Are tones in Chinese songs consistent or arbitrary?

[Jan Finster]

 

I wonder if tones are somewhat "fluid" in Chinese songs?

 

I listened to this: https://www.youtube.com/watch?v=rRCKcsjrsMI&ab_channel=TheMandarinGuy

 

...and even with the lyrics I do not feel the tone is right many times !? (or my ear loses them in the melody?)

 

[Woodford]

I could barely make out any tones in that song at all! I wish I could find the thread here (I tried to search for it), but there was a lively discussion about whether songs have tones in them. Many Chinese people say "yes," while many non-Chinese people say, "No."

 

I was talking with my tutor yesterday about Chinese songs (this one in particular, which somehow became extremely internet-famous in the USA: https://www.youtube.com/watch?v=cnrxJtJcSew). And I said, "Chinese songs are hard for me to understand, because the tones are gone!" And he replied, "Well, the tones are there, but you just can't hear them!" That's far from the first time I've heard someone say that, and I just can't hear the tones.

[Demonic_Duck]

They’re consistently absent (i.e. no tonal information at all, just the pitch of the notes, which is unrelated to tone) in the vast majority of contemporary songs, including both of the ones @Jan Finster and @Woodford posted. The exceptions to this rule are mainly in certain traditional or folk styles of music (京剧、二人转 etc).

 

On 1/26/2022 at 8:19 PM, Woodford said:

"Well, the tones are there, but you just can't hear them!"

 

That’s like saying “the colors are still there, but you just can't see them” about a black-and-white photograph. I mean maybe it's true in some abstract conceptual sense: someone who already knows the tones might “hear” them by reconstructing them in their head, just as someone who already knows what color an object is might “see” that color in the black-and-white photograph. But it’s certainly not true as an observable characteristic of the underlying substrate.

[phills]

On 1/27/2022 at 6:38 AM, Demonic_Duck said:

I mean maybe it's true in some abstract conceptual sense: someone who already knows the tones might “hear” them by reconstructing them in their head, just as someone who already knows what color an object is might “see” that color in the black-and-white photograph. But it’s certainly not true as an observable characteristic of the underlying substrate.

 

That's a good explanation.  But it just might be really subtle. 

 

Or if it's true that tones are "throat positions" and not "pitches" then maybe the singers do clench their throats in the proper way (just by habit) when they belt out the required pitch.

 

Given how playful Chinese humor is, I wonder if someone has written 2 versions of a song using words with the same sounds but different tones, which sound "nominally" the same when sung but has sly differences in meanings when read?

[Jan Finster]

On 1/26/2022 at 11:38 PM, Demonic_Duck said:

They’re consistently absent (i.e. no tonal information at all, just the pitch of the notes, which is inrelated to tone) in the vast majority of contemporary songs, including both of the ones @Jan Finster and @Woodford posted

 

So, if I read the lyrics to a Chinese person and I would use no tone movement (e.g. 5 to 1 or 3 to 5)), but only  said the respective words in the pitch of the end position (e.g. 1 for 4th tone), they would understand?? And the same people do not understand me if I would tone-lessly say "wo ai ni"... ??? ??

[EnergyReaper]

On 1/27/2022 at 2:15 PM, Jan Finster said:

if I read the lyrics to a Chinese person and I would use no tone movement (e.g. 5 to 1 or 3 to 5)), but only  said the respective words in the pitch of the end position (e.g. 1 for 4th tone), they would understand?

Most likely, I can't understand. In fact, even if you read the lyrics perfectly right in tones, I may not understand, and that's usually because the song uses non-common words from everyday Chinese, e.g. 青花瓷 by 周杰伦（Jay Chou）. But lyrics of 中国话 by S.H.E is relatively easy to understand.

[Jan Finster]

On 1/27/2022 at 10:35 AM, EnergyReaper said:

Most likely, I can't understand. In fact, even if you read the lyrics perfectly right in tones

 

Oh great, so we are with back to the English song lyrics problem (https://www.youtube.com/watch?v=7my5baoCVv8) ?

[Demonic_Duck]

On 1/27/2022 at 5:20 AM, phills said:

if it's true that tones are "throat positions" and not "pitches" then maybe the singers do clench their throats in the proper way (just by habit) when they belt out the required pitch.

 

Thinking of them as “throat positions” might help with producing the sounds, but ultimately those “throat positions” are realized as pitch contours, and it's really just the vocal cords rather than any other part of the throat. You can see this by viewing a spectrogram of speech, though speech production is messy and inexact, and intonation also plays a large part, so the results won't map perfectly onto a 4-tone chart.

 

In most styles of music, the use of the vocal cords to produce pitch is purely used for creating melody; alter the tightness of the vocal cords and you alter the pitch and thus the perceived melody. As with speech, you can see the changes in pitch using a spectrogram. If you use a typical mandopop song as input, you'll see the pitch changes map pretty closely to melody, but not at all to tone.

 

On 1/27/2022 at 6:15 AM, Jan Finster said:

So, if I read the lyrics to a Chinese person and I would use no tone movement (e.g. 5 to 1 or 3 to 5)), but only  said the respective words in the pitch of the end position (e.g. 1 for 4th tone), they would understand?? And the same people do not understand me if I would tone-lessly say "wo ai ni"... ??? ??

 

If your pronunciation is otherwise perfect and very clear (clarity is often a problem in any style of singing in any language, not just Chinese), they'll probably get all or most of it, depending how simple and/or formulaic the lyrics are.

 

What they won't be able to do is accurately distinguish between two words differentiated only by tone that are both equally plausible in context. However, such situations are relatively rare. By losing tone, you do lose some information, but it's usually not enough to make an utterance completely unintelligible, unless other information is also missing or corrupted (background noise, weird grammar or word choice, mouth is obstructed, other aspects of pronunciation are off, etc etc).

