Comparing to Chinese (Part 2): Tones

I was glancing through a thread about low and high tones on my forum and it made me realize that we don’t treat tones with as much care as we should in Japanese (ie, virtually none). For example, if I were to describe it in Chinese tones, you really do need to pronounce 日本 with something similar to a second and fourth tone. In contrast, 二本 is more like a fourth and neutral tone. And this really could potentially be an issue. What if you said 日本ください instead of 二本ください? Now you’re asking for Japan instead of two bottles! What a わがまま!

Personally, I’ve had times when I would ask somebody about a new word I just learned and the person would have no idea what I was talking about. Then I’d write the word and he/she would say, “Oh you mean [X]!” and pronounce the word exactly the same way but with different pitches. See, without context you really do need to get the tones right.

And sure, context will cover your ass and prevent any mishaps most of the time but is Chinese any different? You know in Rush Hour 2 when Chris Tucker attempts to speak Chinese? It was hilarious but in real life, if you messed up all the tones, it just becomes gibberish. There are a few insidious homophones like eyeglasses vs eyes: 眼睛(yǎnjīng) / 眼鏡(yǎnjìng), but overall context should take care of one or two mistakes. I’ll have to watch that movie again now that I know some Chinese to see if they were really clever enough to teach Chris the wrong tones correctly to actually say the unintentional but hilarious lines.

Chinese has always had a notorious reputation of being insanely difficult due to the tones but I actually think Japanese is more difficult. With Chinese, at least all the tones are laid out and stay (mostly) the same. In contrast, Japanese really has no rules for pronouncing words with the correct pitch and it would probably change anyway depending on how you’re using it. Unlike Chinese, you’ll probably be understandable even with all the wrong tones, but you will still sound foreign and may even be difficult to understand.

We really should start thinking about patterns in Japanese tones and how we could effectively teach students how to pronounce things correctly not just phonetically but on the tonal level. For example, I’ve noticed that long vowels are often a high and flat tone (first tone in Chinese). Just listen to how train announcers pronounce 東京. (Tones are more clearly enunciated in formal settings like announcements and news broadcasts.) I’m sure by just practicing the long vowel sounds in this manner, you can significantly improve your pronunciation and sound more “Japanese”.

Can you think of any other neat tips for getting the right tones?

22 thoughts on “Comparing to Chinese (Part 2): Tones

  1. I guess it’s something that has to be learned by listening and repeating.

    Also, I believe this varies in the different regions of japan which makes things more confusing.

  2. Yes, which is why we usually start with standard (Tokyo) dialect. Kind of similar to learning Mandarin (main land or otherwise).

    In the end, it all comes down to listening and repeating but students need to be taught how to to listen for and be aware of the tones in order to repeat it correctly.

  3. I think it’s enough if the students just know that there ARE tones in Japanese. I’ve been told several times that there are no tones in Japaense, even by a Japanese teacher! If they know to watch out for tones then they can be more productive in their listening and repeating of native speakers.

  4. Tones are very important, but what I consider to be at least equally important in understanding someone’s speech is rhythm.

    When beginners (or people who’ve just learned a couple of phrases) try to speak Japanese, it’s sometimes very difficult to understand what they’re saying because the syllables are stressed incorrectly.

    Just listen to Bill Murray’s Japanese in the DVD extras of Lost in Translation 🙂

  5. Tones may be a slightly inaccurate way to describe what happens in Japanese. Pitch would be a better way, mainly because it is not standardized across all words, and the same word can change pitch depending on how it’s used (compare 特急 and 通勤特急 as stated by train announcments, if any train lines near you have such trains). As Tae mentioned, without context pitch can be very important. I’ve experienced that several times. 家事and 火事 are two very different things, and there is a pitch difference. I told a friend カジで忙しかった. Of course, I meant 家事, but she heard 火事 because of my incorrect pitch.

  6. Sometimes context isn’t quite enough. Perhaps there wasn’t enough context, but I once told my (native Japanese) friend 果物くれ. He had no clue what I said. I repeated myself, and then he said the exact same thing back to me and started asking me what a 果物 was. I had to give him examples and then he realized what I said. He then corrected my pitches and now I can properly demand fruits from people. 😀

  7. “For example, I’ve noticed that all long vowels are a high and flat tone (first tone in Chinese). Just listen to how train announcers pronounce 東京.”

