Plant-based "milk"

« previous post | next post »

The company Oatly claims to have created a new Chinese word for plant-based milk by placing the grass radical above the character for milk:

Caitlin Schultz, who called this development to my attention, remarks:

Is creating a new character like this legitimate, especially in a corporate context? This feels to me like linguistic colonization or cultural appropriation. Calling plant milks "milk" in English has been contested, as well. If they really feel they need a new word to describe this category of product, isn't a variant of dòujiāng 豆浆 ("soy 'milk'"), such as yànmài jiāng 燕麦浆 ("oat 'milk'"), more appropriate? What does it take to create a new character, and how does it get into character inputting systems? I'm sure this raises other linguistic questions, but those were just a few I had!

All the questions raised by Caitlin are legitimate.  In addition, one wonders how one ought to pronounce this new character.  The same as the original character for "milk", nǎi 奶?  But then, when you speak it, you wouldn't be able to distinguish between the new character and original one.

The photo above comes from this article:  "Plant Milk Brand Oatly Creates Chinese Word Meaning ‘New Milk’ To Drive Health Awareness", by Jenny Star Lor, Green Queen(4/15/19).  In this article, it is asserted that the new character means "new milk", but such an interpretation is not justified on the basis of its composition and derivation.  The article also confuses the difference between "word" and "character", but that is a problem endemic to nearly all writing about Chinese language and script, alas.

When we contemplate this new character, all sorts of other issues assail us.  As I write this, I'm sitting in the food court of a Giant supermarket.  I walk back to the dairy department and I see a whole wall of products called "milk" that are not from the mammary glands of mammals — oat milk, coconut milk, almond milk, walnut milk, not to mention soy milk, which has been around a lot longer (most of the ones I'm seeing now have just popped up in grocery stores within the last few months or years).  Then I walk over to the tea section of the store and I see lots of things labelled "tea" that are not made from Camellia sinensis.  I suppose the same situation would obtain if I strolled over to many other sections of the store as well.

Meanwhile, I just fortified myself with a blueberry coriander lassi and a mango rose water lassi so that I could finish this post. Yum!

P.S.:  If we take the Oatly creation strictly as an advertising gimmick, then I have to admit that it's pretty clever and catchy, but we should not consider it seriously as a new character that we have to add to the already burgeoning Chinese writing system.



  1. Alex DelPriore said,

    April 16, 2019 @ 10:09 am

    Apparently this character composition is actually far from novel, already encoded in Unicode, with a Taiwanese national standard in place since at least 1992 as reference:

    If it were to catch on in text, perhaps it could encourage widespread compatibility with Supplementary Ideographic Plane codepoints, much as emoji did for the Supplementary Multilingual Plane

  2. WSM said,

    April 16, 2019 @ 10:15 am

    A grass radical plus 乃 on the bottom would be more consistent with how 奶 was constructed (not sure if the former doesn't already exist somewhere in the Unisphere). In any case, have fun trying to create a social media campaign centered around a character that's impossible to represent digitally.

  3. Shihchuan said,

    April 16, 2019 @ 10:41 am

    @WSM: I did some search, and as it turns out, there is: 艿 (it's pronounced nǎi as well, and it's part of a word that's the alternative name for "taro": 芋艿 (yùnǎi). Not that I've ever encountered it in everyday life…)

    In EU they actually stipulated that anything that's not produced from mammary gland cannot be labeled "milk" (with several exceptions such as "almond milk" and "coconut milk"……real consistent, isn't it?) At least in France as far as I know, soy milk are now labeled "boisson au soja" (soja drink). Of course, "lait de soja" (soy milk) remains the far more common way for ordinary people to talk about it in everyday life.

    And I doubt that it's really linguistic colonization or cultural appropriation: at least in Taiwanese one also says 豆奶 (tāu-ling, "soy milk") and 米奶 (bí-ling, "rice milk"), with the analogy to cow milk present. It's more of a publicity stunt than anything.

