Second tones usually start level or with a dip

[realmayo]

I've been playing around with the free Praat software a bit recently and have noticed two features of the second tone:

 

a) Second-tone characters are pronounced with either a level sound, or a dip, for 50% or more of the sound. It's only at the end that there is a rise.

 

b) If the final part of the sound is a consonant, the rising tone often only starts when that consonant starts, so the vowel doesn't rise at all.

 

I mention this because if people are trying to get a natural-sounding second tone, and are following the textbook description of a second tone, I imagine they will fail.

 

Of course these are still rising tones: a "tone" is the perception of pitch, and to people immersed in Mandarin, these are perceived as rising. But in reality, it's only the last part of the sound that rises: the pitch of a second tone is frequently more like 325 or 335 than the 35 that is shown in textbooks.

 

I don't think it will be news to anyone that tones as spoken in Chinese are different from tones as written in textbooks. I was just interested that this appears to be a pretty predictable feature of the second tone in spoken Chinese (as opposed to random variation). Also I'm sure you can find the same observations in these forums, but I had fun playing around, as you can see:

 

Here is 来

 

Here is: 请今天来我家吃饭

(来我 is circled in yellow:)

 

Here is: 毛病

Here is 人 in a sentence. About two-thirds of the 人 is level. And what I find super interesting is that the rise only starts on the "n". All of the vowel is flat, it's just the consonant "n" that rises:

 

 

Here the bottom red-dotted line and the second-from-bottom red-dotted lines clearly change after the second green line that I painted in: that change (the second green line) is where the "n" starts. And the blue line, the pitch, only starts rising on the "n".

 

For completeness: here is the 又有人感冒了 part of that sentence: the rising part about midway is the end of the 人 sound:

 

Same with 红色:

Here the green line I added has the "ho" on the left and the "ng" on the right (you can see the difference of the red dotted lines). This makes clear the rise only starts after the "o" is finished.

 

 

Finally, here's a rising tone (high rising)from Cantonese (which is the only reason I started playing with this in the first place):

好衡 hou2 hang4

 

 

[realmayo]

More very tentative observations: I assume it's all nonsense but I would like to know how/why:

 

The simplest sounds, which I think are consonant + monothong, the following may not apply to.

 

But for the majority of syllables in Mandarin which are 'longer' ... i.e. they [consonant + diphthong], or [consonant + monothong or diphthong + final] ... it seems to be the case that:

 

2nd tones: tend to start level (or even dip), then rise

3rd tones: tend to fall immediately, then level out

4th tones: tend to start level*, then fall

 

So on this, it looks like 2nd tones and 3rd tones are opposites of each other (steady then move versus move then steady)

 

This also fits with the idea that there are two registers in Mandarin, a higher register (3 to 5 on the 5 to 1 scale)and a lower register, with Tones 1, 2 and 4 belonging to the higher register, and Tone 3 belonging to the lower register.

 

So:

2nd tones are saying: I below in the high register (i.e. #3 on the 5 to 1 scale) and now that we've established that, I'll go higher.

4th tones are saying from the start: I'm a higher register tone, and now that we've established that, I'll go lower. 

3rd tones are saying: ok maybe my very first starting point is a bit ambiguous but I'm going firmly down into the lower register ASAP, and this is where I belong.

 

I don't know how obvious this is when the tones are pronounced in isolation.

But coming in clear spoken Chinese, rubbing up against other tones, the effect seems clearer. I guess tones have to work harder to identify who they are when they're running on from a completely different tone and therefore aren't in their most natural 'starting place'.

 

*actually more often: start gently falling, then change to sharper fall

[realmayo]

Just to back up my sense that longer syllables such as "ying" in tone 4 will start level and then fall: here is 反应

 

1. 反应  followed by 应 on its own

 

2. the first half of 应 (the "yi" part) repeated 5 times followed by the second half (the "ng" part) repeated 5 times.

 

3. those two sounds stretched out and each repeated 3 times.

 

Here's 反应 in Praat, with some very crude attempt to show the sounds in red letters, and also the green line marking roughly where I cut 应 in half.

Can see the third-tone 反 falling immediately, argubaly it starts to rise on the "n". And then can see the fourth-tone 应 staying level until the "ng" kicks in, when it falls (can ignore the little twiddle at the very end).

