How do you sing a tonal language? 

2:55 correction: 倪 (ní) instead of 兒 (ér)

In Vietnamese, they actually match the tones of the lyrics to the pitch of the music, tones rarely get dropped, because it will be heard super awkwardly. But in Mandarin, especially modern music, they often just ignore the tones, even if a native heard a song for the first time it'd be quite hard to understand the meaning of the lyrics.

In Cantonese, especially "canto pop" in Hong Kong, the music creator and the lyricist are two different professions. Professional lyricists are able to listen to a piece of music, and match up the notes with appropriate words, while maintaining the overall message of the song. This adds another layer of difficulty and depth to the song and to the listeners. Listeners often get a second layer of appreciation and excitement when the words' native notes completely match up with the song in the most natural way, while conveying the message in a crafty manner. This second layer of appreciation doesn't exist while listening to songs in non-tonal languages such as English, basically the artist can write whatever they want into the song with much less constraints. It's great but a good canto pop song lyrics are often appreciated on its own, separated from the music, like a piece of stand alone art, or a poem, because it is really difficult to get it right.

In Mandarin, we completely ignore tones in language and solely concentrate on the musical pitch. Therefore, we must watch the lyrics to avoid confusing the meaning. Here's another interesting thing: Many foreigners struggle to master the tones of Mandarin even after decades of practice. However, when they sing a Mandarin song, their pronunciation can sound as perfect as that of a native speaker because native Mandarin speakers completely ignore tones in music. Next time, if you want to surprise your native Mandarin friend, try singing a Mandarin song🤣

I'm Thai. Singing words in our language tends to preserve the tones integrity as much as the melody allows. When tones and melody don't go together well, it will sound unnatural and awkward, and it's mostly a lyricist job to ensure it doesn't sound too awkwardly. Tones tend to be sung more relatively to how the melody goes, like, you'd find bunch of words with tones raised over mid level when melody ascend in pitch and vice versa. The more commonly used words, especially the functional word and usually unstressed word can hide well in a monotonous line of melody that has same consecutive pitch, since those words and syllable can gain atonality in normal speech.

In Vietnamese the tones of the words always have to match the music, otherwise it'd sound cringey and awkward. Seriously. There's simply no other way around it. I'm honestly quite impressed with how our writers and singers managed to pull it off

Mandarin Chinese -- even with tones, one syllable can have 20 different meanings (just like English's "different meanings for a word"). People understand, in context. Song lyrics can use the pitch of the music and be understood most of the time, because the words are used in context.

thats why for some Chinese songs, people can add subtitles to them using alternative lyrics that twist the pronunciation and tone a little bit to form new lyrics. These new lyrics are often humorous in nature and your brain just cant help but hearing these alternative lyrics because the subtitles suggestively push the brain towards the alternative interpretation of the lyrics. An example in English would be the original song being "I'm 14 carats" and twisted lyric being "I'm a farting carrot". These are easier to do in Chinese as chinese can twist tones in addition to just pronunciation.

I'm from Myanmar, and our language is also tonal. But the tones are actually not that varied since we only got three, which makes writing music while keeping that tonal sound easy. There's rarely any cases where the singer has the change the tone to match the melody.

I had this thought over 10 years ago, but no one understood my question. And here I am, you answered my question. I feel acknowledged!

A lot of music videos also include the subtitles as a part of the design to avoid confusion. But it does make you wonder how native speakers understand a song they've never heard before with no context and a melody that completely violates the natural pitch patterns. 🤔 I guess they're so used to recognizing certain groups of words that they automatically synthesize the meaning despite the lack of correlating tones.

In Norwegian, the tones basically always follow the pitch of the music. Although there are a huge amount of minimal pairs distinguished only by the tone, the members of each pair normally belong to two different grammatical classes (such as one noun and one verb), so the context is enough to understand the meaning in almost every case. There are some exceptions though, such as the words for "farmers" and "beans", and I'm sure you could make a funny song based on that ambiguity.

Somebody who can rap in those languages rhyming both the sound and the tone is automatically superior to Eminem.

I've been taught to regard tones as throat positions rather than fully equating their meaning to pitch. While this helped me with spoken Mandarin, I can fully attest that singing in these languages is still cancer. Xue hua piao piao slaps though.

In a lot of Chinese opera, I've noticed that lyrics are much more understandable when a single word is stretched over a long melodic line. In some ways this happens in languages where there are multiple syllables and this aids the listener in parsing the lyrics because you have more time to interpret a word by placing it in context. However, in Western music, most of the time it's better to match each syllable to one note, because it can add up if you're dividing syllables into multiple notes, and this ends up dragging a song down. In Chinese, since almost all words are single syllable, stretching a word out over a melodic line can go a long way in helping a listener parse. You don't have to do it with every single word to make it comprehensible, and you can even include the word's original tonal direction within that melodic line without disrupting the overall arc. The songs from the classical Dream of Red Mansions employ this very well; one example is ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-7A08gPSyMxo.htmlsi=CvrBCv6ucqNXbE8g. Obviously one problem with this is that it doesn't work so well in a fast-paced song, but even such songs would "linger", so to speak, over an important word to emphasize it, and these accents can go a long way to help the listener know what the overall gist is.

