« previous post | next post »

If you like Japanese food, you are undoubtedly familiar with "teppanyaki", so you probably wouldn't be surprised to see a sign like this in your neighborhood, as did Jim Breen near his home in Melbourne:

The difference is that Jim Breen is a specialist on electronic Japanese / English dictionaries, so the sign made him do a double-take. Before explaining why the sign struck Jim as particularly odd, I should point out that the name "teppanyaki" consists of three elements meaning literally "iron-board-grilled/broiled/pan-fried", i.e., an iron griddle or grill for cooking food.

What we see on this sign are the following three characters: 鐡板焼, though the top right portion of the last character as represented in the font I'm using here is semi-abbreviated; the full (traditional) form, as in the photograph, would look like this: 燒. In Chinese simplified characters the entire name of this type of cooking would appear thus: 铁板烧. In Mandarin, whether written 鐡板燒 or 铁板烧, that would be pronounced tiěbǎnshāo.

In Japanese, the name for this style of cooking — which is, after all, a Japanese type of cuisine — is written 鉄板焼き or just 鉄板焼 without the final kana. in Japan, it would be extremely rare (if ever) to find the 鐡 kanji used to write the first syllable of teppanyaki. 鐡, along with 銕, 鐵 and 鋨 are known in Japan as variants (itaiji 異体字) of 鉄, but they are really only in the kanji standards because they appear sometimes in names such as 鐡男 (Tetsuo) and 鐡穴 (Kanna).

It would seem that what has happened with that Teppanyaki restaurant in Melbourne is that its proprietors (who are Chinese-Australian) simply went to their regular signage supplier and specified what was (for them) the way of writing Teppanyaki in characters, without considering that the result would be quite un-Japanese. If you Google on 鉄板焼(き) you'll get about 30 million hits. 鐡板焼 gets around 60, most of which are referring to establishments in Taiwan, etc.

I would only add that most sushi shops and Japanese restaurants I go to outside of Japan are operated by Chinese. And last week when I want to the Philadelphia Flower Show, all the people working in the big bonsai shop were speaking Chinese!



  1. MonkeyBoy said,

    March 17, 2013 @ 11:02 am

    "most sushi shops and Japanese restaurants I go to outside of Japan are operated by Chinese."

    In the US I've noticed a good number run by Koreans and Korean run Chinese restaurants but I have no idea what the proportion is.

  2. Meena Vathyam said,

    March 17, 2013 @ 11:06 am

    Now, in Shanghai, too there are Teppanyaki and Yakkaniku restaurants. Interesting to read on the origin & interpretation in other places. As for Bonsai, don't the Chinese claim it was their idea all along?

  3. Victor Mair said,

    March 17, 2013 @ 11:09 am


    Yes, lots of Koreans operating Japanese sushi shops and restaurants too. Don't see many Japanese.

  4. Bruce Rusk said,

    March 17, 2013 @ 11:30 am

    The look of the font is very Chinese as well.

  5. Rod Johnson said,

    March 17, 2013 @ 11:32 am

    Get out! Next you'll be telling me Taco Bell's not run by Mexicans.

  6. MonkeyBoy said,

    March 17, 2013 @ 11:49 am

    "Yes, lots of Koreans operating Japanese sushi shops and restaurants too. Don't see many Japanese."

    In big US cities (LA, NYC) some of the "better" Japanese restaurants are Japanese run. I would think a visiting Japanese executive would be offended if taken to a non Japanese run place. I remember one place in Marina Del Ray that had a separate karaoke room which showed large screen videos of topless Japanese women singing traditional songs that the patrons could sing along with.

    Still, most non-Japanese owned places I've been to employ Japanese chefs. All my local grocery stores that have sushi sections import them.

  7. Victor Mair said,

    March 17, 2013 @ 12:12 pm

    From Bob Ramsey:

    Yes, it's true. Just about everywhere in the US most Japanese restaurants are run by Chinese or Koreans or some other ethnicity besides Japanese. Our local Chinese restaurant also has a sushi bar near the entrance—sometimes the itamae are even Latinos! One big exception, though, is southern Manhattan, where you see young Japanese—and I emphasize young—all over, working in noodle places, serving ramen, or just walking around the streets in pairs chattering in Japanese. Some places on St. Mark's in the East Village look like they were lifted out of Tokyo. Many Japanese, it would seem, are drawn to Manhattan much more than to other places in America.

  8. Yolo9 said,

    March 17, 2013 @ 12:47 pm

    Los Angeles has a much larger Japanese enclave than New York. There are even Korean restaurants run by Japanese.

  9. Victor Mair said,

    March 17, 2013 @ 1:39 pm

    @Rod Johnson

    Except that Taco Bell is a huge international chain based in America:


    What we're talking about here are primarily individual restaurants and shops.