  4. michaelyus said,

    April 16, 2019 @ 10:48 am

    Some context: one might note that the milk-mylk in English has been propelled into mainstream knowledge, at least here in London, UK, because of the 2017 EU court ruling that "only liquid from animals could be called milk". 2017 article from The Guardian: Oatly is originally based on technology from Lund University, Sweden.

    As for in Chinese, all of 浆, 奶 and even 汁 apply to coconut milk (leading to confusion with coconut water for 浆 and 汁); for almonds, 奶 and 浆 have both been used [although 奶 considerably more].

    Rice and soy are most commonly 浆, and are of course very well known in the Sinosphere. However, 豆奶 for soy milk is also a common alternative, and in Taiwanese Hokkien is the major form, as well as being more common in HK Cantonese (although that might be because of a certain well-known brand). Oat milk is considerably less well known but at least on the Internet is attested (as 燕麦浆).

  5. Jeff Johnson said,

    April 16, 2019 @ 10:57 am

    The point that “calling plant milks ‘milk’ in English has been contested” is not a very good one. It’s only contested by the dairy industry. In normal usage this is a long-standing and widely accepted meaning. Cf. The Smithsonian here:

  6. Ricky said,

    April 16, 2019 @ 1:44 pm

    I realize I'm falling into a trope, but almond milk has been a part of European culture since the Middle Ages, probably longer than soy milk…

  7. Chandra said,

    April 16, 2019 @ 2:40 pm

    Yes, as Jeff Johnson says it's the dairy industry trying to protect their interests rather than a legitimate protection against false advertising. They've already been successful in other areas, such as requiring the spelling "creme" for anything that isn't dairy-based cream, and "cheeze" or other variants for vegan cheese products.

  8. Kristian said,

    April 16, 2019 @ 3:08 pm

    All of Oatly's ads are cutsy and gimmicky, inventing a Chinese character for themselves follows the same pattern.

  9. Suburbanbanshee said,

    April 16, 2019 @ 3:28 pm

    Given how important that almond milk is, specifically to Hispanic and Portuguese cuisine, as well as to a lot of other Mediterranean cuisines, I have to question the dairy industry's sanity.

  10. AntC said,

    April 16, 2019 @ 5:59 pm

    'milk' used "Of milk-like plant juices or saps from c. 1200." [etymonline]

    Does dòujiāng 豆浆 ("soy 'milk'") — the pronunciation or characters — use the same as (animal) milk? Prof Mair says "milk", nǎi 奶, which looks and sounds unrelated(?)

  11. Alex said,

    April 16, 2019 @ 6:41 pm

    "but we should not consider it seriously as a new character that we have to add to the already burgeoning Chinese writing system."

    This is part of the issue on efficiency to evolve and absorb new words such as burrito, chimichanga, quesadillas, tacos, fettuccine, rigatoni, ravioli, etc

    imagine if the English language had to create new words only by using existing single syllables

    met sick oh roll pan cake

    take a guess what this Mexican food item is.


  12. Alex said,

    April 16, 2019 @ 7:08 pm




    of course perhaps the English might change the pronunciation a bit for Japanese food but the key point is most items are unique thus more easily remembered

    是拉差香甜辣椒酱 baidu
    拉差辣酱 google

    The question is if the language can scale as the world becomes more integrated with so many tourists from China visiting other countries wanting to find those cultural items at home to eat.

    Moreover in literature do the passages flow. For example if one writes about visiting a restaurant on a date and describing the courses

  13. Toby said,

    April 16, 2019 @ 9:28 pm

    "linguistic colonization or cultural appropriation" is not a legitimate question, but represents a distinctly retrogressive and unhumanist point of view. It is distinctly antithetical to this blog as well.