I think you're somewhat missing another tactic that's frequently used, and that is to overlay the tones onto the melody. IE The basic melody is retained, but the tones of the words are micro melodies overlaid on the main melody. It's a bit like adding Vibrato. I've noticed this as a fairly common technique as well.

In Navajo, there is a bit more of a clear cut between tone, but the change in tone usually relates to each other is what I have found, like ‘anii’ meaning face, ‘áníí’ meaning nostril, and ‘aníí’ meaning poop. It only has two tones so the range is not as diverse as it is in Mandarin or Cantonese

Kinda grateful my native language isn't tonal.

Hi, Mandarin and Taiwanese hokkien native speaker here. I rarely comment on videos but I thought this was a interesting point you brought up and a very good question I have came upon myself. I must say that while tones are important in the Chinese language in general, it's also an extremely contextural language, hence why most of the time we would still understand completely when foreigners make tonal mistakes. It's true that words share same syllable and sound but different tones such as the "ma" example you have brought up, but as a native speaker, we are very used to deducting information from sentence structures and patterns that even without tones we are still able to distill the meaning of each phrase even without tonal information. So while that does pose an extra layer of difficulty while understanding pop songs sung in Chinese or Taiwanese, if we listen carefully as a native speaker and listen to the melody one complete phrase at a time, we're still able to get the meaning :-) I'm not sure about other tonal langauge, but this is at least for Mandarin Chinese. Hope that makes sense!

This question has been in the back of my mind for years. Thanks for explaining.

the Hmong language (a fairly small language in China, Laos and Vietnam) has 8 tones, which conveniently fit into the solfege, (do, re, mi, fa, so, la ,ti, do) except that the tones obviously correlate to their own. The Hmong language however also has a very rich tradition of music, which probably explains why that is. Western pop music is a bit harder to correlate with, however traditional songs are always based off of the 8 tones.

bro completely butchered the “Ma” sounds 😂

As a person speaking both Mandarin and Cantonese, I want to share my ideas about **how different their songs are** despite they are neighboring languages: In Mandarin, most artists write songs with no regard of the tones. People just sing without caring anything about the tones. As a result, it is very common to misunderstand the lyrics. Therefore, a lot of people prefer listening to music “with subtitles on”, and every Chinese music software provide synchronized lyric display, even Apple Music in China has synced lyrics! (By the way, watching Chinese subtitles while listening to a program doesn’t interfere with each other as much as phonetic languages do. So a lot of Chinese people watch everything with subtitles.) However, in Cantonese, songwriters pay extra attention to the tones. They make sure the melody matches the tones, and consider tonal matched lyrics of higher artistic value. Cantonese has 9 tones, thus a perfect match is usually impossible. But songwriters do challenge themselves to match as much as they can, especially focusing on keywords and words that may lead to confusion.

I imagine it's a bit like how we don't put English lyrics so they don't sound weird. Even when we talk, we don't place the wrong emPHAsis on the wrong syLLAble. So too when we sing, it sounds strange for the word "happy" to sound like "hapPEE." You know what I'm saying? As a Spanish speaker who learned Spanish as a native tongue, listening to some traditional Peruvian/Bolivian music, it sounds strange as it sounds like the musicians wrote the music first and then put the words into it. The music sounds beautiful but some the words sound "off." Not a native Chinese speaker, but I imagine that to a Chinese song writer, what sounds "off" comes naturally to them. I imagine that a lot of the "modern" music is translations of music written in other languages (English, Korean, Japanese, other dialects of Chinese etc.), and so the Chinese kinda has to be forced into the music, leaving it to sound "off" to Chinese ears. As an example, BTS is very famous in Japan, but some of their songs have been translated, and the Japanese sounds weird sometimes. I speak Japanese somewhat fluently, I'm not a native, but I can hear when the Japanese is strange. Korean has final L for example, but Japanese does not, and it sounds strange in Japanese when a word is cut off.