  10. Victor Mair said,

    March 17, 2013 @ 1:56 pm

    From Cecilia Segawa Seigle:

    I discovered something interesting, thanks to your essay on Teppanyaki. I knew the more complicated 鐵 is considered the old Kanji (旧字), and 鉄 is a new kanji (新漢字 or 当用漢字). So I looked up when this occurred – I vaguely knew it was shortly after the end of the WWII.

    My interesting discovery is that actually the designation of 鉄 for 鐵 predated 1946, when 当用漢字 was officially designated.

    It happened in 1940 February 29 when a limited number of kanji was designated by the Japanese Army for the Arms for the military. The list was called Heiki meisho-yo seigen kanji hyo (兵器 名称用制限漢字表), and included 1235 characters. The breakdown was: 959 kanji from the level 1, which Japanese children learned by the time they were in the 4th grade; plus 276 kanji required for the names of the Arms (level 2). In the list of 1235 kanji included the new kanji for IRON鉄 for 鐵. The kanji iron 鐵 was too complicated and time-consuming for writing, so any arms that required the kanji for iron was to be written by the new character 鉄. 鐵 was not to be used.

    In 1942 (still during the WWII), the Council for the National Language (国語審議会 ) submitted a recommendation for the new Standard Kanji List (標準漢字表) to the Minister of Education (文部大臣) for the use of government agencies as well as the general public. (Prior to this there had been much criticism and recommendations for limiting Kanji for decades, including the entire abolishment of Kanji).

    The new list was commonly used kanji of 1134, plus quasi-standard kanji 1320, plus special kanji 74, the total of 2528 kanji. This list included the new kanji「鉄」but also listed the old kanji in this form: 「鉄(鐵)」. The recommendation of 国語審議会 was the use of the new kanji only.

    The Post-War decision of the 国語審議会 on November 16, 1946 was to limit the total list of commonly used kanji to 1850, and the decision of not using the old 鐵 was definite. Therefore, the official use of 鉄 for iron was legal and permanent. Yet, strangely, the listing of 「鉄(鐵)」was included, merely for a reference.

    On January 1, 1948, when the Permanent Domicile Law was changed, the list of the kanji that could be used for naming babies was announced. They were limited to the 1850 kanji of 当用漢字. 「鉄(鐵)」was included, but the by law the use of the old kanji鐵 was prohibited. For some reason, 鐵 was added in parentheses always as a reference.

    My guess is that the people were so un-used to the 鉄 in the beginning, they didn’t know what it meant. But of course by today everyone knows 鉄 and no one uses 鐵. So, it is very strange that the restaurant used 鐡板焼.

  11. Lazar said,

    March 17, 2013 @ 3:07 pm

    @Bruce Rusk: That's an interesting observation. Disregarding the issue of traditional vs. simplified vs. shinjitai, what are the things that make a font look more Chinese or Japanese?

  12. julie lee said,

    March 17, 2013 @ 3:23 pm

    While many of the sushi shops here in the San Francisco Bay Area are owned and run by Chinese, I notice (with amusement) that many of the people who make Chinese dumplings (baozi包子,etc.) in Chinese supermarkets here are Mexicans or Latinos.

  13. Victor Mair said,

    March 17, 2013 @ 3:44 pm

    @Meena Vathyam

    As with many other imports from abroad (ikebana, origami, the tea ceremony, etc.), bonsai 盆栽 (lit., "basin planting") in Japan developed in a quite different way from 盆景 ("basin scenery") in China.

  14. MonkeyBoy said,

    March 17, 2013 @ 4:26 pm

    A few observations on non-chain stores in the US.

    Japanese food stores are run by Japanese, Chinese by Chinese, while Asian food store are mainly run by Koreans, and Mid-Eastern food stores are run by Lebanese Christians.

    Italian restaurants are run by Italians except the lower tier (sandwiches and pizza) are often run by Greeks and more recently Turks.

    Philly Cheese Steak places located far from Philly and north on the East Coast are often run by Armenians.

  15. Chris Bilson said,

    March 17, 2013 @ 4:48 pm

    Having lived in Taiwan for several years and spent a grand total of around a month in Japan, I would have to say that Japanese food in Taiwan is a lot more like what you see in the United States and other countries I have been too (and better, to my taste) than the equivalent things in Japan (speaking of sushi, sashimi, udon, that type of thing.) A lot of those restauranteurs have migrated to western countries because they know how to make Japanese food that people want there. I've probably eaten at 50 "Japanese" restaurants in California and I would say > 50% are run by Taiwanese. Maybe not authentic Japanese, but very successful in giving people what they want.