 

 

 

Edit: I already realise that at least part of this is not entirely right: on the slowed-down "ng" I can still hear some "i" sound at the start and it's clearly falling, so, although most of the "i" sound remains level in "ying", I think it's roughly correct to say that the pitch starts to fall during the last quarter of the "i" sound....

 

 

[Jan Finster]

Thanks.

(I am not too surprised about your findings (e.g. 红色: since most of the "sound" is in the vowel, I am not surprised it only goes up at the "o") 

 

I did experiment with Praat 2 years ago, but could not really find a useful translation of the curves to real life speaking. Likewise with your curves.

 

The main questions I have is:

a) when you say 反应, do the curves look totally different from that of a native speaker? 

b) would foreignners, who studied Chinese by parroting/shadowing speak instinctively copy the native curve or would their curves look "textbook-ish"?

c) have you analysed the same words from different native speakers in different contexts and do the curves all align?

 

 

[realmayo]

On 3/20/2022 at 6:40 PM, Jan Finster said:

 红色: since most of the "sound" is in the vowel, I am not surprised it only goes up at the "o"

 

Just to be clear, it doesn't go up at the "o" of 红! The "o" is flat (actually it dips). That's the part to the left of the green line. It goes up after the "o".

 

a) They look different: I don't descend soon enough for 反, and my falling-pitch on 应 looks too drastic.

b) If people study that way they should reproduce native-like sounds, more or less, that's the plan right? But if they're told they still don't sound right, they might go "back to basics" and tell themselves "tone 2 is a rising tone" or "tone 4 is a falling tone".

c) Yes, a bit. Curves don't precisely align of course but so far the trends are there (for the delayed start to tone movement in 2 and 4 tones, and immediate movement in tone 3).

 

[YuehanHao]

Thanks, realmayo.  Enjoyed reading this topic. 

Very cool program.  I am not good in Chinese but had to give a try myself.  All along, I have just tried to mimic, without thinking much deeper.  I suppose that is okay, but it is human to wonder, and thus interesting to see it in such a light.  While I won't post any of the junk I did, it was a lot of fun, and I hope it doesn't detract any if I assert having seen some qualitative similarities to your observations above.

[realmayo]

Glad you found it interesting @YuehanHao!

 

But it turns out all my playing around with Pratt was a waste of time - I could have just looked at the wikipedia entry for Chinese tones!

They show the tones exactly like what I found above:

 

https://en.wikipedia.org/wiki/Standard_Chinese_phonology#Tones

 

Wikipedia also says that it's the hardest tone for Mandarin students to produce, and that our 2nd tones often sound like 3rd tones, when spoken before a pause.

This was my experience for a long time.

 

I think the reason we can make it sound like a 3rd tone is because in our mind we're thinking that we must pronounce the 2nd tone so that it rises right from the start, and carries on rising right to the end.

That means a lot of rising! So we figure we'd better start really low. And we force our voices lower, to give us enough "room" to say the entire sound in a rising way.

Forcing our voice lower makes its sound like the beginning of a 3rd tone.

But once you realise you can start off low-ish but level, and then only start rising when it feels comfortable to do so, then the 2nd tone becomes way easier and I think more  natural-sounding.

 

 

From wikipedia:

 

Second tone is a rising tone. It is usually described as a high-rising (˧˥ 35), with the sound that rises from middle to high pitch (like in the English "What?!"). It starts at around 3 or 2 pitch level, and then rises towards the level of the first tone pitch (5 or 4).

 

It may also start with a falling or flat segment, which is quite short in male speakers (a quarter of the total second tone length), but longer in female speakers, reaching nearly half of the total length of the second tone. This initial dip is more apparent in Southern China Mandarin accent, including Standard Taiwanese Mandarin, where the second tone is also lower and alternatively described as dipping or low-rising with overall contour of ˧˨˧ 323 (its start is still slightly lower than its final pitch).^{[24][25][26][27]}

 

This tone is usually one of the most difficult to master for Mandarin learners, as well as the speakers of non-Mandarin Chinese varieties, who often pronounce their second tone close to (full) third tone, especially in the word-final position before a pause.^[