    I might be misunderstanding you, but I don’t think this is true. An example would be 京都, which typically has the 京 part pronounced as a quick high-low transition, with the 都 part kind of tucked in with a low, flat sound.

  8. Yes, you’re right. I’m gonna need to correct that to “often” not “all”.

    See? This is why someone needs to figure this out. Hmm… sounds like a good Phd dissertation.

  9. Another problem is the *how*. How will people go about teaching tones in Japanese? Sure, some dictionaries have tone guides (e.g. 新明解) and NHK has a pronunciation book (NHK日本語発音アクセント辞典), but in any case there is no agreed standard for communicating tone in Japanese. Kana may be phonetic, but it has zero tonal information – unlike pinyin.

    I also have to wonder about consistency of the accent data available – is there actually enough agreement between speakers of 標準語 ?

  10. There are actually accent dictionaries out there, although they usually only cover Tokyo dialect. I’ve also seen Osaka dialect accent dictionaries, and I think even a Tohoku dialect dictionary, but that’s probably about it. As far as being able to know what the intonation of a word is, there are rules for things like how particles or attached words modify the intonation and all that, but they vary between dialects. For example, in Tokyo dialect, each word is either accented or unaccented. If the word is accented on the first mora, that mora is high, while the rest of the word is low. If the accent is on any mora after that, then the first mora is low, every mora after that, up to and including the accented mora is high, and every mora after the one with the accent is low again. If the word is unaccented, then the first mora is low and all the following morae are high. Because of this, unaccented and final-accented words are identical, but only in isolation. When you attach words after them, they act differently depending on the word attached to them. For example, the particle ga doesn’t change the accent of words to which it is attached. This means that if it is attached to an accentless word, the resulting compound will also be accentless, and ga will be said with a high pitch. If the word is final-accented, the accent will remain on that mora, and the pitch will drop, leaving ga pronounced with a low pitch. This is just one simple, if long, example of how intonation functions in Tokyo dialect. There are numerous books devoted entirely to the subject. Unfortunately, I’ve not read any of them, so my practical knowledge is somewhat limited. Hope this has been enlightening, though. If anyone is interested, I can probably provide some actual concrete examples of accentuation patterns in various words.

  11. Thanks for the insightful comment. Some examples would be great since it’s hard for me to understand without them. Hmm… I wonder if they have those books on

    I actually used to own an accent dictionary but I rarely used it. Now if they had an regular electronic dictionary with the accents like my Chinese dictionary, that would be awesome.

  12. Okay, let see here… as far as an example of the effects of ga on accentuation, here’s one:

    Take the words atama and miyako. Both have the pitch pattern low-high-high, so the “a” in atama and “mi” in miyako are pronounced with a lower pitch, while “tama” and “yako” are pronounced with a higher pitch. This means that, according to the rules outlined in my previous comment, the accent could either be on the last mora of each, “ma” and “ko” respectively, or they could be accentless. In order to find out which, you can use the particle ga. In this case, atama-ga has the pitch pattern LHHL while miyako-go has the pattern LHHH. This means that atama is final accented, which is why the pitch drops after “ko” when ga is attached. Meanwhile, miyako is unaccented, so the pitch remains high.

    To give perhaps a better explanation of how the accent functions in Tokyo dialect, since my explanation above probably isn’t the greatest:

    1) The pitch of the first mora in each word will always be low unless the word is accent-initial, meaning the accent is on the first mora. This is because the accented mora, if there is one, is always high in Tokyo dialect.
    2) After the first mora, every mora up to and including the one with the accent will be high pitched. So, if there is no accent, or the accent is on the final mora, all the morae in the word except the first will be high.
    3) Every mora after the accented mora will be low. This is why ga, a particle which doesn’t alter the accentuation of words, will be low on any words except those which are accentless.

    Different particles are subject to different rules: some, like ga, don’t alter accents. Some move the accents. Some remove accents from the attached words entirely. There are also rules for connecting two nouns, which depend on the accentuation of the second word. So, for connecting two nouns:

    1) Remove the accent from the first noun.
    2) If the accent of the second word is on the last or second-to-last mora, move it to the first mora of the second word.
    3) If the second word is unaccented, place an accent on the first mora of the second word.
    4) Otherwise, the accent maintains its position in the second word.

