë (“beehive”), Ancient Greek κνῆκος (knêkos, “safflower”)), Kurdish şan (“beehive”).
This does remind me of Sinitic huáng 黃 ("yellow"), Old Sinitic (OS) *N-kʷˁaŋ (Baxter-Sagart) / *ɡʷaːŋ (Zhengzhang).
Now I need to look more deeply into the intriguing name Huángdì 黃帝 ("Yellow Thearch"), one of the Three Sovereigns and Five Emperors of Chinese myth.
Examples
mì 蜜 ("honey"), OS *mit (Schuessler; Baxter-Sagart), cf. Minnan (Hokkien) bi̍t — From Proto-Tocharian *ḿət(ə), from Proto-Indo-European *médʰu (“mead”). Compare Tocharian B mit (“honey”). [Wiktionary]
shānhú 珊瑚 ("coral") OS *slaːn ɡaː (Zhengzhang) — Borrowed from an Iranian language in Central Asia during the Han Dynasty, probably Khotanese sam̥gga- (“stone”). Compare Modern Persian سنگ (sang, “stone”). [Wiktionary]
Readings
- “Of precious swords and Old Sinitic reconstructions” (3/8/16)
- “Of precious swords and Old Sinitic reconstructions, part 2” (3/12/16)
- “Of precious swords and Old Sinitic reconstructions, part 3” (3/16/16)
- “Of precious swords and Old Sinitic reconstructions, part 4” (3/24/16)
- "Of precious swords and Old Sinitic reconstructions, part 5" (3/28/16)
- "Of armaments and Old Sinitic reconstructions, part 6" (12/23/17)
- "Of shumai and Old Sinitic reconstructions" (7/19/16)
- "Of felt hats, feathers, macaroni, and weasels" (3/13/16)
- "Of dogs and Old Sinitic reconstructions" (3/17/18)
- "Eurasian eureka" (9/12/16)
- "Old Sinitic reconstructions and Tibeto-Burman cognates" (4/18/16
- "Of ganders, geese, and Old Sinitic reconstructions" (10/29/18)
Paul M said,
November 1, 2018 @ 9:25 pm
I have never gotten over the fact that “bee” in mandarin is 蜜蜂 and honey is 蜂蜜. The two characters are simply reversed. The crudeness (or simplicity) of the arrangement would suggest that the words are not of ancient origin. Is it not true? Are there any other such mirrored characters in the language?
Ambarish Sridharanarayanan said,
November 1, 2018 @ 9:59 pm
Etymonline seems to be inaccurate in a few different ways:
* The Sanskrit word for golden they think they want is "kāñcana", not "kancan".
* The Sanskrit word they really want is "kanaka", which is clearly more related to k(e)neko than "kāñcana" is (Sanskrit etymological texts do not seem to relate the two words to one another).
Chris Button said,
November 1, 2018 @ 10:03 pm
While the character 黃 is clearly used as a color term in the oracle-bone inscriptions, it is commonly found in the compound 黃尹 referring to a "Minister Huang" who was clearly venerated.
Scott P. said,
November 1, 2018 @ 10:24 pm
While the character 黃 is clearly used as a color term in the oracle-bone inscriptions, it is commonly found in the compound 黃尹 referring to a "Minister Huang" who was clearly venerated.
I guess they didn't know that Minister Huang did it in the Library with the pipe wrench.
Andy said,
November 2, 2018 @ 3:05 am
The root *melit did manage to hang on in Germanic in the Gothic miliþ (honey) and a couple of words in Old English, most notably the compound 'mildeaw' (mildew), which is also attested in other West Germanic languages.
dainichi said,
November 2, 2018 @ 3:48 am
> “bee” in mandarin is 蜜蜂 and honey is 蜂蜜. The two characters are simply reversed. The crudeness (or simplicity) of the arrangement would suggest that the words are not of ancient origin. Is it not true?
Multiple morphemes per word is a relatively new thing in Mandarin, but that doesn't mean the morphemes aren't of ancient origin. Honeybee, bee-honey. Kinda works in English too.
~flow said,
November 2, 2018 @ 6:11 am
One might add that, incidentally, 密 denotes something secreted, whereas the homophone 祕 denotes a secret. Puzzling.