    On the other hand, Japan has, by far, the best steak restaurants on the planet. You haven't really had steak until you've had it in a really good Japanese steak restaurant. I wish some of those guys would set up shop near me. Then again, I guess, for all the talk, American's really don't value good steak as much as Japanese do.

    in a larger sense though, the idea that a particular style of food is somehow only meant to be produced by a certain ethnicity or race is pretty laughable. Various populations have differing average preferences for styles and types of food and the types and styles you see being consumed by those populations is almost entirely determined by the consumers. Where the person who steps in to fill the demand was born, where their grandparents were born, etc, is almost entirely irrelevant. I've seen this in action many times – like when I start speaking mandarin to the boss of a really good Japanese restaurant and my friends know I don't speak Japanese, they look so sad and disappointed – but I still think it's really funny. Why fixate on some naive notion of what's authentic? Just enjoy food that's good and avoid food that's bad.

  16. B.Ma said,

    March 17, 2013 @ 5:34 pm

    I'm not sure I understand everybody's consternation. It was quite clear to me that the restaurant is a Chinese restaurant which serves Japanese-style food, and the characters on the sign are written and meant to be read in Chinese (probably Cantonese).

    I don't think you would find any Japanese patrons there.

  17. Victor Mair said,

    March 17, 2013 @ 5:37 pm

    @B. Ma

    But then why does it say "Teppanyaki" on the sign in Roman letters? And why is there a banner in the background with what is unmistakably Japanese writing?

  18. Carl said,

    March 17, 2013 @ 6:52 pm

    In Seoul, I often saw signs with Japanese written using a Chinese font or vice versa.

    Here in Honolulu, there are many restaurants that say "Yakiniku" in Roman letters, Japanese, and Hangul (I think they are probably run by Korean Americans).

  19. Jim Breen said,

    March 17, 2013 @ 7:02 pm

    @Cecilia Segawa Seigle

    Thanks for that analysis. All my pre-WWII dictionaries use one of the 旧字; usually 鐡 for that kanji.

    The restaurant in question does pretty good teppanaki. The head chef trained in Japan, etc. Quite a few of the Japanese restaurants around are best regarded as Japanese-themed Chinese restaurants, but there are also plenty owned and staffed by Japanese.

  20. julie lee said,

    March 17, 2013 @ 7:26 pm

    @Rod Johnson

    I was delighted to find a small Thai restaurant near my home and was disconcerted and disappointed when I heard the owner and waiter speak Cantonese. I had the same experience earlier in a Korean restaurant. The pad thai and bi bim bop, etc. were fine though.

  21. Rod Johnson said,

    March 17, 2013 @ 9:14 pm

    Julie–did you mean that to be directed at me?

    (For the record, I know Taco Bell is American, a chain, etc. I was just kidding.)

  22. julie lee said,

    March 17, 2013 @ 10:14 pm

    Actually my comment was directed at the general topic of the ethnic identities of the people who run our ethnic eateries. I knew you were kidding about Taco Bell. The Chinese have a lot of cooking know-how and business know-how too, so we shouldn't be surprised when they capitalize on their forte.

  23. Jean-Michel said,

    March 17, 2013 @ 11:00 pm

    Disregarding the issue of traditional vs. simplified vs. shinjitai, what are the things that make a font look more Chinese or Japanese?

    If I'm not mistaken, the top stroke of 反 (in 板) usually appears, at least in printed Japanese, as a straight horizontal line. I don't see any other tipoffs there, though there probably are some. Speaking generally, there are a number of features typical of Japanese but uncommon in Chinese, like the use of 糸 for the left-hand "silk" radical (糹/纟). Japanese also seems more fond of using the "simple" forms of the "water" and "ice" radicals (氵/冫), whereas printed Chinese usually sticks with the more complex forms. (This difference isn't black and white: some Japanese fonts use the "complex" forms and some Chinese fonts also use the "simple" forms.) It appears to me that Chinese fonts more commonly preserve features of brushwritten characters that Japanese fonts often ignore.

  24. Jean-Michel said,

    March 17, 2013 @ 11:32 pm

    A more succinct summary of what I wrote above might be that Song/Ming typefaces are less common in Japanese than in Chinese; Wiki says that they're the most popular fonts for Japanese as well, but that probably refers more to running text in books/magazines/etc. than the fonts used for signage and logos and such.

  25. Matt said,

    March 18, 2013 @ 12:25 am

    Therefore, the official use of 鉄 for iron was legal and permanent. Yet, strangely, the listing of 「鉄(鐵)」was included, merely for a reference.

    Wait, what's the strange part there? All of those kanji announcements include the designated "old form" after any character that has one, even if there's no direct grapho-etymological relationship.

  26. Akito said,

    March 18, 2013 @ 4:59 am

    The simplified form (鉄) reminds one of losing (失) money (金), so some businesses stick to the old form (I don't know how to post a picture, but you can look at Japanese steel makers' logos to confirm this). There is also a bizarre form combining 金 with 矢.