    1) fuyu + keshiki -> fuyugeshiki
    The accent on keshiki is on the “ke”, so the accent on fuyugeshiki remains on the “ge”.
    2) hanauri + musume -> hanaurimusume
    The accent on musume is on the “me”, so the accent on hanaurimusume shifts to “mu”, the first mora in musume.
    3) chikin + karee -> chikinkaree
    Karee is accentless, so an accent is placed on “ka”, the first mora in karee.

    There you go! A quick and dirty overview of basic accentuation in Tokyo Japanese! Just be sure not to go using these for other dialects, as I know for some of them, and possibly all of them, it’s quite different. Also, if you’re interested in a book, I snagged these examples from the phonetics section of “An Introduction to Japanese Linguistics” by Natsuko Tsujimura. It’s also available on the US Amazon here. It’s a good, if very basic, introduction into Japanese linguistics. Good for people with very little linguistics knowledge, as it really works hard to spell things out. Anyway, if you have any more questions, feel free to ask. I’ve still got a week until summer is over, and I’m bored!

  13. To be fair I think every language has patterns of intonation, and can sound quite wrong if you do not follow the conventional for these. I wouldn’t claim to be an expert at Japanese, but I personally feel the nihon example is misleading (albeit convenient for the purposes of the article). Nihon (日本) – Japan – is one word. Ni hon (二本) – two bottles – is two words; the number two and the measure word for bottles. There are comparable examples in English; “sometime” (being a time at some point indeterminately distant in the future) and “some time” is a period of time. You might say “do you have some time” but you wouldn’t say “do you have sometime”, because that would make no sense. Intonation and emphasis might help you tell these apart but it would be mostly context that would give the game away, and whether the sentence made grammatical sense. That doesn’t mean english is a tonal language.

  14. Victoria,

    English is not a tonal language but an accented language. I’m sure people learning English have to pay attention to accents just like how people learning Japanese should pay attention to tones, pitches, or whatever you want to call them.

  15. There are tonal languages that are very, very complex. In fact the mother of all tonal languages are the languages from Ghana to Southern Nigeria and along the East African region.

    One of the reasons why there are art forms like rap music is based on the way that POETRY IS SAID IN A ‘TONAL’ FORM USED SPECIFICALLY FOR POETRY IN WEST AFRICA.

    The entire system of ‘rapping’ is based on tonality and pitch as well as rhythm. In fact, rap as it is done today is exactly as it was done over the past 10 thousand years in West Africa and the Nile Corridor (Egypt, Cush, Ethiopia, Somalia).

    Rapping was and still is poetry done in a rhythmic, high-low, sing-song fashion accompanied by rhythmic music from two main instruments, drums and a stringed instrument.

    This method is how ancient kings (and modern ones in Africa today) were praised, or how announcements were made to the people in communities.

    Another tonal form is used in the ‘talking drums’ and the talking guitars, harps, ect.

    ‘Blue notes’ in blues music and soul music is based on the ‘low’ (dirty) notes used in ancient times for sending messages in the music (which is still done today in Black American music, African music) and African religions in Africa and the Americas like Shango and Orisha.

    Yoruba language, Ga, Twi, Bantu-Congo, and many East African languages (like Swahili) from Sudan to South Africa uses tonality with great skill. In fact, the use of ‘tonality’ is still very strong in the accents used by African-Americans, Afro-Brazilians, Jamaicans and other Blacks of the Americas who speak European languages for everyday communication and African languages mainly for religious reasons (Yoruba, Fanti, Mandinga, others).

    Some anthropologists/linguists and historians trace some aspects of the Japanese language to the languages of Ghana and the Calabar region of Nigeria. Others point to Tanzania and Kenya. Yet one thing is absolutely certain. There are a very large number of African names of people, places and things in the Japanese languages. There are also many suffix and prefix forms that are identical to those used in parts of Africa.

    One would ask, ‘how did the Japanese gain African words?’