Pamela said,
November 2, 2018 @ 7:43 am
so on the basis of victor's response i am a bit strengthened in my suspicion that "bee" and the "mi-" roots in indo-european streams might actuallly share a source.
yes, my question about *kenəkó was an arch one and i'm glad you agree that huang is immediately suggested. i wonder about others. can k/h go toward s/sh in Turkic and Mongolian? can you get toward "sari" or Manchu "suwayan" –a strange word that doesn't seem to have a very obvious cognate in other Tungusic languages. if it is connected at some point to huang (like, really early, because Manchu turned huang into hûwang, but did Jurchen turn it into suwayan? i don't know about any of the rules there.
Peter B. Golden said,
November 2, 2018 @ 9:12 am
In Turkic the standard word for "honey" is "bal" which appears to have been borrowed from an Indo-European language (see Clauson, "Etym. Diction.: 330; Sevortyan, Etim. slovar' tiurkskikh yazykov, [1978]: 47) with the well-known m~b alternation: *mal > bal. Maḥmūd al-Kāšġarī, writing in the 1070s, in his "Diwān Lughāt al-Turk" ms.: 513, Dankoff ed. II: 228 notes "bal" as particular to the "dialects" of Suvārīn (PBG: a tribal grouping associated with Volga Bulgharia), Qifčāq (Qïpčaq) and Oğuz." In Khāqānī Turkī (the language of the ruling Turkic grouping in the Qarakhanid state) it is called "arï yağï" ("bee oil/fat"), a usage unique to Khāqānī. Arï yağï is not found in the modern Turkic languages.
Chris Button said,
November 2, 2018 @ 1:23 pm
@ ~flow
I think you mean 蜜 denoting something secreted? I’ve actually thought about that in the past too. It seems to me like it is a great mnemonic but is nonetheless coincidental since the sense running through characters like 密 and 祕 is rather associated with motionlessness/silence/compactness than something being set aside as we find with secret(e) with which the word “shit” is ultimately related. Furthermore, speaking of 屎 *ɬə̀jʔ, over on the “Dung Times” thread I linked to from the "Of Ganders…" thread above, Eidolon made the good point that the loanword 車 is a disruptive technology and hence able to be appropriated as a native word with its own word family; that can hardly be said about honey!
@ Pamela
Personally I'm not persuaded that there is any relationship there. Even just in terms of basic phonology it seems way too much of a stretch outside of superficial similarities with daughter languages whose forms are too recent for a graph attested in the oracle bones and with which we have no reason to assume contact in any case.
Jonathan Smith said,
November 2, 2018 @ 4:47 pm
Are people referring (also) to fen1mi4 分泌 'to secrete / secretion'? There could well be a connection to "separate" here and more generally in this group of words — see the first morpheme above, or early character forms and glosses for bi4 必.
Michael Watts said,
November 2, 2018 @ 7:02 pm
I don't see how. As dainichi notes, English is also right-headed and behaves exactly the same way. "Bee honey" would be a type of honey, and a "honey bee" is a type of bee. An "iron helmet" is a helmet of iron, and "helmet iron" is iron for, in, or otherwise associated with a helmet.
The only other example in Chinese that springs immediately to mind is 刷牙 v. "brush [one's] teeth" vs. 牙刷 n. "toothbrush", but the general idea is so completely normal that I would be very surprised to learn there weren't many others.
Eidolon said,
November 2, 2018 @ 7:25 pm
牛奶 cow milk, 奶牛 milk cow
Also works in English.
David Marjanović said,
November 2, 2018 @ 8:05 pm
For that you'd have to postulate a very irregular sound change working in this one root and nowhere else: PIE *m > Proto-Germanic *b is not regular (to put it mildly).
The Turkic phenomenon is different. Turkic languages are generally very unhappy about beginning a word with a nasal consonant unless the next consonant is also nasal; consequently, "I/me" comes out as bi/min.
Chris Button said,
November 2, 2018 @ 9:42 pm
@ Jonathan Smith
What a great point! I think I now need to retract everything I wrote above on the matter apart from the fact that I still feel associating 蜜 "honey" is a nice mnemonic but ultimately incorrect.
The word "certain" (as a gloss for 必) is related to the word "secrete" (as a gloss for 泌) via "discern" which all come from PIE *krei- (as such "excrete" is also related but not "shit" which I mistakenly wrote above that shares the semantics but is from a different root). This allows a further association with 祕 via "secret" and also 密 via extensions into 謐 via the related word "discreet".
It is then all the more tempting to suggest that 蜜 "honey" is also related as a "secreted" substance! However, I still feel that while it makes for a really great mnemonic, the semantic leap remains too big to counter the convincing Tocharian association.