 

Note also that absent tone information, as in singing, is much easier to understand than present but incorrect tone information. Wǒ ài nǐ > wo ai ni > wǒ ái ní (我挨泥？ I suffer mud?)

[anonymoose]

On 1/27/2022 at 5:20 AM, phills said:

Given how playful Chinese humor is, I wonder if someone has written 2 versions of a song using words with the same sounds but different tones, which sound "nominally" the same when sung but has sly differences in meanings when read?

 

I'm not sure why anyone would write two versions of the same song, but in terms of using words with the same sounds but different tones, which sound "nominally" the same when sung but has sly differences in meanings when read, one song that comes to mind is 草泥马之歌.

[Dawei3]

On 1/26/2022 at 3:19 PM, Woodford said:

wish I could find the thread here

You might be thinking of a previous discussion in which we discussed whether Chinese had trouble understanding words used in songs due to less distinct tones.  

 

For myself, there are rock & rolls songs in English for which it took me years to actually understand what they were actually saying (and I'm a native English speaker).  

[alantin]

Is there anyone who can understand the lyrics of "Cotton Eyed Joe" by ear only and without having ever heard it before?

[carlo]

I like writing songs and have sometimes done it in Chinese for fun. I haven't studied this seriously (pro linguists might have). Purely based on my experience:

 

In any language, there are good and bad ways to set words to music. The rule is that the melody should not clash with the natural shape (prosody, highs and lows, rhythm) of the spoken language. Wang Feng sings 北京北京 as G-Eb-G-C. If you flip the melody around and sing 北 on the highest note, it's either awkward or incomprehensible.

 

Translating lyrics is hard because you must fit the translation to a melody you can't change. But it's the same idea. There is no direct translation between tones and melodic intervals, but it's not arbitrary either: there are some tone combinations that work better than others on a given melody. That's probably why native speakers still hear the tones in the song.

[Woodford]

On 1/31/2022 at 10:20 AM, carlo said:

There is no direct translation between tones and melodic intervals, but it's not arbitrary either

 

That's such an interesting observation. It makes me think: English has some tones in it (we use something like the 4th tone to place an emphasis on the syllable of a word, while the non-emphasized syllables have something like the 5th/neutral tone). Yet when it's put to music, those "tones" are "lost" as well. But like you said, if the music directly clashes with the underlying tones, it sounds awkward. I remember having a friendly debate with my wife over two different potential melodies to a song. One melody (the one I liked) used a quarter note, eighth note, and then another eighth note over the word "eternal." So....long-short-short. My wife said, "No, that's awkward. Nobody says, 'EEEEE-ternal' in regular speech." Ah, yes, I suppose that's right.

[Publius]

On 2/1/2022 at 12:20 AM, carlo said:

In any language, there are good and bad ways to set words to music. The rule is that the melody should not clash with the natural shape (prosody, highs and lows, rhythm) of the spoken language.

Quite so.

A fine example is the Cantonese lyrics to 《滄海一聲笑》.

The tune is very simple, using five notes in the traditional pentatonic scale (C D E G A / do re mi so la / 1 2 3 5 6).

It runs like this (with small variations):

6• 5 3 2 1 – | 3• 2 1 6̣ 5̣  – | 5̣• 6̣ 5̣ 6̣ 1• 2 3 5 | 6• 5 32 1 2 – ||

And the lyrics (for better visualization I changed Jyutping tone numbers to IPA tone letters):

滄海一聲笑//滔滔兩岸潮
cong˥ hoi˧˥ jat˥ sing˥ siu˧ // tou˥ tou˥ loeng˩˧ ngon˨ ciu˨˩
浮沉隨浪//只記今朝 (6 56 53 21 2 –)
fau˨˩ cam˨˩ ceoi˨˩ long˨ // zi˧˥ gei˧ gam˥ ziu˥

 

(6 65 3 2 1 –)

蒼天笑//紛紛世上潮
cong˥ tin˥ siu˧ // fan˥ fan˥ sai˧ soeng˨ ciu˨˩
誰負誰勝出//天知曉
seoi˨˩ fu˨ seoi˨˩ sing˧ ceot˥ // tin˥ zi˥ hiu˧˥

 

江山笑//煙雨遙
gong˥ saan˥ siu˧ // jin˥ jyu˩˧ jiu˨˩
濤浪淘盡紅塵俗世//幾多嬌
tou˨˩ long˨ tou˨˩ zeon˨ hung˨˩ can˨˩ zuk˨ sai˧ // gei˧˥ do˥ giu˥

 

清風笑//竟惹寂寥
cing˥ fung˥ siu˧ // ging˧˥ je˩˧ zik˨ liu˨˩
豪情還剩了//一襟晚照 (6• 5 32 1 1 –)
hou˨˩ cing˨˩ waan˨˩ sing˨ liu˩˧ // jat˥ kam˥ maan˩˧ ziu˧

 

蒼生笑//不再寂寥
cong˥ sang˥ siu˧ // bat˥ zoi˧ zik˨ liu˨˩
豪情仍在//癡癡笑笑
hou˨˩ cing˨˩ jing˨˩ zoi˨ // ci˥ ci˥ siu˧ siu˧

 

Note how the "creaky" tones 潮, 遙, 寥, 浮沉, 豪情, and濤浪淘盡 fit especially well with the lowest part of the tune? Unfortunately it doesn't work if sung in Mandarin.

[carlo]

Cantonese with its richer system of tones probably takes it to another level. Japanese has a similar issue with pitch accents (今 vs 居間). Not just tones, but rhythm, syllable length, vowel shape, intonation etc have musical implications. This is allegedly why people with different native languages write distinctive melodies, even when the piece is purely instrumental (likely through the influence of folk music).