Interesting corrective on tones in normal Chinese

[realmayo]

I came across some youtube videos setting out very clearly the ways in which Chinese tones, when spoken naturally, are different from what the textbooks tell us. I think most people would pick these rules up naturally over time, but maybe discard them when they are trying to pay particular attention to pronunciation!  What do people think?

 

They are from an online teacher called Ben, aka Learn Chinese With Ben. Youtube playlist here: https://www.youtube.com/watch?v=Ty4eoBtsIFY&list=PLW4RM43CAPvyhxq-whqmKI1LFpvqS5_XR&index=1

 

Basic rules guidelines

Last sound in a chunk is stressed (unless it's a neutral tone, in which case it's the previous sound that's stressed)

That extra stress is created by: higher pitch and longer duration

All the sounds apart from the stressed sound generally have lower pitch, shorter duration, and less tonal expression (i.e. less 'height' or 'range' in e.g. second tone)

 

 

Three layers:

 

Most sounds come in what you might call three "layers": 1st layer is 4 to 5, 2nd layer is 3 to 4, 3rd layer is 2 to 3.

Basically: unstressed is 3rd layer, stressed is in 2nd layer.

But: one 'chunk' in a sentence will itself be emphasised (below: 吃中国菜): that emphasis comes by moving all the sounds in that chunk up one layer, which is where the 1st layer comes into play (the 吃 and 菜 below):

 

 

 

The two screenshots are from the youtube videos, the rest is my attempt to summarise, which may well be wrong! I recommend having a look.

[sanchuan]

Oh, if only. And I truly mean it. It's just so much more complex than that. 

 

But let me first thank you for taking the time to drag this out of the youtubes and review it for us. Always glad to see issues of prosody brought up to a wider audience and it was good to watch. 

 

And that's because it was well made. He made it all clear and easy to follow (except for the angry fruit salad of colours on his slides...). He's defined a nice few rules, as you summed up, which I think are gonna make excellent cruches for the beginner stuck in tone-by-tone pronunciation. He deserves more exposure. 

 

Past beginner levels, however, those 'rules' risk becoming a new straightjacket. There's no clear consensus about them: let's just think of San Duanmu's life work on these issues which, briefly speaking, shows words are always trochaic and only distorted by utterance(chunk?)-final lengthening). Questions over whether modifiers are accentuated in declarative contexts, or otherwise, is either inconsistent or often throws into question what we should consider a modifier in Chinese. If you also consider that neutral declarative sentences are really difficult to find in the wild and that every other clause will feature words with some level of contrastive/emphatic accentuation (which flips the rules upside down), then what you're left with is a mess of contradictions and exceptions. 

 

I concur with the take-home message though: tones are flattened in the flow of speech and only emerge in their full range when receiving a primary or secondary (or tertiary, he seems to say?) sentence stress, which is determined by semantics and information structure.

 

But to go on setting down syntactic rules for it, eg that verbs like 是 are normally always stressed fully (which we know is only true in very limited declarative contexts) means raising more questions than you're answering. Now, I'm personally 100% keen on going down such rabbit holes, but most will just take these 'rules' at face value. And before you know it, eager beavers will add them to online grammar wikis (or Instagram graphics?) and we'll have yet more zombie textbook rules.

[vellocet]

On 10/9/2022 at 12:27 AM, sanchuan said:

except for the angry fruit salad of colours on his slides...

Those are the commonly used tone colors, used to aid memory.

 

On 10/9/2022 at 12:27 AM, sanchuan said:

Past beginner levels, however, those 'rules' risk becoming a new straightjacket. There's no clear consensus about them

Who cares?  Beginners really need this stuff, because the usual textbooks (like anything produced by BLCUP) don't just fail to teach it, they straight-up lie to students.  And 95% of all students of Chinese never make it past beginner level.  You're talking from the clouds, very few achieve the mastery that you do. The rest of us need rules.

[fabiothebest]

On 10/8/2022 at 2:20 PM, realmayo said:

I came across some youtube videos setting out very clearly the ways in which Chinese tones, when spoken naturally, are different from what the textbooks tell us. I think most people would pick these rules up naturally over time, but maybe discard them when they are trying to pay particular attention to pronunciation!  What do people think?