Pamela said,
November 3, 2018 @ 7:35 am
"Personally I'm not persuaded that there is any relationship there. Even just in terms of basic phonology it seems way too much of a stretch outside of superficial similarities with daughter languages whose forms are too recent for a graph attested in the oracle bones and with which we have no reason to assume contact in any case."
so that makes me ask a question about something i know nothing about –when it comes to loan words, how does phonology really work? without the patterns govening changes within related languages, how do random elements like loan words predictably change?
Christian Weisgerber said,
November 3, 2018 @ 4:39 pm
Honigbiene and Bienenhonig are both common in German despite feeling fairly redundant. I guess Honigbiene would distinguish the insect from other kinds of bees (wild bees, solitary bees) that don't produce appreciable amounts of honey. Bienenhonig seems even more puzzling, since all honey is produced by bees, but I think I've seen some vague references to ersatz honey (made from sugar beet syrup?) in East Germany or generally poorer times… and now I see German Wikipedia even has an entry Kunsthonig ('artificial honey')… according to which the designation Kunsthonig has been banned in West Germany since the late 1970s.
Chris Button said,
November 3, 2018 @ 9:34 pm
@ Pamela
With (reasonably) fixed sound laws out the window, I suppose it all comes down to what is phonologically/phonetically plausible. I'm struggling to find any connection in that regard with 黃 *ʁáŋ (I prefer a uvular approximant /ʁ/ that conditioned labialisation where Zhengzhang has /ɡʷ/ but it makes little difference to the issue here). It's also paramount to use reconstructed forms rather than later reflexes (e.g. semantic shifts aside, the fact that Mandarin huáng "yellow" and Swedish honung "honey" might have a superficial resemblance is neither here nor there).
BZ said,
November 5, 2018 @ 12:03 pm
@Christian Weisgerber
There is something called date honey (https://en.wikipedia.org/wiki/Date_honey). In fact, when the Bible refers to the land of Israel as the "land flowing with milk and honey", it is referring to date honey.
Chris Button said,
November 5, 2018 @ 12:55 pm
@ Jonathan Smith
Further to the above, I've been looking a little closer at the graphic forms. While the Shuowen needs to be used with a lot of caution, it's analysis of 必 as being composed of 弋 and 八 seems worth considering (the bronze forms have 弋 with dashes either side which doesn't give me any reason to trust the traditional analysis of it as originally representing 柲). Treating 八 *prját as the phonetic and assuming an etymological relationship with 別 *prjàt "divide, distinguish", we then have a nice "ə/a" ablaut with 必 *pə̀c < *pjə̀t. I haven't actually covered the 必 series in my dictionary yet so this may of course change (I'll post anything here as long as the comments remain open), but I'm definitely going to make it my next entry….
Victor Mair said,
November 5, 2018 @ 10:13 pm
"The Word for ‘Honey’ in Chinese, Tocharian and Sino-Vietnamese"
Kristin Meier and Michaël Peyrot
Zeitschrift der Deutschen Morgenländischen Gesellschaft
Vol. 167, No. 1 (2017), pp. 7-22
https://www.jstor.org/stable/10.13173/zeitdeutmorggese.167.1.0007?seq=1#page_scan_tab_contents
Abstract
According to a widely accepted etymology, Chinese mì 蜜 ‘honey’ (MChin. mjit, OChin. *mit) is a borrowing from the Tocharian etymon represented by Toch. B mit ‘honey’. Recently, Jacques (2014) has argued on the basis of evidence from Sino-Vietnamese and Lakkja that the Chinese word should rather be reconstructed as MChin. mit, OChin. *mrit. He suggests that this word was borrowed from another Indo-European word for ‘honey’ (related to Greek μέλι etc.) that is not so far attested in Tocharian, but might have had a form close to OChin. *mrit. In our view, Jacques' arguments do not stand close scrutiny: the Old Chinese word cannot have been *mrit and the traditional Tocharian etymology still provides by far the best explanation for the Chinese etymon.
Jonathan Smith said,
November 6, 2018 @ 11:47 am
@ Chris Button Ah, interesting. Yeah structurally early "必" is reminiscent of "分". And I agree 別 is a similar word, however one cares to reconstruct the pair in OC. I think the form "八" is not simply phonetic as it must itself have first written a verb like 'divide', with rebus application to the numeral (same process for all of 5-9 presumably.)