  27. Vanya said,

    March 18, 2013 @ 6:25 am

    The last Sushi place I visited in the US that was actually owned and staffed by Japanese nationals was in, of all places, Wilkes-Barre, PA.

  28. Vanya said,

    March 18, 2013 @ 6:33 am

    Of course, Japanese food without Japanese people is not unusual. Pizza places in my suburban Massachusetts neighborhood seem to be owned by any Mediterranean nationality except Italian – Greek, Lebanese, Egyptian, Bosnian, etc. But invariably the walls are decorated with photos of the Colosseum, Venice, Florence, etc.

  29. Rod Johnson said,

    March 18, 2013 @ 11:14 am

    As Franz Boas taught us, "race, language and culture" are essentially independent variables, so it shouldn't be a surprise to find that cuisine (=culture) and ethnicity (="race") tend to become increasingly decoupled over time. It's really only the recency of the spread of Asian and other local cuisines that leads us to expect Japanese restaurants to be run by Japanese.

    I grew up eating (mostly shitty) naively Americanized versions of Mexican, Chinese, Italian, Greek, French and Indian food alongside my native "American" (which included such good old American delicacies as Spanish rice, goulash, macaroni, "French bread," and corned beef and cabbage). Some of them emphasized the signifiers of ethnicity (Chef Boy-Ar-Dee, sombreros, "Chinese" typefaces) more than others. It may be that the presence of "Asians" of whatever sort in Asian restaurants is such a signifier to Americans. The tension between assimilation and "authenticity" is no less evident here than in many other areas of American life.

    (Sorry to be emphasizing "American" here–I just don't feel competent to comment on how this plays out in other countries.)

  30. julie lee said,

    March 18, 2013 @ 12:09 pm

    @Rod Johnson:
    "cuisine (=culture) and ethnicity (="race") tend to become increasingly decoupled over time"

    This decoupling happens in other areas too of course. A New York Times article recently said that college admissions people in top schools are now classifying piano and violin as "Asian instruments" because so many Asian college applicants are accomplished in these instruments. I see this classification as subtle or not-so-subtle discrimination against Asian applicants. Squash used to be a white sport. A big Wall Street Journal article just a few days ago said a lot of mothers are very intense about getting their children into the top colleges through squash. The WSJ article featured a large picture of what looked like an Indian playing squash. (Yes, squash is very big among Indians, and increasingly among Chinese). That article, to me, was just one step away from calling squash "an Asian sport". You can call the symphony orchestra "an Asian thing" here in the Bay Area because the youth orchestras of the East Bay and West Bay (youth orchestras of Cupertino, San Jose, Oakland, Berkeley, San Francisco, etc.) are conspicuously preponderately Asian. They are huge orchestras. Marin County north of San Francisco is very white and its youth orchestra is, by contrast, very small, and preponderately white. I asked my daughter why it was so small, occupying only half the stage. She said white children tend to go into bands not symphony orchestras. This was all new to me. I hope I'm not being racist by mentioning it, and I hope I haven't ruffled any feathers or put anyone's nose out of joint.

  31. Ellen K. said,

    March 18, 2013 @ 1:45 pm

    Julie Lee, Indian = people from India or Native Americans? I can't tell from your post and I'm curious.

  32. julie lee said,

    March 18, 2013 @ 2:34 pm

    Indians from India. I realize that my comment just now and my reaction to the WSJ article (of which I read only the first few paragraphs) were somewhat peevish because it made me think of "Asian instruments". Yes, the big annual Bach Festival of young musicians (mostly piano and violin) in Berkeley struck me as predominantly Asian. I guess you'd call Bach and Beethoven "Asian composers" now.
    I guess young non-Asian Americans tend to go for pop, jazz and rock.

  33. J.W. Brewer said,

    March 18, 2013 @ 5:44 pm

    I have no idea of the ethnicity of the actual owners (who I may or may not ever have seen) or showman/chefs (who are all some sort of East Asian but I haven't inquired as to Japanese v. Chinese v. Korean) of our local suburban teppanyaki place, but I can report that its name is universally mispronounced by the white/black/Hispanic clientele (it's called "Edo" but pronounced like a hypothetical Japanese proper noun "Ido"). Of course, if that's what the local non-Japanese customers have in their ignorance of standard transliteration conventions come to call it, it makes sense for the owners/employees (regardless of whether or not they know Japanese or are of Japanese origin themselves) to play along rather than be all prescriptivist. After a few years, I altered my own pronunciation to accord with local custom, accepting that there's no reason "Edo" has to be pronounced uniformly regardless of context, just as Houston St. in NYC need not be pronounced same as Houston, Tex. (Just don't get me started on the idiosyncratic tri-syllabic pronunciation of "Tokyo" in that Deep Purple song.)