 

They are from an online teacher called Ben, aka Learn Chinese With Ben. Youtube playlist here: https://www.youtube.com/watch?v=Ty4eoBtsIFY&list=PLW4RM43CAPvyhxq-whqmKI1LFpvqS5_XR&index=1

Thank you for posting this. Ben 老师 is right, we often are taught about tones in isolation without considering whole sentences. I'm going to watch all the videos in the playlist, I hope I can improve. I will need to practice reading and speaking more though and I hope some native speaker will correct my mistakes.

[realmayo]

On 10/8/2022 at 1:20 PM, realmayo said:

The two screenshots are from the youtube videos, the rest is my attempt to summarise, which may well be wrong!

Thinking about it, I shouldn't have used the word "rules: - "guidelines" would have been better. I can't remember how Ben老师 characterises them.

 

On 10/8/2022 at 8:47 PM, vellocet said:

The rest of us need rules.

 

On 10/8/2022 at 5:27 PM, sanchuan said:

Past beginner levels, however, those 'rules' risk becoming a new straightjacket

This is a risk, as you say, but I think it's one well worth taking.

 

Personally I think beginners need to learn the 'robotic' pronunciation first: they need to learn the pitches and pitch contours associated with the textbook tones, for words spoken in isolation. But I reckon that after a certain point, perhaps when students are being usefully exposed to plenty of listening material, the guidelines I posted above need to be explicitly taught.

 

Because the bigger risk is that the brain, taught the textbook tones, refuses to believe what the ears are hearing.

 

And if the best way to develop good pronunciation is listening plus mimicry/shadowing, then students need to be 'given permission' to reproduce what they are hearing. And the first step is explaining why and how natural speech differs from textbook tones.

 

I kind of tested myself and like everyone else I guess I have already absorbed those guidlines above. But I'd never previously identified them. That means that where I'm specifically trying to improve my pronunciation, I'm actually sounding less natural and harder to understand. Also, having identified them in theory, I now automatically identify examples of those guidelines-in-action all the time in real life.

 

(In English the equivalent might be: first teaching people how to pronounce "t", but later reassuring them that a native speaker saying "I just don't want to do it tomorrow, OK?" isn't 'wrong' and, once you've mastered the other sounds, is probably worth imitating.)

 

Has anyone seen these guidelines or something similar being explicitly taught to students, and to beginner-ish students in particular? I have a feeling that most Chinese speakers, and perhaps many Chinese teachers in China, would initially struggle to accept them.

[sanchuan]

On 10/8/2022 at 9:47 PM, vellocet said:

Beginners really need this stuff, [...] You're talking from the clouds, very few achieve the mastery that you do. The rest of us need rules.

I was very quick to say this is excellent content for those beginning to speak Chinese but stuck pronouncing things tone by tone. 

 

But, yes, I also noted upper intermediate students to advanced (and above!) may benefit from similar (but different) guidelines. This is an important point to make, and I'll stand by it, because the industry and the web are absolutely awash with resources for Mandarin beginners and the scarcity of attention paid to middling intermediates is a more frequent reason to quit than the insufficient number of prosodic rules given to beginners.

 

On 10/9/2022 at 7:50 AM, realmayo said:

Has anyone seen these guidelines or something similar being explicitly taught to students, and to beginner-ish students in particular?

I don't know about other YouTubers or all the newfangled language schools about but otherwise, sure: the last chapter of Chinese: A Comprehensive Grammar, by Yip Po-Ching and Don Rimmington, goes over all the features explained in this video and more. It's more for early intermediate and above, I'd say, but there you go. And that's from 2004. 

[sanchuan]

On 10/8/2022 at 9:47 PM, vellocet said:

On 10/8/2022 at 6:27 PM, sanchuan said:

except for the angry fruit salad of colours on his slides...

Those are the commonly used tone colors, used to aid memory.

True! Thanks. Crass of me not to notice, I suppose. Mixing tones and colours never really worked for me but I know there's a standard developing there so it's great that Ben 老师 should use it.

[malazann]

big ups Mayo. love the first video. I knew about the general concept of "destressing" but the more you are exposed to an idea in different ways, the deeper you begin to understand them.

I came across this series on this forum but I'll share it since its related (by SpongeFlower on word level stress and sentence level stress) - https://www.youtube.com/playlist?list=PL6DUmreytzopaSKrO--EaVbue7OIpPCRa

[vellocet]

On 10/9/2022 at 2:40 PM, sanchuan said:

the scarcity of attention paid to middling intermediates is a more frequent reason to quit than the insufficient number of prosodic rules given to beginners.

