Onigiri > Onigilly

« previous post | next post »

Brand-name transliteration (in Embarcadero Center, San Francisco), courtesy of Nancy Friedman:

Koji, a first generation immigrant from Japan and the founder of ONIGILLY, surely knows how to pronounce "onigiri おにぎり" ("rice ball") the Japanese way.  So it's a humorous mystery to me why he decided to transliterate it as "onigilly".

Hypotheses:

1. he thought spelling the name that way would help hapless foreigners pronounce it to sound more like Japanese than if he had spelled it the standard way as "onigiri" — lord knows he's trying hard enough with his pronunciation guide to instruct us how to say the name:  oh-KNEE-ghee-lee

2. brand-name distinctiveness and recognition

Whether they're "onigiri" or "onigilly", they certainly look delicious and nutritious!



47 Comments

  1. Suzanne Valkemirer said,

    November 8, 2018 @ 11:16 am

    Is the answer not that since an allophone of Japanese /r/ is [l] (though it does not have a phoneme */l/, distinct from the phoneme /r/), speakers of the language who cannot pronounce English well enough to distinguish the English phonemes /r/ and /l/ consider those English phonemes to be allophones of the same phoneme and therefore consider the English letters and to be allographs of the same grapheme, with the result that they may use one or the other letter indifferently?

  2. Coby Lubliner said,

    November 8, 2018 @ 11:43 am

    There is a Japanese restaurant in Berkeley called Kirala, though the Japanese name is キララ (kirara). I have long wondered about the reason for the L substitution for the second R. It's something that sometimes happens in Spanish: 'vegetable' is verdura, but 'vegetable seller' is verdulero.

  3. Bloix said,

    November 8, 2018 @ 11:59 am

    This reminds of Piccadilly, most likely from Piccadilly Hall, the derisive nickname given to a mansion built by a one-time tailor who was said to have made his fortune selling piccadills, or lace collars – from Spanish picadillo, from picado (pierced), which had a corresponding Spanish word, picadura (also a kind of lace collar) – an l/r substitution at least superficially similar to one at work here.

  4. Athel Cornish-Bowden said,

    November 8, 2018 @ 12:01 pm

    I'm not sure how common it is in Spain, but in Tenerife there is a supermarket chain called Alcampo. When I saw its logo I thought it looked familiar, and then I realized it was the same as that of the large French chain Auchan. Clearly "Alcampo" is a calque of "aux champs" (or maybe "au champ", but somehow the plural looks better, despite the singular Spanish).

  5. Bloix said,

    November 8, 2018 @ 12:02 pm

    Coby Lubliner – I posted my comment without reading your observation regarding l/r in Spanish – apologies for not acknowledging it.

  6. Athel Cornish-Bowden said,

    November 8, 2018 @ 12:05 pm

    @Coby Lubliner. Consider Latin arbor (r-r), which is árbol (r-l) in Spanish and albero (l-r) in Italian. Apparently modern Latin speakers don't like r-r in one word. French arbre probably doesn't count as a French r isn't remotely like a Spanish or Italian r.

  7. Gabriel Holbrow said,

    November 8, 2018 @ 12:19 pm

    I think the key is that "Onigilly" would be pronounced exactly as オニギリ (onigiri, same as the delicious rice ball) when put into Japanese phonology, as Ms Valkemirer pretty much points out above. That is, from a Japanese perspective "Onigilly" is an unambiguous and acceptable romanization of おにぎり.

    Remember also that there is quite a tradition of commercial and brand names in Japan, especially names of restaurants, using creative romanizations to write familiar Japanese words. The first example that comes to my mind is the shipping mall on Odaiba (お台場) island in Tokyo called "Diver City", where "diver" is intended as a romanization of 台場 (daiba, as in the name of the island) and also incorporates a pun on "diversity" to suggest the wide variety of shops and eateries there. https://mitsui-shopping-park.com/divercity-tokyo/

  8. Cervantes said,

    November 8, 2018 @ 12:29 pm

    Maybe not really on topic, but Puerto Ricans pronounce "r" much like "l". It seems pretty easy to slip between those sounds. In any case, transliteration is not standardized and if you think onigilly is a better representation of the Japanese pronunciation go for it. Note that there is no standard at all for transliteration between Arabic and English. I saw the name of the late Libyan dictator spelled Kadafi, Kaddhafi, Gaddhafi, Gadafi, Kadaffi, Gadaffi and probably other variations as well. (I don't know anything about Arabic but evidently the Arabic consonant is between K and G. People pronounce Qatar variously as cotter, gutter, catarrh, and gotter. Maybe somebody knows which is closest to correct.)

