Tones have stress information

[chalimac]

Mandarin has stress. This is a much neglected observation, but most bi-syllabic words have a distinctive stress. I came across the paper Tone Pattern and Word Stress in Mandarin that documents how tones are correlated with stress.

The researchers experiments with continuous speech could be summarized in this simple rules to find where the stress lies:

1. If there is a T1 in the word it coincides with the stress:

他们 TAmen

危险 WEIxian

聪明 CONGming

but

国家 guoJIA

年轻 nianQING

好吃 haoCHI

2. If the word is made of two T1, there are two equally distinctive stresses:

发音 FA YIN

伤心 SHANG XIN

3. If there is no T1 in the word, the stress is in the first syllable:

容易 RONGyi

重要 ZHONGyao

快乐 KUAIle

... and tons of words

Why the first syllable is emphasized in so many words is explained in another paper called Quantitative measurement of prosodic strength in Mandarin

If you hear the beginning of a word you have more information about the next syllable. On the other hand knowing the last syllable of a word is not as helpful in predicting the beginning of the next word [...] This is part of the explanation of the higher prosodic strength we obtained in the word initial position

[Shi Tong]

There is stress in Chinese.

I would say that the rules that you have mentioned here seem to appear less in Taiwanese guoyu, IMO.

Quite a lot of people here have mentioned the lesser extent of this kind of stress in Taiwan, and quite a lot of instances where qingsheng would apply in China, they do not appear as much in Taiwan.

There is still certainly stress, emphasised with sentance end particles in Taiwan guoyu though.

[gougou]

容易 RONGyi

I would have thought that one to be the other way around.

[buanryoh]

Makes sense to stress (or pronounce with clarity) the first syllable as a way to make sure the listen understands what you are saying - particulary when stating a word that shares the same syllables with many other words yet has different tones.

[lokki]

2. If the word is made of two T1, there are two equally distinctive stresses:

In the case of Tian an men (T1+T1+T2), what I hear is the following stress pattern:

天安门 tian AN mEN

ie. I hear a very clear main stress on the second syllable (first tone) and some secondary stress towards the end of the third syllable (second tone) while the first syllable (first tone) feels completely unstressed in comparison to the rest of the word. I also imagine I hear the second syllable at clearly higher pitch than the first, though both are first tone.

(Note: Here I am talking about the standard Mandarin pronunciation, such as can be heard in audio recordings of language courses or listening to bjradio on the net.)

Am I wrong?

I also think it is very tricky to talk about stress and tone as separate entities, at least the way we understand the concept of stress in non-tonal languages such as English. We use pitch patterns, as well as volume and duration, to achieve stress, meaning tone is part of what we perceive as stress.

Maybe this is why I also have a hard time hearing 容易 as RONG yi. I too hear it as roNG YI ie. with the main stress on the second syllable and some secondary stress towards the end of the first syllable. I am pretty sure this is because the fourth tone has a pitch pattern that agrees with what I normally identify as a stressed syllable.

It would be interesting to see exactly how they determine stress, as opposed to tone, in that report. I'll make a note to read it through in detail some time.

Edited March 31, 2010 at 01:58 PM by lokki

[gougou]

In the case of Tian an men

The original poster only seems to be talking about bi-syllabic words.

[Shi Tong]

In Taiwan Tiananmen is pronounced with a similar if not the same stress for every word.

The only thing which I think is noticeable in two T1 words in a row is that the second T1 tone lowers slightly.

[lokki]

The original poster only seems to be talking about bi-syllabic words.

True enough.

[chalimac]

The rules I proposed are a rough approximation. They would cover most cases, but not all, because stress in mandarin is perception based and not inflexible.

In the case of 容易 we have the pattern T2+T4. According to the study this pattern will yield:

- 86% of words with the stress in the first syllable.

- 14% of words with the stress in the second.

I have checked native recordings of 容易 and there is some variation:

This seems to stress the first:

http://hua.umf.maine.edu/Chinese/Language/Sound4b/4908lyz.wav

This the second:

http://hua.umf.maine.edu/Chinese/Language/Sound0c/755cl.wav

But in connected speech most emphasize the first:

汉语不容易学

http://hua.umf.maine.edu/Chinese/Language/Sound9a/9152lyr.wav

http://hua.umf.maine.edu/Chinese/Language/Sound9b/9152jz.wav

http://hua.umf.maine.edu/Chinese/Language/Sound9d/9152hjl.wav