  34. Chris C. said,

    March 18, 2013 @ 6:36 pm

    I wanted to mention that, whenever I've felt the desire to fake knowing Japanese in online discussions, I've found Jim Breen's website to be an invaluable reference.

    We have a family in my area — the San Lorenzo Valley of Santa Cruz County, California — which owns 1) a Chinese restaurant in the valley's southernmost town, 2) a newly-opened Japanese restaurant in the valley's northernmost town, and 3) a "Japanese" import shop just over the hill in San Jose, although I suspect most of the items are of Chinese manufacture for the Japanese market. Having never inquired after their family name, and with no understanding of the chatter I hear coming out of the kitchen, I have absolutely no idea what their actual ethnicity is.

  35. Anthony said,

    March 18, 2013 @ 9:52 pm

    At least in the United States, running a restaurant is a hard, miserable business compared to other jobs which might pay as well. A friend of mine noted that we're losing Polish, Russian, etc., restaurants in California, because recent immigrants from those countries aren't desperate enough to try running a restaurant. This probably explains the dearth of Japanese restaurants run by actual Japanese, except at the very high end – most ethnic Japanese in California are now very assimilated, and end up mostly being salaried professionals, while most Japanese immigrants these days end up in Silicon Valley, making far more than restauranteurs might.

  36. Victor Mair said,

    March 19, 2013 @ 7:24 am

    The correct standard Japanese pronunciation for 鉄板 is teppan. Although I've studied a fair amount of Japanese, I've never given much thought to the rather radical sound changes that occur when individual kanji are combined into words. The pronunciation of 鉄 by itself is tetsu or kurogane. The pronunciation of 板 by itself is han, ban, or ita. Since kurogane and ita are indigenous pronunciations, while teppan 鉄板 is based on borrowings of Sinitic sounds for the two kanji, for teppan we are only concerned with tetsu and han / ban.

    It is rather remarkable that tetsu + han / ban = teppan. What I'm wondering is the extent to which such changes exist in the language as a whole. I'm sure that Japanese phonologists have these things worked out in enormous detail, but would like to directed to at least a summary account of how extensive they are.

    I remember once, when I was doing research for a book review of a dictionary of Korean Buddhist texts, that the sound changes that occur when various hanja are combined are also wide-ranging. It was especially frustrating that different authorities argued over the best way to represent them in Romanization.

    For Mandarin, the rules governing sound changes that occur when various hanzi are brought together are much simpler and basically have to do with tone sandhi. At Harvard, Rulan (Iris) Chao Pian had a handbook for the pronunciation of Mandarin that was quite long (I think that it was based on the work of her father, Y[uen] R[en] Chao), and many pages were devoted to the explication of rather arcane sequences of tones. One that is not too complicated may be illustrated by my own name, Méi Wéihéng 梅惟恆 (with three second tones), which is actually pronounced as Méi Wēihéng (2-1-2).

    Usually, however, Mandarin tone sandhi rules are reduced to just three or four.



  37. Cecilia Segawa Seigle said,

    March 19, 2013 @ 9:04 am

    To Akito,
    You are right! Old-fashioned merchants are very superstitious, and they would hate losing (失) money (金) – so they might prefer using the old kanji. I never thought of the make up of the character. Thank you.

    To Matt,
    I thought it was strange to persist on listing the character that had become illegal, right behind the newly designated character. If you are not permitted to use it, why keep listing it? In the early days of changes they probably did it because people didn't know the meaning of the new kanji 鉄. But they kept listing the old kanji for a loooong time. That is what was strange to me.

  38. chris said,

    March 19, 2013 @ 9:48 am

    It is rather remarkable that tetsu + han / ban = teppan.

    Is it? I thought I remembered hearing or reading somewhere that "tsu" was frequently elided, and surely "b" isn't that far from "p", phonetically. If you assume "tebban" as an intermediate stage, it doesn't seem that unusual as a pronunciation shift, to me. Although I'm certainly no expert.

    Also, this isn't relevant to the main point of the post, but is it normal in Australia for a phone number to have 8 digits?

  39. Mark Dunan said,

    March 19, 2013 @ 9:58 am

    @Chris, most of Australia went to two-digit area codes and eight-digit phone numbers for land lines in the 1990s. I agree, though; seeing those eight digits run together like that certainly contributes to the Chinese-like look!

  40. Mr Punch said,

    March 19, 2013 @ 10:30 am

    My father, an immigrant to the US from France, propounded a "French restaurant theory" of social mobility. Noting that in the US, except at the very high end, restaurants serving French food were often run by Italians, pizza places by Greeks, Greek restaurants by Lebanese, etc., he suggested that later immigrant groups tend to assimilate by association with earlier-arriving nationalities with whom they share some cultural traits.