John Chew said,
November 6, 2018 @ 5:49 pm
My Chinese isn't good enough to search for lists of reversible compounds (although if I were just slightly less lazy, I'd write a Perl script to do it). In Japanese though, a search for 逆さ二字熟語 gives webpages listing examples such as:
https://penoppe.com/archives/8408
蜜蜂 乳牛 肉牛 社会 情熱 議論 風雨 誕生 物品 左右 歌唱 階段 虚空 日本 中国 地下 科学 温室 温水 回数 分数 数奇 数倍 数日 数人 数個 十三
http://yhsi.hatenablog.com/entry/2013/09/16/103302
野原 原始 分子 心中 揮発 質素 生誕 機動 事情 故事 紙色 男色 男優 相手 火口 水着 天晴 世辞 数寄 半生 身分 露悪 高座 明文 温気
社会 進行 行旅 理論 議論 争論 理学 力学 里山 山沢 山猿 会議 議決 立国 毎日 頃日 手合 合併 一合 利権 利便 奇怪 宮中 道中 道筋 番茶 水流 海外 外人 名人 夫人 傑人 知人 証人 証書 筆名 名文 名声 天上 天下 足下 足高 死病 平和 細微 虚空 空中 関連 読解 解明 明澄 現実 質実 写実 著大 事大 事変 事物 異変 変転 王国 王女 子女 転回 転機 明言 根菜 性急 根性 毛根 線路 科学 学才 野外 家長 家出 家名 名実 年長 長身 長所 在所 所見 薄手 順手 草本 本元 手元 手先 先祖 末端 脈動 動作 作製 作詩 背中 後背 光栄 光彩 光風 子種 弟子 弟兄 兄貴 末期 口早 数日 数多 段階 楽器 地平 勢威 勢力 戦力 力水 冷水 温水 温室 宇宙 分野 導引 導電 電荷 花火 種火 苦労 苦痛 出前 出産 出演 出発 進発 発熱 熱暑 展出 魚雷 間隙 重体 体得 心得 重荷 形象 形体 息子 音子 落下 陸上 頂天 練習 習慣 貫通 馬子 声音 親父 親近 後生 生殺 生卵 住居 立木 木材 俗習 俗世 離別 格別 個別 頭目 目尻 目上 波音 波長 見識 見物 干物 置物 運命 運気 気心 気血 座主 君主 下臣 臣従 問訊 難詰 風防 風雨 余剰 欷歔 礼儀 礼拝 拝跪 祭司 結集 敬愛 求愛 愛恋 凹凸 幕内 観客 憤激 当該 当日 時日 政治 治療 感情 人情 評定 規定 評論 表裏 対応 地境 争闘 始終 途方 食餌 託宣 滅法 肉食 物色 羽毛 陵丘 呼称 勇武 版図 服喪 容認 庭石 石化 増加 僧尼 厄災 説教 泉源 福祉 歌唱 同類 虜囚 育成 次席 保留 蜂蜜 追訴
日本 中国 保健 安保 内部 外部 車外 孫子 数字 通電 太極 陰陽 日向
Victor Mair said,
November 6, 2018 @ 9:41 pm
From Axel Schuessler:
Several insignificant reactions:
(1) The Toch.B word mit ‘honey’ is derived from an IE root *medh- ‘sweet’, with nominal suffix *médhu- referring to various sweetish substances, Skt mádhu ‘honey', Gk. μέθυ ‘some alcoholic beverage’ (presumably other than wine), German Met (long ē), Engl. mead. Small problem: the Toch B word that actually corresponds phonologically to *medhu- is Toch B mot (<*mēdhu-) ‘alcoholic drink’ (Mallory&Adams p. 262). So what form exactly does Toch B mit go back to? (rather than guessing, let a Toch. specialist answer that). In any case, if the Toch.-OC relationship is one of borrowing and not chance resemblance, Chinese must have borrowed it from Toch. Why? The Chinese must have been aware of honey since time immemorial. But: languages engage in strangest borrowing, thus Uralic languages borrowed from PIE words for ‘water’ and ‘woman’. // (2) There is no way to put an *r into OC *mit ‘honey’, no matter who claims that and what marginal language may suggest this. And the speculation about the connection with a root, as found in Gk. μέλιτ- is pointless. How and via which languages or travelers is that supposed to have reached the East? As an idle exercise, I spent 5 minutes recently to come up with more than a dozen Gk.-OC correspondences that looked even better than Sagart’s AN-OC equations; but this Gk-OC is of course complete nonsense, as can be shown when probing into details. The world’s languages are full of lookalikes. // Shuowen mentions a graphic variant for 蜜 with 鼏 (Qieyun System miek) as phonetic. I am surprised that so far no one has seized upon this. Yet if this variant should have some merit (which I doubt), then mi ‘honey’ might have had a final OC *-k which would eliminate the connection with Tocharian. // (3) Both IE stems that can among others mean ‘honey’ (*mélit-, *médhu-) have survived in Germanic: one in Engl. “mildew", one in “mead”. In both instances, the meanings specialized, thus creating a semantic gap that was then filled with ‘honey’. This shift or loss of meaning, even a word, in a branch of a language family, as in this case, is nothing mysterious and particularly noteworthy; it happens all the time. The IE root for ‘dog’ survived in Engl. in the specialized word ‘hound’, therefore another word for ‘dog’ filled its semantic place, i.e. ‘dog’ of uncertain origin. And so on. // Michael Witzel (“Early sources for South Asian Substrate Languages," 1999, p.66) suggested that the roots *mélit-/*médh- entered IE from some unknown source in Eastern Europe or Central Asia (so where did OC *mit really come from?). // (4) The connection of ‘honey’ with Gk. κνηκός ‘yellowish’ also κνῆκος ’saffron’?, Skt. kánaka ‘gold’, kāñcanám ‘gold’ looks doubtful to me. Significantly Mayrhofer in his Indo-Aryan etymological dictionaries just mentioned these ideas without endorsing them. To the contrary, kánaka and its cognate kanala ‘shiny’ (‘scheinend, glänzend’) are suspect of being Dravidian or Austroasiatic loans (so Mayrhofer), these words certainly look like such and cannot have anything to do with the Western items. The reconstructed IE *kṇ̄ǝk (or whatever) does not look IE at all. Also, Skt. unpalatalized /k/ would be odd in an IE etymon. So these words seem like loans (even the meanings 'gold, saffron’ are suggestive of that). And extending these to ‘honey’ in Germanic seems a bit of a stretch, unless regular phonological details can support this. — Or might ‘honey’ be borrowed from Balto-Finnic??
Victor Mair said,
November 6, 2018 @ 10:33 pm
Some notes on early forms of huáng 黃 ("yellow"):
Phono-semantic compound (形聲, OC *ɡʷaːŋ): phonetic 炗 + semantic 田 (“field”) – the color of earth.
https://en.wiktionary.org/wiki/%E9%BB%83#Glyph_origin
From Matt Anderson:
That analysis from wiktionary is from the Shuo wen, and it is clearly wrong as far as early forms of 黃 — in the OBIs, for example, it contains no element even resembling 炗, and there is no 田 either, though at least there’s a component that sometimes looks like 田.
As far as what it actually signifies, I’m not entirely sure, and there is lot of disagreement about it. I think Jonathan Smith wrote in his dissertation that it represents a person with a jade ring around them, which I think is as likely as any other suggestion (such that it is a picture of a jade pendant or that it’s a “shaman” with a full belly, among others). It’s closely related to the graphs shǐ 矢 ‘arrow’ and yín 寅 ‘an earthly branch’, which helps but also confuses matters. Both shǐ and yín were at different points written interchangeably with each other, and one form of huáng is sometimes interchangeable with one form of yín, but no form of huáng is ever interchangeable with shǐ.
Jonathan Smith:
I did write on this, elaborating on Guo Moruo's line of thought in his study "Shi huang" 釋黃 ("…黃即佩玉，自殷代以來所舊有。後假為黃白字，卒至假借義行而本義廢，乃造珩若璜以代之，或更假用衡字。後世佩玉之制廢珩璜字義各限於佩玉之一體…"). More specifically, the graph apparently first wrote a word meaning '(jade) ring (pendant)', with 'yellow' happening to be a close homophone.
Suburbanbanshee said,
November 7, 2018 @ 9:40 am
There is a category of preserves and jelly where you make a gooey, not very long-lived sort of fruit jelly that is called a "honey." Quince honey is the kind I have made, because it has a lot of natural pectin.
Chris Button said,
November 7, 2018 @ 1:06 pm
Personally I've always like Qiu Xigui's discussion of the graphic confusion of 黃 with 交.
I suppose that could be put down to the -t / -k variation in 幭 as another way to write 鼏.
Chris Button said,
November 8, 2018 @ 10:33 pm
It might be noted that the Shuowen analysis is nonetheless "etymologically" most likely correct in terms of the relationship between the words represented by 黃 *ʁáŋ as a color term and 光 (炗) *qáŋ even though there is no original graphic association.
It seems like the article is not yet available on Jstor. Perhaps too recent? In any case, in light of the article, Jacques seems to have abandoned his proposal:
https://panchr.hypotheses.org/1894