  9. Max Wheeler said,

    November 8, 2018 @ 12:36 pm

    If he wants it stressed on the second syllable, writing it with – ll- is perverse.

  10. Ginger Yellow said,

    November 8, 2018 @ 12:44 pm

    I suspect it's a rather prosaic reason, namely that Onigilly is trademarkable and Onigiri wouldn't be.

  11. James Unger said,

    November 8, 2018 @ 1:14 pm

    When we lived in New Zealand, I once phoned in a prescription for my then fiancee, whose family name is Okumura. The chemist at the other end didn't flinch at all. I learned why when I picked up the prescription; the label on the bottle read "Miss O'Cumureagh." O'Nigilly sounds like a name Dickens would have made up.

  12. Victor Mair said,

    November 8, 2018 @ 1:20 pm

    Cf. Genghis Khan, alternatively transliterated as Chinggis Khaan, /ˈɡɛŋɡɪs ˈkɑːn, ˈdʒɛŋ-/; Mongolian: Чингис хаан, Çingis hán [t͡ʃʰiŋɡɪs xaːŋ].

  13. cameron said,

    November 8, 2018 @ 1:24 pm

    I agree with Max Wheeler above. The choice of spelling is directly at odds with the suggested pronunciation.

  14. Philip Taylor said,

    November 8, 2018 @ 2:41 pm

    Professor ZSZ, from SISU, asserted that "the Japanese cannot pronounce the sound /l/, which is why they mistranslitterate Nala as Nara" when I spent a few weeks with him and his family in Kyoto.

    And. for me, Qatar and catarrh are pronounced identically (sorry, no IPA from this accursed Android keyboard).

  15. Philip Taylor said,

    November 8, 2018 @ 2:43 pm

    Professor Shanghzhi from SISU asserted that "the Japanese cannot pronounce the sound /l/, which is why they mistranslitterate Nala as Nara" when I spent a few weeks with him and his family in Kyoto.

    And. for me, Qatar and catarrh are pronounced identically (sorry, no IPA from this accursed Android keyboard).

  16. astrange said,

    November 8, 2018 @ 2:44 pm

    I would just like to inform readers of this comment section that there is a Japanese curry restaurant in Portland called "Kalé".

  17. John from Cincinnati said,

    November 8, 2018 @ 3:34 pm

    @Athel Cornish-Bowden. re: Apparently modern Latin speakers don't like r-r in one word. Spanish torero, Italian torero, Portuguese toureiro, all from Latin taurarius, bullfighter.

  18. Gabriel Holbrow said,

    November 8, 2018 @ 4:09 pm

    Pardon my commenting again, but I just thought of the all-time best (in my opinion) creative romanization of a common Japanese word as a brand name: Mazda for マツダ, from the name 松田 (conventionally romanized Matsuda).

  19. Allen Thrasher said,

    November 8, 2018 @ 4:47 pm

    Several disjunct comments:

    How do they claim to know that, as the sign in the photo says, onigiri "have a history that spans 2,300 years"? Is it mentioned in some Chinese classic of 300 BC or earlier? If the person who composed the sign was referring to a history in Japan, I have no idea how anyone could know this, unless onigiri had been discovered in a tomb of that date. The frequent assumption that every 'Eastern' culture or civilization is more ancient than 'Western civilization' irritates me., since IMHO it is an error.

    In reference to astrange's reference to "a Japanese curry restaurant in Portland," I am curious to know if there is such a thing as Japanese curry, in the sense of an special variety of curry originated by and cooked by Japanese, or is it just that a restaurant. In Portland specialising in curry is owned and run by an ethnic Japanese. If the former, the DC area has done poorly in the diversity sweepstakes. At least I have not heard of such a restaurant in the region.

  20. Jim Breen said,

    November 8, 2018 @ 5:07 pm

    @Allen Thrasher, Japanese has a style all of its own. There was an article in the Japan Times a year or so ago by the outgoing Indian ambassador, in which he described his investigations into Japanese curry. He discovered it came into the Japanese cuisine via the navy, which had adopted it from the (British) Royal Navy in the late 19th/early 20th century.

  21. Victor Mair said,

    November 8, 2018 @ 5:19 pm

    House Curry is a very big deal in Japan, with many restaurants throughout the country. I like to eat in them when I'm in Japan, because I know that I'll get a reasonably wholesome, tasty meal for very little money (relatively speaking). They kinda remind me of Waffle House.