  41. Victor Mair said,

    March 19, 2013 @ 10:56 am

    I've seen so many Greek pizza places that at one point I almost got the idea that pizza was a Greek thing, or at least that there was a decidedly Greek variant of pizza.

  42. ardaleth said,

    March 19, 2013 @ 11:16 am

    I can't point to any book or site that discusses the issue, but if I remember correctly the tetsu + han = teppan equation can be easily explained diachronically.

    Final -tsu in on-yomi reflects an original Chinese reading ending in -t. Middle Chinese for 鉄 is *thet, so the Japanese was probably something like *tet with the final vowel added to comply with native phonotactics.

    On the other hand, Old Japanese /p/ > Modern Japanese /h/, and given that the Middle Chinese for 板 is *paen I would expect the original Japanese reading to be something like *pan.

    Now *tet + *pan = teppan is not such a big leap!

    In fact this -chi/-tsu/-ku + stop/p > h = geminated stop assimilation seems pretty common in Sino-Japanese e.g. Nippon, Hokkaido.

    Just my two cents, I'm no expert so feel free to correct me.

  43. Akito said,

    March 19, 2013 @ 11:45 am

    The underlying form of "tetsu" is "tet". Early Japanese learners of Chinese couldn't pronounce a stop consonant without a vowel support, so they added "u" (sometimes "i") to make the second "t" audible, the high vowel affricating the preceding "t" at the same time. (Mandarin speakers dropped the final stop. Koreans often add "eu" to English words ending in a stop consonant.) The added syllable was not necessary for compounds, so we have "tetpan", which by assimilation slips into "teppan". Simple.

    Why "pan" and not "han"? Modern Japanese "h" evolved out of bilabial "f", which (along with "w") is thought to have developed from Nara-period "p". We have been de-bilabializing. (One joke says that "haha" (mother) was originally "papa".) Deep in the brain we still remember the form with "p" as a combining form. When the rule for voicing applies, "p" is realized as "b", as in "kanban". Simple.

  44. MonkeyBoy said,

    March 19, 2013 @ 12:32 pm

    "a decidedly Greek variant of pizza"

    The Greek Pizza cooking variation has long been recognized – rather than cooking the pizzas directly on the floor of an oven (traditional Italian) they are cooked in oiled pans. Greek made pizza also tends to have a spicier sauce and maybe a higher sauce to cheese ratio.

    Both of these techniques involve traditional pizza ovens. Most new places use conveyor belt ovens which offend many pizza purists, but then again most pizza consumers are hardly knowledgeable in the various traditional styles.

    A Turkish neighbor I had really didn't know or care about traditional styles or methods and mostly like to point out how much better his conveyor belts were than ovens in churning out pizzas.

    Maybe people from outside a tradition are better at adopting innovations though purists may claim this degrades quality.

  45. Victor Mair said,

    March 19, 2013 @ 4:16 pm

    From Bob Ramsey:

    I don’t know who else you’ve asked about this Japanese sound change, but you can find discussions of it lots of places, including in language textbooks (which can sometimes be the better sources of information than academic papers). In any case, in dealing with internal reconstruction in my course on the history of Japanese, I usually summarize information from Martin’s book, The Japanese Language Through Time (Yale). I’m pasting below some notes I use taken from that source.

    The part you’ve explicitly asked about is in bold script. You can see, I think, that what you’re dealing with is a common and regular feature of Japanese, and one that has resulted from the historical change p > h at the beginning of a word (or, in the case of Sino-Japanese, of a morpheme). (I might add parenthetically that a /p/ in Old Japanese occurring in medial, intervocalic position was different; there it developed into -w-, and that’s the source of the irregular conjugation of, for example, kau ‘to buy’, so that you get, e.g., kaimasu but a /w/ in kawanai. You probably already know these things, but I thought I’d pass them along anyway, just in case.)


    One of the more peculiar aspects of the modern Japanese sound system is the relationship between h (which is a velar consonant, pronounced in the back of the mouth) and the labial consonants p and b. For reasons that we will explore, b behaves as the voiced equivalent of h.