It's worth noting that non-tonal languages have similar songwriting challenges. There is often some amount of tonal information, such as indicating a question in English, as well as conventional tone patterns in how words are said depending on how they are used. But probably the biggest consideration for a non-tonal language is rhythm. There is a natural rhythm to how words and phrases are said, and a song can sound very bad if it doesn't adhere to it. As I understand it, the singer's melody and lyrics are usually made together so that they are consistent with each other in this regard. I'm not saying tonal languages don't also need to consider this. I would presume that they do. But I am saying that it's a very similar problem, despite it not getting the same "wait, but how?" Thoughts

For contour tone languages like Chinese languages, the tone on a syllable can be dropping/rising etc, and some music does a great job of representing this in the melody, and I believe this fluctuation of pitch even within one note really contributes to the vibe of traditional oriental music

As many have already pointed out, dropping tones in Mandarin isn’t rare at all. Meaning comes from context. While “ma” can mean different things in different tones, each tone in itself also carries from a few to dozens of homophones which cannot be differentiated phonetically even the tone is clarified. However, unlike classical Chinese, modern Chinese isn’t a monosyllabic language. A big portion of its vocabulary consists of two-syllable and multi-syllable words. While “wo” and “ming” could mean many things, “wode”, “mingzi” almost always mean “my” and “name”. With enough context provided, tones can be dropped completely without causing a single ambiguity. In the earlier years of Chinese sci-fi movies and children’s programs, robots and computers often spoke with a completely toneless voice, and it didn’t really cause any confusion.

My sister married a guy who was from Vietnam. He and his immediate family have lived in the U.S. since 1975 so of course they all speak English, but they have often hosted extended family and other friends/family from Vietnam who don’t speak English. I figured it would be a nice gesture to try to learn some Vietnamese so I could at least have a rudimentary conversation with visitors. I’m a professional opera musician who deals with translations of operas for surtitles, so I’ve learned quite a bit of Italian, German, and French, plus the Spanish I studied throughout high school. I’m fairly adept at picking up the basics of languages. That wasn’t the case with Vietnamese. I understand the concept of tonal vowels, and when reading Vietnamese on a page I can understand some of it and even say a few things correctly. When spoken, though, I can’t understand anything, nor can I seem to say a whole sentence correctly. In listening to Vietnamese music with my brother-in-law’s mother, I’ve asked her about inflection and how the tonal vowels work when sung. She had no earthly idea how to explain it. She is not a musician at all so my musical "language" made no sense to her, just as her attempts to explain the sung inflection made no sense to me. The best understanding I could get was that the music usually follows the inflection. It seems similar to Handel’s use of word-painting, just done on a universal scale rather than being tossed in haphazardly for color. Your video here make it a bit clearer, since I would imagine it’s very difficult to always make the music match the inflection. I’m still a little confused as to how it works in practice; I’d love to hear a native Vietnamese speaker sing some examples of, explaining it as they go. But great video! Very fascinating topic.

A partial English equivalent to this phenomenon is how we often need to align the emphasized syllables of words the the beat of the song. I notice in a lot of Japanese music that is dubbed into English, the translators will either need to use very loose translations or just disregard this rule, leading to some funny-sounding lyrics with the em-PHA-sis in the wrong syl-LAB-le to preserve the rhythm and word order.

My son spoke Thai from birth but learned English at 12 (he is 20 now). He is very critical of the Thai language and gets grumpy if he has to speak it with his mom. He can't understand why Thais never invented punctuation or spaces between words and it bugs the crap out of him. He reinforced the concept that a new language fosters new thoughts ways of thinking that are somehow less likey to occur to someone who speaks only one language. I observe it in others who learn English as well.

An easy way to relativise the complexity of melody lyrics matching is to compare tonality to cadence in non-tonal languages (e.g. English). Fitting lyrics with the right cadence (i.e. the right lyrical meter) to match the melody is kinda similar in complexity to fitting tonality to melody, both are super complex when taken out of context but they’re both “natural” for song-making in those respective cultures

In mandarin the tones are all ignored. You can still understand mandarin w/o tones quite well

I'm Thai and never think of that before. there are many accent and tones within the country some are familiar with most people and some not so singing in Thai doesn't mean that you have to have right exact tone cause it imposible cause it can shift as long as it doesnt shift too much from the standard it still can be understood completely fine because when you listen to music you listen to the whole sentence not just one word it hard to be misunderstood even few of the word ain't tone correct, the other tonal language might be the same I guess.

I asked this same question many years ago to my Chinese friends and nobody could properly explain it to me. I think I get clearer answers from the comment section here, that in general Mandarin songs ignore the tones while Cantonese ones try to match the tones. But I don’t know how wise it is to learn from RU-vid comment section…

cool video. i've been taking voice classes in non-tonal languages (english/italian/german), and even in them we alter our vowels pretty frequently from the actual spoken ones for artistic effect.

This really isn't as much of an issue as one might think. Tones are only one of many ways tonal languages transmit information, and information transmitted by tones is often accompanied by other information for redundancy. For example, many words in Chinese are compound words made of more than one syllable, so with additional syllables as well as logical associations of other words elsewhere in the sentence, you can still distinguish meaning without tones. I find that this concern with the ambiguity of tones is mostly a phobia for the new students who only learn and use single, simple words often made of single syllables. Once a higher degree of proficiency is achieved nobody really thinks about the tones anymore. Though that is not to say you can get away unnoticed with using the wrong tones; they can guess what you mean but any tonal mistakes will still be obvious, just like making a typo in writing. Therefore, imagine an English line, where for artistic effect, some words are misspelt. Evan thorough the randition is wrung, you wood still unterstand, thorough maybie with somme diffaculty. Likewise, one would still be able to parse the lyrics of a tonal language with its tones messed up by the music.