    Japanese style house curry spread to America already in the early 80s, with one of the most famous brands being Vermont Curry. I have no idea why they chose "Vermont" for their brand name, but it supposedly offers the "genuine flavor of Japanese cooking".

    http://flavorofjapan.com/

    Another favorite type of Japanese curry that I sometimes indulge in is karee raisu カレーライス ("rice with curry").

    http://languagelog.ldc.upenn.edu/nll/?p=21160#comment-1500726

  22. Victor Mair said,

    November 8, 2018 @ 5:23 pm

    From Roger Allen:

    There definitely is a standard transliteration system from Arabic to English, that established by the Library of Congress. Every dissertation on Arabic submitted in the United States must conform with it. That does not mean, of course, that everyone is aware of it or uses it….

    As for Qatar, the closest pronunciation of it in English is the word "gutter." Many, if not most, people pronounce it as "catarrgh," but, at all events, the emphasis in the word is on the first of the two syllables, not the second.

  23. AntC said,

    November 8, 2018 @ 5:25 pm

    I am curious to know if there is such a thing as Japanese curry

    To reinforce @Jim's comments: yes there is. It's distinctly curry and distinctly not like Indian or other S.E. Asian currys.

    Lots of cumin, modest amounts of chilli, and quite sweet. Also often served with lumps of (processed) cheese — a very pale imitation of paneer.

    In Asian grocery stores, you can get a stock cube/paste "smooth and creamy"/"rich roux with caramelized onions".

  24. Brett said,

    November 8, 2018 @ 5:28 pm

    To me, those onigiri really, really look like seaweed hamantaschen.

  25. Krogerfoot said,

    November 8, 2018 @ 5:39 pm

    I think both hypotheses in the OP are correct.

    Also, from my NAm English speaker perspective, the double-L spelling seems to encourage a closer approximation to the Japanese word than a single-L version would. "Onigilly" seems more readable than "onigily," to me, and even if readers end up stressing the third syllable instead of the second, it's still pretty close to the Japanese. It additionally seems likely that a single-L version would elicit a /j/ sound for the g in a lot of cases.

  26. Victor Mair said,

    November 8, 2018 @ 5:54 pm

    From Joe Lowry (re Arabic):

    There are standard transliteration systems in use in academia, but they are not popularly in use (newspapers, non-academics), they are technical (because they use diacritical marks that are in some cases unusual in English), and then many native speakers want to spell their own names in the way they think best and those forms get frozen (and who are we to tell someone that they misspelled their name!). Sometimes ambiguities get frozen, too, where there are two versions in popular circulation. Then, historically, German, French, and English transliteration systems differ because of the different phonetic properties of those languages, which makes for further diversity. And then colonial influence and Arabic regionalisms play a role–the way people who were colonized by the French spell their names is very different–for the same name–than the way those colonized by the English might spell theirs.

  27. Victor Mair said,

    November 8, 2018 @ 6:11 pm

    Today in my Poetry and Prose class, in a discussion of the genre known as "parallel prose" (piántǐ wén 駢體文), I said these texts use "verbal pyrotechnics", and one of the students, who is very smart and has good English, wrote in her notes "verbal parallel techniques".

  28. David Marjanović said,

    November 8, 2018 @ 6:58 pm

    Pardon my commenting again, but I just thought of the all-time best (in my opinion) creative romanization of a common Japanese word as a brand name: Mazda for マツダ, from the name 松田 (conventionally romanized Matsuda).

    Makes perfect sense in German.

  29. David Marjanović said,

    November 8, 2018 @ 7:02 pm

    re: Apparently modern Latin speakers don't like r-r in one word. Spanish torero, Italian torero, Portuguese toureiro, all from Latin taurarius, bullfighter.

    I think the difference is the morpheme boundary. The "tree" word doesn't have one.

    And yes, arbre did count, because what is now known as the French R isn't very old – it still hasn't reached parts of Québec or of Burgundy. However, -re behind a consonant is generally dropped in colloquial usage, and apparently has been for hundreds of years.

  30. Jerry Friedman said,

    November 8, 2018 @ 7:04 pm

    Athel: Consider Latin arbor (r-r), which is árbol (r-l) in Spanish and albero (l-r) in Italian. Apparently modern Latin speakers don't like r-r in one word.