I think it has much more to do with the fact that most learners of Chinese never make it to intermediate level.  Thus, applying emphasis to beginner material is correct.

 

They don't quit because there's no material.  They quit because Chinese is far harder than they ever thought possible; they leave China; they get a new hobby other than studying Chinese; or they attain a rudimentary grasp of Chinese enough to order at a restaurant and don't want to continue the torture.

 

Even on Ben's video it states: "Goal: to become a fluent Chinese speaker!"  I think this is incorrect and the vast majority of people who learn Chinese learn it so that they can communicate with Chinese people.  Fluency is a difficult, distant goal and no matter how far you go there's always more to study.  Obviously this doesn't apply to academics, and guess who writes the textbooks?

[anonymoose]

On 10/9/2022 at 7:40 AM, sanchuan said:

I don't know about other YouTubers or all the newfangled language schools about but otherwise, sure: the last chapter of Chinese: A Comprehensive Grammar, by Yip Po-Ching and Don Rimmington, goes over all the features explained in this video and more.

 

Which video are you referring to? I watched the video linked to in the first post which was mainly about pronunciation, of tones in particular. The last chapter of the book you referenced is on "stylistic considerations in syntactic constructions". It doesn't mention pronunciation at all, let alone tones.

[sanchuan]

@anonymoose

Well, quite. It was the second-to-last chapter:

 

Prosodic Features

26.1 The overall rhythm of Chinese speech 

26.2 Syntactic sequences and their underlying rhythm 

26.2.1 'Verb + object'patterns

26.2.2 The 'attributive + headword' pattern

26.3 Echoing patterns of rhythm 

26.4 Expanding, condensing and padding to get into the

appropriate rhythm 

 

The videos in the OP discuss tones in the sense of reductions or accentuations based on rhythm and syntactic sequences, which is what that chapter is all about.

 

The videos only focus on the general rhythm of the sentence, though: wherever the teacher decides to pause in a sentence, thereby causing a pre-pause lengthening in the preceding syllable, there you'll be 'ruled' to utter a stressed tone, which will be one of at least two or three in a sentence. In other words, there's a logic to the rules in the videos that in large part rests on how to chunk a sentence, yet how to chunk a sentence is left unexplained and to the student's discretion, as if you could chunk anywhere - which you can't. For example, some confusion may arise when, as I recall, you're taught that 非常帅 is always feiCHANG||shuai, only to be encouraged immediately after with a hearty 非常好 pronounced as feichangHAO||.So in my opinion the rules are a bit hard and fast. He seldom connects the rhythm of a 'chunk' to its syntax (eg words, sub+V+Obj, clauses etc are all often simply called "words") and, where he does, it feels too absolute (eg is it true that degree adverbs always receive primary sentence stress? Is it always "HEN hao" or "HEN piaoliang"?).

 

So maybe these videos offer more 'rules' than other sources, but I'd be a bit wary of taking them all at face value when the underlying logic is contradictory or unexplained or just frankly difficult for even experts to pin down. What the videos do make wonderfully clear to the student, though, (which is rarely presented as explicitly, including in that grammar book I mentioned) is how the pitch of tones is smoothed out in conversation by often starting the following tone from where you left off in the previous one (eg 3 second tones represented one above the other, etc). But then we're moving away from material that's beneficial to "beginner-ish students", as said in the OP: I'm not sure someone who's only just learning to keep tones distinct in conversation should be encouraged to rush through to the final syllable just yet. 

 

But I think for intermediate students and above this is really good content. Whether looking for rules or exceptions, it encourages you to pay much more attention to tones and sentence rhythm and that's all for the good.

 

[realmayo]

As well as wrongly saying "rules", I'm also thinking my use of the word "stress" was ill-chosen! Instead I should have written something like "full tonal realisation". In contrast to a separate feature perhaps called "rhythmic stress".  Arguably the former is pre-pause and the latter is post-pause (very short pauses)?

 

Interesting that the foremost experts can't agree on lots of this stuff.

 

On 10/10/2022 at 10:55 AM, sanchuan said:

how the pitch of tones is smoothed out in conversation by often starting the following tone from where you left off in the previous one

 

Definitely: I think a two-syllable word with tones 1+2 is the most dangerous for learners because the textbooks force you into this crazy-up-high first tone, followed by a throat-twisting judder downwards so you can begin a rising-from-the-mid-range second tone. Just start the first tone in the mid-range if you know it's immediately followed by a second tone. Seems to me the only time pitches don't join naturally on from one another is for neutral tones?