Oh the jokes we have in Chinese just from the mismatched lyrics and song melodies

This is an interesting question and a clear description -- but what I'd really like is some audio examples where we can hear a line of a song as spoken and then as sung. Maybe in an upcoming video?

Traditional Taiwanese music/opera do tend to have music follow the tones. It's not that hard actually. Pick the cord, if it's rising tone, match a 3 cord upward. Thing is, if you comepletely ignore tones, the song itself becomes not understandable without reading lyrics.

There are also secondary qualities to the tone pronunciation, such as length, creaky voice, or stress on certain parts of the syllable, and often they're enough to tell the tones apart in a song (at least from my experience in Mandarin n Cantonese)

0:13 You aren't that far off! Your first tone is pretty good. Your 3rd tone sounds close to what the 2nd tone should actually be and your 5th tone sounds like what the 4th tone should actually be (the 5th tone is kind of like an unemphasized 4th tone). Your 2nd and 4th tones are a little off - the 2nd starts low and rises and the 4th starts high and falls in pitch, so they aren't just higher/lower pitched versions of the 1st tone. Source: Native Mandarin speaker

you should definitely make a video about how different arabic dialects are, why they are so, and what a diglossia is, as a native speaker it'll be very cool if you bring light to it! shokran.

Tonal language actually make it harder to make lyrics, since there's fewer words to choice, and using wrong tone make the song impossible to understand at first time listening. Languages like Japanese are ones of the most easy to make, since tone isn't an important role, and each word's simple enough to make a sound, and there's alot of sounds in one sentence

Spanish might not be defined as a tonal language, but words can change meaning depending on the accent. As a simple example we can take a look at the words práctico/practico/practicó, each with the accent in each of the vowels Práctico = Practical Practico = I practice Practicó He/She practiced As the accent is very important in these words, songs are usually composed with that in mind, as not following the accents in the words can make it sound weird at best In fact, there's this song by Enrique Iglesias and Juan Luis Guerra called "Cuando me enamoro" (accent in the first o in enamoro; "When I fall in love"), but in the actual song the melody makes it sound more like "Cuando me enamoró" ("When he/she made me fall in love with him/her")

it's truly amazed me cause I've never thought there's this kind of difficulty when it comes to tonal languages. I guess Vietnamese song writers have been doing a pretty good job overcoming this challenge that it doesn't seem like a problem to be taken in to account at all.

If you listen to Afrobeat, Yoruba lyrics completely ignore the tone.

In classic Chinese literacy, there is a format called "Ci", which basically mean "lyrics". It comes with different "titles", so you can have multiple implementation under the same "title". Each title has a specific rule where each word has to follow some tonal rules, so each implementation, though has different words, will have similar tonal transition/progression. So basically each title should have a music comes with it, so the same music can be sung with different lyrics( under the same/similar tonal transition/progression). Countless "Ci" were produced by poets and were sung by vocalists. But too sad, all the music from those titles are lost in time, only the lyrics remain. Still some great Chinese/Cantonese songs still follow that similar rules so it won't feel weird to sing.

0:12 the way you pronounce the five tones sound all like the first tone😵

You somehow managed to rope in the Dene-Yeniseian hypothesis in a video about singing in tonal languages. Instantly subscribed

I came with one curious single question in mind and now I ended up more confused after your over the top answers.

this reminds me of when my mandarin teacher said in my first semester of mandarin that the only time we'd get to escape tones was if we were singing and then this girl asked if she could just sing all the time to not have to remember tones 😆

I see this question as asking if a language with vowel length distraction can't be sung. I assume it works the same: the length is ignored in modern music, but traditional music trend to match the length of the words to the music, or the music to the length of the words.