    That answer is simple, particular to those of Naija-Congo and Mandinga-Cush origins. According to our history, Black Anu from the Sahara (who once lived in the Sahara, Egypt, Sudan, East Africa) migrated to Japan about 10,000 years ago. The Anu were African Negroids similar to the people of today’s New Guinea and Melanesia. They were master seamen and sailed an open inland-sea in part of the Sahara. Descendents of the Anu still live in Africa today, they are called Anuak, Tibbou and other groups.

    More on ancient migrations, see “Susu Economics,”

  16. my shinyuu from toukyou taught me that ame (n.rain;rain fall.) falls down in pitch H-L like rain falls from the sky where as with ame(adj.candy;sweet)not to be confused with amai (1.sweet 2.indulgent) this patern is inverted yielding L-H N.B. amai and ame only seem slightly different in native speakers pronunciation i thought tadional formal japanese was meant to be completly flat but all the young japanese here in oxford seem to use many pithes! doumo arigatou honane matene sayounara Jib

  17. “Now if they had an regular electronic dictionary with the accents like my Chinese dictionary, that would be awesome.”

    In case no one has mentioned it yet, pitch (or tone, or accent, whichever is your preferred terminology) is indicated in “Super Daijirin” on my Canon Wordtank V80. See, e.g., 雨 (rain) has a 1 (enclosed in a square) next to it to indicate accent on the first mora, whereas 飴 (candy), which has a zero to indicate no accent. See also 橋 (bridge), which has a 2, and 箸 (chopstick), which has a 1.

    This feature has been very helpful to me.

    My JP friends say that the Osaka pattern for rain-candy is the opposite: so instead of “rain is falling,” it sounds like “candy is falling” to the “standard ear.” They also say that in the Ibaraki area (northeast of Tokyo), pitch/tone is not an issue, i.e. bridge and chopsticks sound the same.

    Hope this helps.

  18. Like English stress accent patterns, Japanese pitch accent patterns have to be learned on a word-by-word basis. That’s the bad news. The good news is over 3/4 of the words in the Japanese language are unaccented, and once you know which mora receives an accent, the pattern is predictable. (Although things get a little more complicated when dealing with compounded words.)
    But that only applies to words spoken in isolation. In actual speech, things get a little more complicated. Particles, conjugations, and intonation patterns larger than the word-level will all effect the pitch accent pattern of an utterance.
    There’s an incredibly rich body of linguistics literature written in English on Japanese prosody, but even as a linguistics minor, I can barely stomach it. From a pedological perspective, there’s a book published in 1979 by Bonjinsha called “Japanese Pronunciation Guide for English Speakers.” It looks like there was only one edition, so it’s not easy to come by, but it’s the best book on the subject I’ve seen in English.

  19. When reading about this, I can’t help but think how difficult it is to try and figure out English pronunciation just from a book. This is difficult enough for native English speakers, and it’s the reason it can be hard to understand Japanese speaking English sometimes.

    So it doesn’t surprise me in the least that it holds true for Japanese, as well. I find I have a much better grasp of pronouncing Japanese words after I hear someone say them and get a feel for the correct intonation. No different from any new English word I learn.

    I don’t think any books can really teach this well, at least not without becoming so overly complicated to the point where they’re more confusing and a waste of time. It’s probably best to make friends with people and learn from them, which is what most people do with their first language anyway.

  20. My Japanese teacher at school used different pitch patterns, probably Osaka or something. I learned Japanese copying his speech. Later on, when I came across Essential Japanese by Samuel Martin which shows the accent pattern of every Japanese word, I discovered that I was saying things wrong, at least from a Tokyo standpoint.

    So now my Japanese is a mixture of my teacher’s dialect whatever it was, standard Tokyo dialect, and just plain random because I have no idea of correct accent on most words. I must sound horrible.

  21. Ha ha, but that kind of thing can be easily solved once everybody around you speaks the same dialect whether you want to speak that dialect or not!

  22. The ‘problem’ with some people is that they think they need to learn the word *and* the tone, but in reality it’s just a word and there’s no other way to pronounce it. Let’s say, in English you want to say ‘hat’, but you say ‘hot’. It looks like ‘hat’, but it’s a different word. Why? Because of the other sound. In other words: another vowel. It’s exactly the same.

    Oh, about Chris Tucker: he tried to speak Cantonese. Jackie Chan speaks Cantonese as his native tongue, so does the majority in Hong Kong 😉

Comments are closed.