    Neither did Old English speakers. 'Purple' is from 'Old English purpul, dissimilation (first recorded in Northumbrian, in Lindisfarne gospel) of purpure "purple dye, a purple garment," purpuren (adj.) "purple," a borrowing by 9c. from Latin purpura "purple color, purple-dyed cloak, purple dye," […] Purpur continued as a parallel form until 15c., and through 19c. in heraldry.

    A nice r-l exchange in Spanish is milagro from Latin miraculum.

  31. Coby Lubliner said,

    November 8, 2018 @ 8:34 pm

    Cervantes:
    1. Puerto Ricans pronounce "r" much like "l". Only syllable-final (I have heard "starter" pronounced ehtáltel). Andalusians do the opposite (juerga comes from huelga).
    2. The Iraqi family name Chalabi is جلبي (jalabi) in Arabic (which has no / t͡ʃ/); the romanization probably reflects its origin in the Turkish Çelebi.

  32. Michele Sharik Pituley said,

    November 8, 2018 @ 11:32 pm

    There is a concert hall in Gifu, Japan, which is transliterated as both Paruru Hall and Palulu Hall. So “onigilly” does not surprise me.

    @Brett – “hamantaschen” — they do, don’t they? Lol!

    Yes, Japanese curry is “A Thing”, as the kids say these days.

  33. Martha said,

    November 9, 2018 @ 1:27 am

    This reminds me of a Japanese teacher I had who told us that if we couldn't make the Japanese l/r sound, to just use l because it sounds closer to the Japanese sound than r does.

    I also agree with Max Wheeler regarding the stress, and I'm also curious about the vowel in the third syllable. My assumption is that all of the i's in onigiri would be pronounced the same, but when I see the word "onigilly," I wouldn't pronounce the second i with that same vowel.

  34. cameron said,

    November 9, 2018 @ 2:25 am

    Cory Lubliner mentions: "The Iraqi family name Chalabi is جلبي (jalabi) in Arabic (which has no / t͡ʃ/)"

    Another name that has been much in the news recently is that of the murdered journalist Jamal Khashoggi (that's the usual transliteration in the US media). The double 'g' in an Arabic name is a definite weird feature, since standard Arabic lacks /g/ (/g/ definitely exists in various versions of spoken Arabic, notably in Egyptian, where ⟨ج⟩ is pronounced as /g/, and also in many versions of Gulf Arabic, where qaf (ق) is pronounced as /g/).

    I don't speak Arabic, but I grew up speaking Persian. The name "Qâshoqchi" makes perfect sense to me. qâshoq means spoon. The suffix -chi means seller. (I'm not sure if -chi is a Turkish morpheme that's used in Persian, or a Persian morpheme that's used in Turkish – in any case it's used in both languages). As Cory Lubliner mentions with respect to the name Chalabi, in Arabic the /t͡ʃ/ often becomes a ⟨ج⟩ (which, as I mentioned above, would be pronounced /g/ by an Egyptian). Thus a qâshoqchi would be a spoon-seller, or spoonmonger.

    In Persian, qaf (ق) is pronounced as /q/.

    Jamal Khashoggi's father was of Turkish ancestry. The last name is spelled خاشقجي in Arabic. As mentioned above, the transformation of the -chi suffix into /ji/ is somewhat normal. Note that the first 'g' in he transliteration corresponds to a qaf (ق). So the double 'g' is meant to represent an Arabic qaf-jim sequence, where the jim is a substitute for Persian/Turkish /t͡ʃ/.

    But the mystery to me is why the Turkish qâshoq became خاشق in Arabic. Why'd the initial /q/ become represented as /x/ (خ)?

  35. AG said,

    November 9, 2018 @ 2:28 am

    if this is – as it appears – a kind of benign reappropriation of the R/L issue in Japanese, I'm kind of charmed by it.

    Americans have gotten racist amusement from stereotypical Asian accents of the "Flied Lice", "velly good" and "Oh, herro" sort for generations. Is it a possibly good sign that this, er, possibly good sign is kind of playing on that but not in a malicious way?

    …and the depth of Japanese love of curry is still astonishing to me. In a lot of movies, comics etc. I've encountered, it's THE nostalgic, home-cooked dish that reminds a lot of people of their youth.

  36. Athel Cornish-Bowden said,

    November 9, 2018 @ 3:22 am

    I imagine the name Vermont Curry arose because someone liked the sound.

    In Montevideo you can eat a canadiense or a chileno in a fast-food place. No connection that I could discern with Canada or Chile, and the items in question don't exist in Canada or Chile.