其實在我年少時，因為政府不鼓勵學習廣東話， 所以也不是太明白的！但是後來遊歷全球，經多方觀察，才明白粵語確是非常獨特的！可以說：世界上沒有一種語言在作曲填詞方面可以與粵語相比的！請勿誤會，當然粵語歌也有不好的。我是說在作曲填詞的最高境界上，現在是找不到比粵語更好的語言。我自幼便深通普國語，至今任何人對我說普國語；我都會聽懂超過99％。可是我不明白為什麼有時我在聽一些普國語歌時，竟然連一個字也聽不懂！原來超過八成普國語的字都是雙聲字而那兩聲通常都是相差很遠的（例如：雙字，shu ang)！九成半以上粵語字都是單聲字（例如：雙，seung)！而且粵語字通常都是九聲齊全！所以在粵語來說，要找一個字的發音與曲調完全相同比普國語是容易很多。反觀所有普國語歌，幾乎每句曲詞都有拗音字（即改變字的發音去遷就曲調）。有些普國語歌幾乎每字拗音！所以我便聽不懂！歐洲的拼音式多聲字便更無法相比！所以好的粵語歌是可以做到詞曲合一，差不多沒有拗音字；唱歌和平常說話幾乎相同。全世界沒有另一種語言有這個粵語的特色，普國語也不例外！可是粵語一定會在中國失傳！因為現在中國政府的統治模式是必然使多數語言消亡的！這不一定是中國政府白紙黑字的政策；但是中國人的統治者為了防止地方有任何實權，便規定各地方領導都不准是當地人！非當地人統治便不鼓勵當地的語言！這是不難理解的！反觀民主國家如印尼，雖然有全民皆懂的國語（Bahasa Indonesia）,但地方語言如(Javanese)並沒有消失！又如菲律賓雖有幾乎全民皆通的(Tagalog)，但地方語言如(Ilocano)沒有消失！甚至美國的西裔人還是經常以西班牙語溝通的！反觀中國，上海話從1949年至現在只七十年便幾乎全面消亡！放棄粵語便是放棄人類文明的重要遺產！長遠來說，廣東人是要到外國去學習粵語！誠可嘆也！ 普通話國語與歐洲的多聲字語言一樣；在唱歌時都離不開拗音，所以有時便全曲拗音，成為一種獨特的歌唱形式，叫藝術歌曲！如果沒有字幕，普通人是完全聽不懂的；但歌者還引以為榮！可是廣東話因為可以做到詞曲合一，不需拗音。所以粵語歌是沒有（或幾乎沒有）藝術歌曲的！但是就是因為世上沒有粵語藝術歌曲，很多北方的歌唱家都看不起粵語歌，認為沒有藝術！其實這是大錯特錯！因為全世界除了粵語歌外，沒有其他的歌曲可以做到詞曲合一；不需拗音！光是這個特式便足證粵語歌的藝術成就。同時正正就是因為粵語歌的藝術成就非凡，大部份廣東人都不太喜歡聽非粵語歌曲，因為非粵語歌曲都有很多拗音字！這樣使廣東人看起來好像排他性很強，其實都是誤解！ 每個中文字的粵語發聲最多可以有九聲。普通人是不一定可以發出到九聲；但是普通人最少也可以發六聲。例如那個(Si)部的粵語發音由高至低可以表達如下：（1）師（2）史（3）試（4）市（5）是（6）時。(Fun)部：（1）分（2）粉（3）訓（4）奮（5）份（6）焚。(Ma）部: (1)媽（2）嫲（3）嗎（4）馬（5）罵（6）麻。(Se)部：（1）些（2）寫（3）瀉（4）社（5）射（6）蛇。(Sui)部：（1）需（2）水（3）歲（4）緒（5）睡（6）誰。（Yi）部：（1）依（2）椅（3）意（4）以（5）二（6）宜。(Yuen)部：（1)冤（2）丸（3）怨（4）遠（5）願（6）園。(Ngoi)部：（1）哀（2）靄（3）愛（4）有聲無字（5）外（6）呆。(Lo)部 ：（1）撈（2）佬（3）有聲無字（4）老（5）路（6）勞。(Tou)部：（1）滔（2）土（3）吐（4）肚（5）有聲無字（6）途。(Wai)部：（1）威（2）位（3）尉（4）偉（5）惠（6）維。(Wong)部：（1）汪（2）枉（3）有聲無字（4）往（5）旺 (6) 黃。(Wu）部：（1）烏（2）糊（3）惡（4）有聲無字（5）護（6）胡。(Yim) 部: (1) 閹（2）掩 (3) 厭 （4）染 (5) 艷 （6）嚴。(Yung)部：（1）翁（2）擁（3）有聲無字（4）勇（5）用（6）容。又(Si)部九聲是（1）師（2）史（3）試（4）市（5）是（6）時（7）舌（d尾）（8）攝（p尾）（9）蝕（t尾）；後三聲是尾聲。沒有一個普通話國語的字的發聲是如此齊全的；因為所有普通話國語字都只有四聲！所以粵語纔是完整的漢語；普通話國語是真正的方言！各位還記得尤雅那首著名的『往事只能回味』吧？她第一句的第一個字是（時）字，同時她第三句的第四個字也是（時）字。可是這兩個相同的字尤雅是用完全不同的聲調唱出來的！因為她在唱第二個（時）拗了音去遷就曲調。這首歌的曲調不錯；但是每句曲詞都有不同程度的拗音字。全部普通話國語歌都有這個問題！反觀粵語歌，雖然偶然也有拗音字，但拗音情況是極少的,粵語歌的拗音字例子如下:關正傑那首經典(情愛幾多哀)最後三句:(為換到,她的愛,甘心衝進恨海)那個(愛)字,他唱成(哀)字!但是關正傑這首名作全首歌只有這一個拗音字而已!普通話國語幾乎沒有一首歌的作詞可以有這種全曲只有一個拗音的水準!因為所有普通話國語歌幾乎每句曲詞都有拗音字!歐式語言的歌曲就更不足論。且聽Diana Ross那首名曲 If We Hold On Together 幾乎全曲拗音。 因為粵語歌普遍無拗音，唱粤語歌跟平常說粵語沒有很大的分別。所以與全世界的語言不同；最佳學習粵語的方法就是學唱粵語歌！ 兹列舉數句尤雅此曲中的拗音字如下: 時(光）一去永不回；往事只能（回）味!憶童年（時）足馬清（梅），兩少（無）（猜）日（夜）相随。（春）（風）又（吹）紅（了）（花）蕊，（你）也經已添了新歲！ 廣東人作了數千首完全沒有拗音字的粵語歌!例如葉振棠那首『江湖行』唱起來跟平時講廣東話完全無分別！但是完全沒有拗音字的普通話國語歌絕無僅有，找來找去只有那首廣東人作的「滄海一聲笑」！將來粵語絕跡於中國後，中國便不會再有第二首無拗音字的普語歌了；因為只有廣東人纔會創作無拗音的歌曲！