  14. Victor Mair said,

    April 16, 2019 @ 11:37 pm

    "'linguistic colonization or cultural appropriation' is not a legitimate question"

    I beg your pardon? Are you saying that it should not be permitted to discuss such matters on Language Log? Even if one disagrees with those who raise such issues, are you saying that we should shout them down and stifle their voices? They should not be allowed to speak?

  15. Chas Belov said,

    April 17, 2019 @ 1:58 am

    Conversely, a slang term for cow milk is "moo juice."

  16. Greg Price said,

    April 17, 2019 @ 3:17 am

    Thanks to Alex DelPriore for pointing out, above, that this character is not newly invented — it already exists! The character is U+2B1F5 in Unicode; and Unicode identifies that with 12-4F52 (aka "12th plane, 4F52") in CNS11643, which is a Taiwanese standard.

    Here it is: but the fonts I have don't seem to have a glyph for it.

    In addition to the link on the Unicode website that Alex provided, a Taiwanese government site provides information from the CNS11643 standard, including glyphs in several styles:
    The glyphs do indeed match the composition described, with 艹 over 奶.

    I'd be curious to learn where and how it's been used, or what sources led to its inclusion in the Taiwanese government standard. It's possible that information is there on that CNS11643 website, for someone literate in Chinese to find.

  17. Mimi K said,

    April 17, 2019 @ 6:32 am

    Regarding "products called 'milk' that are not from the mammary glands of mammals":

    Some plant-based milk brands use the spelling "mylk" just for this reason. Similar to your conjectures about the 艹 + 奶 character, the difference is imperceptible in speech.

    Some very vegan person even wrote an Urban Dictionary entry for "mylk":

  18. Robot Therapist said,

    April 17, 2019 @ 6:46 am

    And we are probably going to have a similar battle about "meat" grown in tanks.

  19. Greg Price said,

    April 17, 2019 @ 7:15 pm

    Oh also from that page on the CNS11643 site: the pronunciation is listed as "nǎi", so indeed the same as 奶.

  20. Bloix said,

    April 17, 2019 @ 9:01 pm

    On English "milk" as applicable to non-dairy products, see Samuel Johnson, A Dictionary of the English Language (1755):
    milk, n,
    1) the liquor with which animals feed their young from the breast;
    2) emulsion made by the contusion of seeds, as in, milk of almonds.

  21. R. Fenwick said,

    April 18, 2019 @ 1:48 am

    @Bloix, English usages for "milk" as relevant to plant-based fluids go right back to the Old English rendition of Pseudo-Apuleius (c. AD 1000-1100), where ylcan wyrte meolc "milk (=sap) of spurge-wort" is offered as a treatment for warts (cix:154).

    For a non-English parallel I also can't help but think of Latin lac "milk", which does primarily refer to mammalian milk (and in that sense supplies the Latin idiom lac gallinaceum "chicken's milk" = "something exceedingly rare", apparently an equivalent to English hen's teeth or rocking-horse shit), but is freely used to refer to milky plant saps as well:

    pubentes herbae nigri cum lacte veneni "soft black herbs with venomous milk" (Virgil, Aeneid, 4:514)
    lac caprifici "milk of the wild fig" (Celsus, De Medicina, 5:7)

    and then of course there's that supreme milk-bearer, the lactuca—or in its modern English descendant, lettuce.

  22. Adam F said,

    April 18, 2019 @ 2:57 am

    I was under the impression that milk (other than from humans for babies) was generally considered an unpleasant secretion (or something like that) and not generally consumed in East Asian cultures, so I'm a little surprised that they would market it in China as "milk".

    (I'm open to correction on any of that.)

  23. Philip Taylor said,

    April 18, 2019 @ 6:10 am

    Adam — condensed milk is very popular in Việt Nam, where it is routinely used as an ingredient in white coffee (both cà phê sữa đá and cà phê sữa nóng).