    We begin by looking at what the Japanese call “sequential voicing” (rendaku 連濁). Here is an example of sequential voicing: The word kami ‘paper’ begins with the voiceless consonant k. But when it appears in a compound word such as oriĝami, ‘folding paper’ (ori- ‘folding’ + kami), the k gains voice and becomes ĝ. The word teĝami is compound of te ‘hand’ plus kami. Similarly, toki ‘time’ + toki ‘time’ becomes tokidoki ‘some­times’. Here are a few more examples:

    Basic word Word in compound

    kuti ‘mouth’ deĝuti ‘exit’ (also, iriĝuti; kaisatuĝuti, etc.)

    kaeru ‘change’ kiĝaeru ‘change clothing’

    kawa ‘river’ Sumida-ĝawa ‘the River Sumida’

    sakana ‘fish’ yaki-zakana ‘broiled fish’

    sima ‘island’ Ioo-zima ‘Iwojima’

    tana ‘shelf’ hondana ‘bookshelf’

    tera ‘temple’ kiyomizu-dera ‘Kiyomizu Temple’

    tokee ‘clock’ mezamasi-dokee ‘alarm clock’

    tonari ‘next door’ hidaridonari ‘next door on the left’

    Not all compound words work this way, but sequential voicing is common enough to be recognized by most elementary students of the language. The process is regular: Add the feature of voice to a voiceless consonant, which is otherwise pronounced the same way in the same position. Thus, k becomes g; s becomes z; t becomes d.

    By this same process of sequential voicing, h becomes b:

    Basic word Word in compound

    hasi ‘chopsticks’ wari-basi ‘disposable chopsticks’

    hasi ‘bridge’ nihon-basi ‘Japan Bridge’

    hana ‘flower’ ikebana ‘(the art of) flower arranging’

    huton ‘futon’ sikibuton ‘bottom quilt’

    What is curious is that, unlike the other pairs of consonants (k and g, for example), h and b are two consonants that have almost no features in common. They are pronounced at opposite ends of the vocal tract. A lot more than sequential voicing relates h to p and b. Here is a look at some of the variety:

    h p b

    Like the velars, the labials form reduplicative mimetic sets.

    hata-hata pata-pata bata-bata (fluttering)

    huka-huka ‘fluffy’ puka-puka ‘lightly’ buka-buka ‘baggy’

    h b


    haya-baya huka-buka hoso-boso hito-bito

    h p

    Many different processes relate h with p.

    (1) Sino-Japanese compounds. If the Sino-Japanese pronunciation (the on reading) of a Chinese character begins with h, that h will become p following the mora nasal, or tu or ti; thus iti ‘one’ + hun ‘minute’ gives ip-pun. (The tu or ti becomes a mora consonant.)

    -hun ‘minute’ ip-pun ‘one minute’ san-pun ‘three minutes’

    -hatu ‘departure’ syuppatu ‘depart’ (syutu + hatu)

    hookoku ‘report’ koohoo ‘official report’ denpoo ‘telegram’

    (2) Intensification. Some words can be pronounced emphatically by doubling a medial consonant. Thus totemo pronounced emphatically is tottemo; amari becomes anmari. But if the medial consonant that is doubled is h, the h becomes p:

    yahari yappari

    yohodo yoppodo

    — karappeta (from kara ‘empty’ + heta ‘unskilled’)

    kuihagure kuippagure

    (3) The intensive prefix ma(t)-. This prefix turns kuroi ‘black’ into makkuroi ‘jet black’; siroi ‘white’ into massiroi ‘snow white’; taira ‘flat’ into mattaira; etc. When it is combined with words that begin an h, it changes the h into p.

    mapputatu ‘just exactly in two’ (hutatu ‘two’)

    mappadaka ‘stark naked’ (hadaka ‘naked’)

    mappiruma ‘broad daylight’ (hiruma ‘daytime’)

    (4) The prefix ko(t)- ‘small’.

    koppidoku ‘a little harshly’ (hidoku ‘harshly’)

    (5) Related words.

    ha ‘leaf’ happa ‘leaf’ (colloquial)

    hayai ‘quick’ kenkappayai ‘quick to quarrel’ (kenka ‘quarrel’)

    hiroi ‘wide’ dadappiroi ‘unduly wide’

    hana ‘nose’ sisippana ~ sisibana ‘pug nose’

    hara ‘stomach’ karappara, sukippara ‘empty stomach’

    — yoppite ‘all night long’ (from yoppitoi, which is said to be from yo hito-yo ‘one night of night’

    hikaru ‘shine’ pika-pika hikaru ‘flashes with a flash’

    (6) Contractions.

    Phrases like omo ni miru ‘see (as) important’ contract to omonmiru ‘reflect carefully’, and omo ni suru ‘make important’ to omonziru ‘honor, respect’. In the same way, omo ni hakaru ‘estimate (as) important’ contracts to omonpakaru ‘consider’.