As a Cantonese speaker. Filling in the lyrics has always been hard to do, so hard in fact it is an actual separate job than the singer/composer in the music industry. But the result is beautiful.

In Myanmar, we take account the tone of the word to the lyrics and sound. Cus if we ignore the tones of the words, the song might sound a bit awkward and weird even tho we know the meaning. But basically, we only have 3 tones (eg. မ, မာ, မား) so it might a bit easier than other languages, I think.

In addition to the five tones, Thai also have the alive and dead syllables. Different tones will allow for some movement in pitch, but if they don't fit it will sound completely out of tune, or misinterpreted

Theres more to tones than just pitch; there can also be shifts in sound quality ("swallowed" sounds, for example), and variation in tone duration. So it can often be possible ro still express tones through these kinds of distinctions while singing, as long as the melody doesn't interfere too severely

I'm very late but please remember that even in the northeast part of India they also use tonal languages and consists of Sino-Tibeto Burman dialets. There are about 5 and other states where tonal language is spoken,such as Mizoram, Sikkim, Manipur, Arunachal Pradesh and Nagaland etc.

When you're listening to Vietnamese, the tones are very rarely dropped. But even if they were you could still make out most of the lyrics using context clues

Actually, you can understand based on context. You may get confused with individual word but hearing it in the full sentence will make sense. For example in the line “trời còn làm mưa, mưa rơi mênh mang” from Trịnh Công Sơn’s “Tuổi Đá Buồn”, the “còn” is actually a whole step above “trời”, even though both have the falling tone, which makes it sound like “con”, or child, rather than “còn”, or still. But the listener knows that the whole like is “the sky is STILL making raining” and not “the sky is CHILD making raining”. Mưa (neutral tone, no tone) also sounds like mứa (rising tone), because it’s a whole step and half step above from trời. But mứa isn’t even a word. So it’s all just context. Kind of like how in English you’ll know if it’s read past tense or read present tense based on the context of the whole sentence. Oh and even then, in the example I used, there were two syllables with the falling tone but one is higher. This happens, but what will NEVER happen is the next syllable is a higher tone but a lower note. Never. It can be same tone higher note, higher tone same note, but it can never go in the opposite or it will sound super awkward and unnatural, even if we can understand based in context, as my fellow Vietnamese have pointed out in the comments.

as someone who listens to lots of chinese pop music, they don't really think about it when writing the songs I think, you kinda garner the meaning through context

Sometimes ignoring the tones makes vagueness thats poetic.For example theres a Xhosa song Senzeni Na? Because of the none clear tones it can be interpret as either "What must we do?" Or "What have we done?".Which are both relevant in the context of it.

Looking forward to Mixtec. Mesoamerican languages are very underrated

There are other possibilities. In Western music, it's common to fit one syllable to one note (often with the stressed syllable in English coinciding with the stressed beat in the bar). But you do also get a number of syllables on one note (at the extreme in Anglican Chant) or one syllable spread across several notes (in Italian and Eastern songs). Either of these could be tried with tone languages to try and fit the words and music better together. Two other ideas - and I'm not sure they would work! - 1. Share the time, alternating words and melody in any way desired (shown here as doh-re-mi for music), e.g. doh - twinkle - soh - twinkle - lah te doh te lah - little - soh - star [I demonstrate that notes and syllables are not related]. or 2. Much easier, have one person sing the melody while another person speaks the words over it in a coordinated fashion. The song "Desiderata" works a bit like this.