 

Speaking of joining-naturally, I think I'm right in saying that, often, in a two-syllable word, the tone of the second syllable will 'reverse-bleed' into the first syllable? That is, the tone expected from the second syllable will actually have started by the end of the first syllable. Everyone knows this 'forward-looking' feature occurs with e.g. two third tones in a row, or working out which tones to pronounce 不 or 一, depending on what tone follows them. But I assume it's a feature of all sounds, not just those ones.

[sanchuan]

On 10/10/2022 at 2:28 PM, realmayo said:

the tone of the second syllable will 'reverse-bleed' into the first syllable

Yes, I agree. I think that's another way of saying that the contour of the whole lexical item (or indeed any longer chunk) is determined long* before production so the first syllable appears already set "in the key", if you excuse my analogy, of the whole chunk. As you say, that's what makes tone sandhi possible, whether it's in a string of three tones, in a phrase starting with 不 or 一, in feiCHANGhao turning to feichangHAO, or even in English ,real 'estate turning into ,realestate 'market. But I wouldn't necessarily say that the suprasegmentals (tones, stress, accentuations, intonation, etc) of some segments "bleed" back into other segments. I'd say they form one prosodic contour per chunk of speech and it's this contour that governs the range/pitch/etc of all the segments within: unstressed items will be made to accommodate stressed ones and so forth.

 

That's why it's all down to how you chunk it. For example, as has been said many times, it's not true that all the syllables in a string of multiple third tones will always accommodate the last one, because it depends on how you parse the string. I just happen to think that's also true of some of the rules in the videos that started this discussion.

 

* In synaptical terms!

[Moshen]

I'm wondering if people who are excited about these ideas believe you'll be able to apply these "rules" while you are speaking?  Wouldn't it be more effective simply to imitate native speakers - who, I'm sure, do not sound exactly the same as one another?  After all, the native speakers aren't following any rules in their head when they speak.  In addition, I'm sure that the "rules" bend quite a bit according to what a person is trying to say and emphasize in each sentence.

 

For me, this discussion reminds me of when I was studying classical flute;  my teacher and I spent a lot of our time working on what he called "phrasing," which is the equivalent of how tones shade into one another in actual Chinese speech.  He hardly ever talked to me of any rules.  Instead, he demonstrated different ways to play something and gave me imaginative analogies.  The result: a musical line that sounded more natural and meaningful.

 

I suspect that Chinese learners would sound more natural getting trained to listen and implement what they heard rather than try to apply these "rules."

 

(Sorry to be so skeptical...)

[Flickserve]

It’s good for background knowledge.

 

You will end up doing it anyway if you listen and shadow a lot. 

[realmayo]

On 10/10/2022 at 9:01 PM, Moshen said:

I suspect that Chinese learners would sound more natural getting trained to listen and implement what they heard rather than try to apply these "rules."

 

Do you regret that you were taught the 'rule' that in two consecutive third tones, the first one becomes a rising tone?

 

If learners find that 'rule' useful, wouldn't they also find it useful to know that with two consecutive second tones, they don't have to go up and then down and then up again, but can just form one single upward rise? At what point in your study of Chinese did you realise you don't have to pronounce most of your first tones high - and might it have been beneficial to have learned that a little bit earlier?

 

Humans wouldn't be able to enjoy music or science or stories - or learn language - without being interested in patterns (although high interest in patterns is typically male and extremely high interest in patterns is typically autistic, I think).

 

Personally I think identifying these patterns/rules is useful because

(a) they can be a shortcut for beginners to have better pronunciation, and

(b) they encourage you to trust your ears when paying real attention to what you're hearing ("yes you heard right, that third tone did turn into a rising tone and it wasn't random but part of a pattern associated with two consecutive third tones, therefore continue mimicking this sentence, it's good Chinese").

 

... and that should speed up of internalising them so they become second nature when you speak and you forget that there is actually a rule after all. After all, most people stop automatically at a red signal and go automatically at a green one, without thinking about any traffic lights rule, although they did once have it explicitly explained to them.

[sanchuan]

I must second @realmayo's answer, though I understand where @Moshen's scepticism comes from.

 

These discussions are somewhat academic. But that's the whole point. (Unless you think research and detail is useless or that some details are more important than others - which isb just not how science works or moves forwards).