    More examples of verbal contractions:

    hiki- ‘pull’ + haru ‘stretch’ → hipparu ‘pull’

    + hagu ‘strip off’ → hippagu ‘strip’

    + hataku ‘slap’→ hippataku ‘slap’

    oi- ‘shoo away’ + hazimeru ‘begin’ → oppazimeru ‘begin’

    + hooridasu ‘throw out’ → opporidasu ‘throw out’

    yoi- ‘get drunk’ + harau ‘clear away’ → yopparau ‘get drunk’

    (7) The suffix -ppanasi ‘(leaving something) just as it is’. This suffix is derived from the verbal forms hanasi ‘leaving it’.


    tuke-ppanasi ‘leave turned on’

    ake-ppanasi ‘leave open’

    koware-ppanasi ‘leave unrepaired’

    oki-ppanasi ‘leave lying somewhere’

    tati-ppanasi ‘remain standing’

    (hon o) kari-ppanasi ‘(books are) unreturned’

  46. Victor Mair said,

    March 19, 2013 @ 6:41 pm

    From Hiroko Sherry:

    For Kanji compound words:

    For general Japanese pronunciation rules:

  47. Matt said,

    March 20, 2013 @ 7:57 pm

    Hi Cecilia,

    Thanks for the reply! Well, like I said, it's standard to list all "old kanji" in parentheses in the "official kanji" announcements, so it's not like 鉄 is getting special treatment. If you look at the latest Joyo Kanji listing (available via here, updated 2010), you'll see that it still has entries like "温 (溫)" and ”担 (擔)" in the same format as "鉄 (鐵)".

    The underlying form of "tetsu" is "tet". Early Japanese learners of Chinese couldn't pronounce a stop consonant without a vowel support, so they added "u" (sometimes "i") to make the second "t" audible, the high vowel affricating the preceding "t" at the same time.

    Although I agree that the underlying form of "testsu" is "tet", the rest of this paragraph is not necessarily true! It depends on your definition of "early", but if we look at the early 17th-C Jesuit dictionary, we see that the Japanese speakers of the time seem to actually have been better at pronouncing stop consonants than modern-day speakers.

    It is difficult to say for sure because of course Japanese orthography (in its traditional incarnation) has no way of unambiguously representing a stop consonant other than /n/, but for example many words ending in つ are represented with a final <t>. 金鉄 is in there as <Qintet>; 火鉄 is in there as <Quatet>. 親切 is <Xinxet> and 愛別 is <Aibet>. 往日の is spelt <Vǒjitno> — not even an epenthetic vowel!

    On the other hand, native Japanese words have the final <u> (actually, <çu>): 泡が立つ is <Auaga tatçu>; あいつ and かいつ are <Aitçu> and <Caitçu> respectively.

    Some of this has filtered down to the present day, for example in certain sound changes in Ryukyuan languages, and specialized set phrases on the mainland. Traditional theater's こんにった as a pronunciation for 今日は must surely derive from /koN.nit.pa/, not /koN.niti.pa/ (discussion of what the /p/ was at that point carefully avoided) — although the fact that 日 is pretty regularly <nichi> in the Jesuit dictionary complicates things a bit…

    There's room for argument over how final <t> was actually pronounced, but I am given to understand that (for example) close examination of how the various final つs were written in kana at the time reveals systematic distinctions there too. (I think Frellesvig covers this, don't have the reference handy, sorry.)

    So given that 鉄板 has been in the Japanese language since at least the days of the Jesuits (in fact, it's in their dictionary, as <Teppan>!), it seems overwhelmingly likely that when the word /teppan/ was coined, the first syllable was simply /tet/, both underlyingly and on the surface. The change from /h/ to /p/ is explainable as Akito says: /h/ was originally a bilabial plosive and it was still en route to its current pronunciation back in the day.

  48. Manabu said,

    March 21, 2013 @ 9:09 am

    I am a Japanese citizen and I have focused on cooking French and California Fusion here in San Francisco. I also do not speak French. Many of my colleagues in the area are Japanese and successfully making French cuisine at their homes and running world-class restaurants.

    It never occurred to me that the my nationality got anything to do with what I am allowed to cook or offend anyone who enjoy my creation in the kitchen. We could also make grammatical or spelling error in naming our restaurants or menu items.

    The best sushi chef in the world, according to a recent competition is a Swedish person.

    Japanese food can be cooked and its art further developed by people born from all nations.

  49. Akito said,

    March 21, 2013 @ 11:29 pm

    Thank you for looking some words up in the Jesuit dictionary. I can only think there was variation in how the syllable-final stops were manifested (or not manifested) in Japanese accent. It is plausible that (near-)original pronunciations were retained in some vocabulary among, for example, the clergy in Buddhist temples. At the same time, I wonder how many of such learned words were within the reach of the common folk. It may be relevant to ask whose pronunciation the Jesuits were recording.

    The cannon of the CV structure is so strong that one pharmacy in my childhood neighborhood had a sign with the furigana やくきょく for 薬局. The publishing house 小学館 still throws in a "u" after the first "k" in their English name.

RSS feed for comments on this post

Leave a Comment