I was walking with some people at the Hồ Chí Minh mausoleum and a vietnamese friend explained that one of the songs we were hearing was written with the tones in mind. When the song said his name, the melody was So, Mi, Do. Low, High, Middle note. Which matched the tones of his name. I asked if the music always matched the tones. And she said usually yes because of not, it's considered bad song writing.

Actually even though a syllable may have a different tone what counts for the tone to be understood is not the mean pitch but the shape of the pitch change : if the tone is gliding, falling, rising, dipping it doesn't matter from or to what height it is gliding, falling, rising or dipping.

Another consideration in lyrical syntax. Languages like English, with strict syntactic rules, do not allow for words to easily be transposed. For languages with looser syntax, it's possible to rearrange the words to better fit the meter, rhyme, and also syllabic tone.

Interesting topic - would’ve been helpful to have some examples to make the point.

Never thought YT had an answer for this question

In my experience of Thai the only tone which is really sung is the high tone to get your horse, whereas the others are a bit more accommodating of pitch difference so I would call the Thai falling tone is an emphatic tone and the rising tone is a questioning tone as well which helps. As for singing Thai there is a distinction that all Thai poetry can be sung but there an allotted melody for a specific meter of poetry.

1:54 that exploding head drawing made me laugh so much I subscribed.

When I came to China, I became aware of this when most of the Chinese people I met said they couldn't understand a song if they just heard it and didn't have the lyrics to read or check the lyrics prior. Say for instance we are in a bar and a Chinese song they never heard before stars playing, and you ask them what is about or to make a translation, they almost always say they cannot understand the lyrics. And later I realized because they are a tonal language, if the tone of the word they are singing doesn't match the "tone" [note] of the music, they cannot really understand it. So they might have to check the written lyrics to actually understand the song.

I'm Norwegian, and I've been asked this before; how do we sing with pitch accent? (I have a lot of language interested friends) Usually, as words that are spelled the same, but pronounced differently based on pitch, grammar, context, expression, etc, needs to be sung, we either, as stated, just don't care about the pitch, or we use synonyms to make the song cleaner. Instead of "Får får får? Nei får får ikke får, får får lam" ("Do sheep get sheep? No, sheep don't get sheep, sheep get lamb" all differently pronounced), it could be written as "Do Sheep birth sheep? No, Sheed do not birth sheep, sheep birth lamb". Some honourable pitch-mentiones are "Danger and The Father", "The stable and the law", "The Step and The ladder", Etc. As a return to this, I've asked a chinese speaking friend the same; how do you sing a chinese song? Usually, they don't care whatsoever. He explained it like this: It has a lot to do about the kind of words you use. Saying "Wo shi" (I am (I don't know how to do accent marks, sorry)) at the beginning of the sentence in any tone in a song, we know you mean "I am", instead of something different, like "Crouching rock" (wo shi, different tones). It only makes sense, and we make that joke a lot. Context with what kind of words you use, and even grammar and what kind of verbs, are important; as well as tones. But when you can't use tones properly, you still have the context to help.

i was actually just thinking of this very dilemma the other day. great video!

A lot would depend on the number of tones. Mandarin Chinese has 5, but many African languages have only two: high and low. (Their drum language involves two drums, one with a high and another with a low tone.) Even when an African language has 3 or 4 tones, their words usually have several syllables, so there are fewer pairs of words separated only by tone.

Interesting. I've actually wondered about this for a long time. FunFact(?): I've read that people who speak tone languages are more likely to have perfect pitch

i chose chinese as my major in the university in 2023. we have had 2-3 lessons at the moment but we started learning chinese(mandarin) with the easiest keys? and hieroglyphs. tones.. oh god yes i hear the slight difference but i really cant say them correctly or like a native(chinese sounds are interesting and im glad there's no hard R). it takes too much time for my brain to remember which tone i should use so singing with tones would be a pain for me

I brought this question up at university almost 30 years ago but nobody could answer (it was to a researcher on thai music). Glad you finally answered this and for including the sources. First link doesnt work for me though if you can repost.

Native speaker of a pitch accent language - Swedish - here. Composing songs in swedish, and singing them, are not made harder by our pitch accent system. For all (at least all that I can think of at the moment) word pairs that are spelled the same and pronounced the same with exception of tone, the meanings are so different so that the brain instantly chooses the only word in that word pair that can make any sort of sense. For example anden/anden, meaning the duck/the holy spirit. Only one of those words makes sense in a religious song. Note that we express definite singular with a suffix, not a separate word - the - like in English.

This is one of those questions I’ve had in my mind for years now. Thanks for the video!

note: some other famous pitch accent languages include Japanese and ancient Greek.

do you got actual examples for any of this

This is a great find! I'm going through all your videos and really loving them.