  24. AntC said,

    April 18, 2019 @ 6:37 am

    I was under the impression that milk (other than from humans for babies) was generally considered an unpleasant secretion (or something like that) and not generally consumed in East Asian cultures, …

    Heck you're waaay out of date. As part of getting wealthier/more Westernised, the Chinese are consuming vast amounts of milk — particularly as baby formula; but increasingly also as an ingredient in Western style fastfoods. New Zealand exports nearly all its milk production to China/East Asia, and has turned over previously uneconomic land to graze cows.

    This has needed enormous irrigation schemes and vast amounts of fertilisers, totally changing the nature of the (ex-)'back country'. And it's ruining our ecology (see reports in the press today).

    Doubtless in a decade or so, the Chinese will be suffering all the health problems of an over-refined diet. And who-knows what digestive challenges because many are lactose-intolerant.

  25. Victor Mair said,

    April 18, 2019 @ 7:00 am

    As a child, I was fascinated by the plant with the name "milkweed".

  26. Victor Mair said,

    April 18, 2019 @ 8:36 am

    From the Twitter account of Chenchen ZH:

    "Tbh this looks really bad to anyone who's familiar with internet subculture. Think 草 as in 草泥马 and 奶 as in boobs."

    Decoding this, we have:

    nǎi 奶 ("milk; breast; boob; lady")

    cǎo 草 ("grass; vegetation")

    cǎonímǎ 草泥马 ("grass mud horse") — ubiquitous internet meme for the near-homophonous "cāonǐmā 操你妈" ("f*ck your mother")


    "Grass-Mud Horse Lexicon Classics" (8/30/13)

    "Polysyllabic characters in Chinese writing" (8/2/11) — especially this comment

    "Franco-Croatian Squid in pepper sauce" (3/12/09)

    Also mentioned in countless other Language Log posts.

  27. Luo said,

    April 18, 2019 @ 8:48 am

    "I'd be curious to learn where and how it's been used, or what sources led to its inclusion in the Taiwanese government standard. It's possible that information is there on that CNS11643 website, for someone literate in Chinese to find."

    @Greg Price

    Based on the CNS11643 link (, it states that the source comes from "內政部戶政用字", which is Department of Household Registration (

    A further review into New Character Application in the CNS website (, in footnote 2 it states that "內政部戶政司" (Department of Household Registration) is responsible for new character application of "姓名用字" (i.e. a person's name), so very likely the character is created for someone's name purpose.

    (Please correct me if I'm wrong).

  28. Jichang Lulu said,

    April 18, 2019 @ 9:15 am

    Although it has already been cited, I cannot resist relinking to that post on wolf's milk.

  29. Chandra said,

    April 18, 2019 @ 12:18 pm

    @R. Fenwick – I wasn't familiar with the expression lac gallinaceum, but now I wonder if it explains the French name for eggnog, "lait de poule", which I've always found quite amusing. I had always assumed it referred to the combination of eggs and cream in the drink.

  30. Trogluddite said,

    April 18, 2019 @ 3:06 pm

    In English, at least, there are also several inorganic "milks" – e.g. milk of magnesia (a suspension of Magnesium Hydroxide), or moonmilk (a mineral deposit found in limestone caves.)

  31. Victor Mair said,

    April 18, 2019 @ 4:11 pm

    An anonymous correspondent asks:


    Is the grass radical used to allude to "grassmudhorse" or something like that more generally? (Or was it when that euphemism became popular?)

    I ask because I have a vague recollection of a Chinese person telling me that young people would sometimes put 艹 on top of 我, so using 莪 with a first-person meaning, but I didn't get a clear sense of why. I don't know, but if 艹 was an obscene shorthand, it would help make sense why the Twitter user made that connection.


    The first thing I must say is that the character é 莪 has been around for at least several hundred years and is used to signify various types of artemisia and other plants as well as an alternative designation for fungi.

    However, for those who have a dirty mind, one could interpret it as "grass me", where the "grass radical 艸 / 艹 / 䒑" stands for the near homophone cāo 操 ("exercise; drill; hold; do; act; grasp; grip; grapple; gripe; behave; conduct; clutch; take in hand; manage"), which is internet slang and a euphemism for cào 肏 (“f*ck”).