  37. AG said,

    November 9, 2018 @ 3:29 am

    as usual, someone on the internet has already spent a probably unhealthy amount of time looking into just this issue:

    http://justhungry.com/house-vermont-curry-mystery

  38. Athel Cornish-Bowden said,

    November 9, 2018 @ 8:04 am

    @David Marjanović (arbre).

    Many years ago (1980, I think), before I lived in France or knew much Spanish, I met a Frenchman in Chile who said that pájaro was the word he found most difficult to say.
    I found that strange, as I have always found pájaro easy enough to say. Since hearing a lot more French than I was hearing then I have realized that most French people think that a Spanish j or a Russian х is essentially a French r (we all remember the well known statesman Mireille Gorbachev): I can see that i>páraro might be hard to say. (Myself, I don't like to have u and ou too close together, especially if preceded by r, as in the address of the Pasteur Institute, La rue du Dr Roux.)

  39. Rodger C said,

    November 9, 2018 @ 9:04 am

    Gojira : Godzilla.

  40. Chris Button said,

    November 9, 2018 @ 1:36 pm

    The closest thing to a Japanese /r/ in Standard American English is /t/ when it surfaces as an alveolar flap [ɾ] (and not merged with /d/) that is not far from the Japanese post-alveolar [ɽ] (not retroflex). However, writing "Onigitti" wouldn't really capture that very well!

    The closest thing to the Japanese high pitch on the second syllable "ni" is indeed an English stress distinction since the two are often mistakenly conflated by speakers on either side. However, a transcription with "ni" in superscript and the rest in subscript probably wouldn't mean much to people either.

  41. John Swindle said,

    November 10, 2018 @ 12:20 am

    In English it would be "Originally," which is indeed stressed on the second syllable.

  42. Krogerfoot said,

    November 10, 2018 @ 7:18 am

    The second-syllable stress on onigiri is quite subtle to English speakers. I can attest that English speakers in Japan, who hear onigiri spoken by native Japanese speakers all the time, still tend to pronounce おにぎり with the stress on the third syllable. Whoever coined "Onigilly" has a keener ear for the way English speakers would handle a Japanese word than that of the people criticizing the choice on this thread, in my opinion.

  43. Chris Button said,

    November 10, 2018 @ 12:38 pm

    @ Krogerfoot

    Yes, I think japanese lexical tones are hard to distinguish for English speakers and often ignored. Although when acknowledged, the association tends to be mistakenly made with English stress. As for "onigiri", the nuclear tone on the stressed penultimate syllable "gi" in English seems to stem from applying English rules to a Japanese word since the full-voweled "o" syllable at the beginning attracts an accent as a stressed syllable such that the natural rhythmic place for the next accent (with thr nuclear tone) is on "gi".

  44. Chris Button said,

    November 10, 2018 @ 1:00 pm

    Having said that, such a pronunciation is still retaining the Japanese CV.CV syllable structure as o.ni.gi.ri since a more English sounding o.nig.ir.i could then place the sole stress and hence the nuclear tone on "nig".

  45. Nicolas said,

    November 12, 2018 @ 7:10 am

    @Athel Cornish-Bowden I suspect the issue with pájaro stemmed not so much from an approximation as /paʁaʁo/ (which shouldn't pose problem to a French native), but either from a tendency to overcorrect one of the foreign phonemes so they both end up identical (i.e. aiming for /paɾaxo/ and ending up with [paxaxo] or [paɾaɾo]) or because they kept a French voiced uvular approximant for /ɾ/ and were aiming for a voiceless velar or uvular fricative for /x/ and getting those realisations mixed up since they're allophone in their native language. (/ɣr/ clusters in Dutch are often problematic for French learners that realise the Dutch /r/ as uvular for the same reasons)

    I recently had a conversation with other (Belgian) French natives about the Spanish word reloj /relox/ and how difficult it was to pronounce, whether aiming for [χɛlɔx] or for [rɛlɔx], because of the proximity of either [χ] and [x] or [r] and [l].

  46. Internetan Gratis said,

    November 14, 2018 @ 5:46 pm

    Why is it that native speakers of Japanese have a hard time pronouncing "l"? Whenever a western word contains "L" I see that they pronounce it "ru", "ra", "ro", "ri", or "re".

  47. eli said,

    November 15, 2018 @ 12:47 am

    There's a shop with a very similar name in Nakameguro, Tokyo. Taken by me last year: https://i.imgur.com/3A4U6kT.png

RSS feed for comments on this post