What about Mandarin or Cantonese covers of Western pop songs? The melody is already pre-written. For example, 鄧麗君 Teresa Teng singing《比翼鳥》, a Mandarin cover of "San Francisco (Be Sure to Wear Flowers in Your Hair)", an American pop music song written by John Phillips and sung by Scott McKenzie. Or 陳百強 Danny Chan singing《粉紅色的一生》, a Cantonese cover of "La Vie en Rose". So the Mandarin/Cantonese lyrics are written regardless of tone.

Very nice video, despite speaking mandarin I never thought of this before!

This isn't a problem for native speakers of tonal languages. As a native speaker of Thai, I can sing Thai, Lao, Vietnamese, Mandarin, Cantonese and Fujianese songs naturally. I don't really need the solutions proposed in this video. Moreover, I know that many Vietnamese can sing Thai, Lao and Mandarin songs naturally, too.

This question has passed through my mind and i am exited for the answer

Japanese uses pitch accent, and it can be pretty important when trying to understand homophones

Great video. Very interesting concept I had no idea of. I would love if you could expand on this concept with actual music examples.... although I can see how unfeasible that might be.

Not every Thai words can varies of all 5 tone sounds. So it's not a big problem at all to put the lyric to match the melody since people can guest what should be the correct intended words if it's just slightly tone shifting.

Taiwan isn’t part of the PRC. It should be a different color on the map

Well,after analyzing a couple Thai songs with same words,which had same vowels,I have concluded that Thai singer and musician care more about the tone of the music rather than vowels. Example is, the word น่ารัก(narak-means cute) in the song ‘เหมือนเคย’ and ‘เกินต้าน’ sounds completely different,in เหมือนเคย the word Narak sounds like it uses a higher vowel than the word narak in เกินต้าน You can try and catch this yourself if you search up these two songs. Edit: so,I forgot to write my conclusion. I think Thai songs cares more about pitches rather than natural vowels. The music link: เกินต้าน: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-9zXeLZTBd80.htmlsi=VH4SgPSrL8IuIhHw ;เหมือนเคย: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-94vAkI2AvNk.htmlsi=0SwOUp3-XE8QG0Ta

0:31 FYI, the North-Eastern States of India come under this as well. Many of the indigenous tribals use similar tonal languages including myself. Most derived from the Indo-Tibetan languages. =)

So it’s kind of like how you have to select the words with the right amount of syllables and rhyme to match the rhythm of the song, or if the lyrics has already been written, to write the tune that matches the built in rhythm.

In my own experience, an older songs especially church hymns/school songs, the tones are completely ignored, the lyrics are hard to understand. In modern pop songs, the lyrics are written to fit the music in tones (and if its only a couple of tones not fitted in a sentence, we mostly can understand what is being sung). It's more work but thats why a good lyricist is valued in Hong Kong.

Very interesting question, I was thinking about it recently but was clueless. Thanks man!

So in Vietnamese for example would you need to modulate the pitch by a semi tone in some cases, and 2 in others, depending on the Scale. For example C Major, if singing a C and singing tình you'd bend down a semitone to B, but for the word thúc you'd bend up 2 semitones to a D. And if it's a pentatonic scale would you modulate 3 semitones in some case.

been looking forward to this vid

1:54 You say the first solution is OBVIOUSLY pretty difficult. If you are a native speaker, I dont see why it would all that difficult. I don't speak a tonal language, and know very little about them at all, but I consider myself an expert on music composition and know quite a bit about how librettist/composers and singer-songwriters go about writing music, and I compose myself. (a librettist/composer is sort like the "high art" version of the singer-songwriter's "folk" version of the same occupation. Think musicals and operas.) Let me tell you, it not that hard, and I would say it is basically required if you write songs. I mean in English. Obviously with tonal languages it might be MORE difficult, but compare it to, say, the accents in poems. In English, the natural flow is unstressed, followed by stressed, unstressed, stessed, ect. Not always though. Use a poem with a different stress pattern, but with the same syllable patern and melody as Amazing Grace (8, 6, 8, 6). It won't work. Look at the unstressed syllables of lyrics of most music. On average the pitch is lower rather than higher than the next stressed syllable. And matching the melodies direction to the lyrics is very common. How many lyrics do you know have words like "reaching higher" or "high above" or "i wanna fly like the birds?" Think about it. Was the melody ever moving downwards? The same is common with raising the pitch if at the end if the words form a question. This is called Prosody (actually an umbrella term for other things as well...) There a some very academic rules and concepts on how to do this "correctly" but 95% of people who compose by ear, with no music theory under their belts, just do this subconsciously without even knowing it. Think of some famous tv and movie themes. Now try to sing title of the it using they melody. Did it work? It often does! And the composer didn't even need to do this as these are meant to be insrumentals! TV

In Vietnamese, if you speak the right way, it is singing. A song can be written with lyrics first or note first. With note first, the versatile of lyric might be reduced.