 

Aside from the all-important benefits of this for AI, text to speech, voice recognition and the like, this aspect of language is just as important for language acquisition.  Knowing some rules (and I only mean the couple we know hold true, not the morass of random guidelines given in those videos perhaps) can massively improve not just production but also listening - sometimes almost overnight. 

 

We all met non-native speakers of English for instance in the UK or the US who get by in the language but seriously struggle to understand or make themselves understood by native speakers, no matter how many decades of "baby-like language immersion" they've spent in the country. Studying such rules would have made, and does make, a massive difference for them because, to answer your question, yes, things like rhythm and scansion are actually very easy to learn and apply, if you know the basic patterns. 

 

You wouldn't say the same about grammar, would you - rules like knowing a verb from a noun. You wouldn't ask if learners will ever be able to "apply these rules while they are speaking", as you say. Why think phonetics is any different? It's certainly not less important: it makes output understandable and input comprehensible. Isn't that what language is all about? Why not make it easier on people by teaching them how?

 

We actually all agree that such aspects of language should only be taught with a couple of fun, general and digestible guidelines. It's just... finding few easy rules that can solve many hard problems is a difficult process (precisely because of that variation in speech you mentioned), but it's not an impossible or useless endeavour. For English, for instance, we pretty much have it down pat.

 

Yes, the detailed discussions to get to them may get impossibly knotty. But not all learners need to see that. 

[Moshen]

OK, so if you believe you're able to apply these rules, I accept that.

 

But I don't agree that this nitpicky approach is something that needs to be taught to everyone, or that longstanding pronunciation or intonation problems would necessarily disappear from being taught these rules.

 

Many people learn more holistically and would better improve from listening, imitation and correction. 

 

Besides, many people speak understandably even if they don't sound exactly like native speakers.  (Others make mistakes that prevent them from being understood.  Some kinds of mistakes are more consequential than others.)  Is the goal to sound like a native speaker of Chinese or just to communicate well and correctly and be understood? 

 

 

[realmayo]

On 10/11/2022 at 12:13 PM, Moshen said:

Many people learn more holistically and would better improve from listening, imitation and correction. 

 

I agree, that's how I learn, and I think awareness of these established patterns (rules/guidelines/observations) of native speech could help students and teachers alike, if introduced at right time.

 

For instance: one reason students' tone 2 can sound like tone 3 if it follows a tone 1 (or follows a tone 2) is because a native speaker would end the preceeding sound mid-range, but a student will end it high. So a native speaker naturally runs the pitch of the first syllable into the pitch of the second syllable, but the poor student feels obliged to go up, down and back up again - producing what the native speaker hears as tone 3, not tone 2.

 

So in fact this 'rule' is actually saying: 'you should rethink a little those rules about tones you were taught on Day 1'. Of course if you advocate never being explicitly taught about tones in the first place, but just picking them up (and I quite like the sound of that) then fine, none of this applies!

 

 

Edit: I still think the prime use of these guidelines is to help you make sense of what you hear.

[sanchuan]

On 10/11/2022 at 1:13 PM, Moshen said:

I don't agree that this nitpicky approach is something that needs to be taught to everyone

We are in agreement that academic "nitpicking" aimed at understanding how language is spoken is different, and best kept separate, from instructional guidelines aimed at learners. This thread is just indulging in a bit of the former in order to review an instance of the latter.

 

On 10/11/2022 at 1:13 PM, Moshen said:

Many people learn more holistically

Absolutely. That is not only necessary, but may even be sufficient! Understanding patterns of speech more explicitly is only helpful in addition to that - to correct and/or improve. But  I believe it's a demonstrably beneficial addition in those cases.

 

On 10/11/2022 at 1:13 PM, Moshen said:

Is the goal to sound like a native speaker of Chinese or just to communicate well

That sounds like a false dichotomy. You can do lots to improve the latter before you're even close to achieving the former.

 

This whole exchange frankly feels a bit like a false debate, too. Didn't you just say that a holistic approach should be correction plus imitation? How's that different to what we're talking about? All we're saying is it's best to correct patterns (where research can find any), rather than single words at random. After all, patterns is all your brain is ever going to learn - the question is whether to learn some more explicitly than others. You're right to say that not all learners will benefit from only explicit pattern instruction just like I'm right to say that not all learners will benefit from only random ad-hoc corrections.