    For references, see comment #25 (the sixth comment above this one).

  32. Greg Price said,

    April 18, 2019 @ 7:11 pm

    Thanks @Luo! Very interesting.

    I wonder if glyphs for this character appear in fonts commonly used in Taiwan, or by the ROC government. (There's no such glyph in the fonts I have on my machine, particularly the Noto family.) It seems like it'd be awfully inconvenient to have a character in one's name that many systems will just not know how to display.

    The 12th plane of the CNS11643 standard, where this character appears, didn't have any assigned characters in the 1992 version of the standard, but it did in the 2007 version. So most likely this character was there by 2007 — but in 1992 would have been basically impossible to enter on a computer.

    In decades past, when less of daily life and bureaucracy happened on a computer and people's names were mainly hand-written (*), I imagine there was more room to use a rare or novel character (so long as it was composed of familiar pieces in a familiar kind of pattern, like this one's is.) I wonder if this character was really new when the 內政部戶政司 requested to add it to the CNS11643 list (i.e. sometime in the '90s or '00s) — or if some people had been using it in their names for a long time, and it was late only in its inclusion in the computerization of things.

    (*) Which was still largely true in the US of my childhood in the '90s. I'm guessing the timeline in Taiwan wasn't radically earlier.

  33. Greg Price said,

    April 18, 2019 @ 7:17 pm

    > The 12th plane of the CNS11643 standard, where this character appears, didn't have any assigned characters in the 1992 version of the standard, but it did in the 2007 version.

    (There's some information on the web mistakenly attributing all the planes to the 1992 version; even a page on the Unicode website has this wrong. But see e.g. this document in the workings of the Unicode committee, which includes a request from the Taiwanese member body to correct this small error: )

  34. Luo said,

    April 18, 2019 @ 7:30 pm

    Thanks @Greg.

    That's an interesting remark, character could be more creative when it were mainly hand-written in the past decades, now we are bound by the limitation of computer font.

    Re your question, as far as I know, glyphs for this character is not available as in any common font I have.

  35. Andrew Usher said,

    April 18, 2019 @ 10:07 pm

    But surely, the animal use of 'milk' is primary and other senses are by derivation or metaphor. I think many people would agree with me that 'milk' sensu stricto is what comes from a tit and the others always require a qualifier without unambiguous context.

    Also note that the word 'juice' esp. in its main modern sense is not that old and probably today it would have been used for the edible/potable plant-based products; 'milk' in that sense only has antiquity.

    k_over_hbarc at

  36. Keith Ivey said,

    April 19, 2019 @ 10:00 am

    Are dairy lobbyists also attacking "cream of wheat"?

  37. Rodger C said,

    April 20, 2019 @ 8:49 am

    the word 'juice' esp. in its main modern sense is not that old and probably today it would have been used for the edible/potable plant-based products

    To me, juices are squeezed. Almond milk and soy milk, are, as Dr. Johnson said, contused (contunded? contounded?).

  38. Adam F said,

    April 22, 2019 @ 6:04 am

    @AntC Thanks for the update. I'd noticed some Chinese people at our university eating dairy products, but I assumed it was a habit they had picked up in the UK.

  39. J. Goard said,

    April 23, 2019 @ 12:42 am

    Funnily, Naver's English-Korean dictionary translates English "rice milk" as:

    쌀가루를 섞어 끓인 우유 'boiled (cow) milk with rice flour mixed in'


    우유죽 '(cow) milk porridge'.

    Although Korean uses the Sino-Korean root 유 /ju/ 'milk' in the established 두유 /tuju/ 'soy milk', as a sometime vegan in Korea, I'm used to seeing names of all the recently popular plant milks borrowing 밀크(milk) from English.

RSS feed for comments on this post