Near-Native foreign accents

[leosmith]

Hi all,

I've been doing some poking around, trying to find a good way to start my studies next year. I'd like to develope an understandeable accent in Chinese.

There seems to be a lot of scams out there about perfecting foreign accents. It's possible some of their methods work, but I found lots of sites that use a lot of hard-selling techniques and "overly ridicule" traditional learning techniques.

Do you know of a good technique for acquiring proper Chinese pronunciation? If so, post it here!

Here is my favorite so far. This gentleman is on the leading edge of perfecting foreign language accents. (Notice his credentials, and the lack of products for sale on his site. )This is very technical, but also very useful.

http://olle-kjellin.com/SpeechDoctor/ProcLP98.html

I found it interesting that he is aware of the "natural" method, like most of these sites claim to be, but doesn't suggest silly things like sitting idle for 400 hrs before you start to talk.

[geraldc]

The method they discuss in that paper, i.e. an entire class repeating phrases in chorus, was the traditional way that Chinese was taught. It was the way my dad learnt to read 50 years ago etc. However I don't think adult foreign learners are happy with that method of learning.

To be honest most accents are understandable, there are very few Chinese with standard accents, and in the same way, a foreigner shouldn't be embarrassed with a non standard accent. As long as you're understandable, I don't really see the need to try speak the same as the locals. Americans never feel the need to pick up an English accent and vice versa.

[skunkpuppy]

In my view, the question of whether a non-native speaker tries to pick up the native accent or not is usually a social/racial/cultural question.

From my experience, in my country --Singapore-- English is widely taught and spoken, although it is not the native tongue for the population. So English is spoken with a strong local accent.

Interestingly, when locals come into contact with caucasian foreigners-- who need not necessarily be American, British, or any sort of native English speaker-- many locals feel the need to affect a British/American accent. The result is often a very unfortunate and sometimes unintelligible accent comprising local/American/British influences.

This phenomenon is widely seen in many Asian countries, when asian locals come into contact with caucasians. My view is that is, consciously or even unconsciously, the effect of the colonial mentality of "white is right."

More broadly, it stems from a very common and natural phenomenon that occurs in all conversations. When two people talk, often they will begin to modulate their speech patterns to resemble each other. This might happen consciously or unconsciously, but I figure that it is really an effort to make both parties feel more comfortable. The effect with two people of similar linguistic background is not very dramatic, but still detectable.

However, with regard to the adoption of faux accents, I think a large component is the perception that English spoken with a non-British/American accent is socially inferior. Many locals in my country justify themselves by saying that they only do it to sound more comprehensible to the other party, or, if they are in US/UK, because they wish to "do as the Romans do."

Neither reason is really plausible. First, they are often even more incomprehensible after attempting the complex process of adopting a totally foreign accent, and often sound nothing like any native English speaker. Second, the reverse phenomenon is not usually seen. British/American speakers in Asian countries, speaking English to Asians, feel no need to adjust their accents to sound more local, even when they are not always easily understood by the locals. The most that would happen is that they would speak sloooooowly and LOUDLY.

So I say, it's a matter of cultural confidence that affects which accents people choose to imitate.

[gato]

As geraldc said, it is a complicated trying to sound like a native when learning Mandarin because there are many native accents. A popular approach to pick the Beijing accent and try to perfect that. But as long as you can get the tones and consonants right and speak clearly, you'll be understood.

Singporeans' English accent seem to be very much influenced by how Chinese is spoken. They seem to stress almost every syllable and use a different intonation than what the North Americans, Brits, and Aussies would use. This is only natural, of course. It's similar to how Indians and Pakistanis would speak English with the intonation of the local dialect.

However, it's interesting that most HKers I've met don't have a Singapore accent, that is, they don't speak English with Chinese cadence and intonation. Maybe it's because most Chinese HKers don't speak English as their primary language and can therefore keep English and Cantonese in separate mental compartments. Many Singaporean Chinese, in contrast, aren't fluent in Chinese and do speak English as their primary language. Maybe as a result, they can't keep the two languages separate as easily, and it feels quite natural to speak English using Chinese intonations.

Or perhaps, more simply, it's because in HK the more educated do try to imitate the UK or American accent, whereas in Singapore because English is a native/home language for many, they don't try to imitate the UK/American accent.

[bomaci]

I found it interesting that he is aware of the "natural" method, like most of these sites claim to be, but doesn't suggest silly things like sitting idle for 400 hrs before you start to talk.

I actually tried the method he proposes about a year ago, but I think it is a bit frustrating for the self learner to only work on a sentence at a time in such a manner so I gave it up.

However I don't agree that listening for 400 hrs before you start talking is silly. You are in fact not sitting idle, you are practicing your listening comprehension. Both the natural method and Olle Kjellins are bascially about the same thing. To build a solid image of the intonation of the language in the learners mind. Also note that even with Olle Kjellins method the learners first listen to the sentence a couple of times before saying it. I would say that the best method to develop native-like pronounciation is to get a course that comes with a lot of recorded dialouges (Teach Yourself, Assimil, Collouqial, etc), then strip away everything but the dialouges from the recordings. Then listen to these over and over many many times. When you understand all the dialouges in the course, you could proceed to simultaneously repeat along with the recording.

I would also advise that if you use this method try to avoid all explanations of tones like the plauge, first train your hears to listen to the language. Then when you have already been exposed to alot of input you can read explanations of how the tones work and which tone is which.

[rose~]

I myself speak Mandarin with a slight Taiwanese accent, of course it's Taiwanese mixed wth foreigner and never thought anything of it.No-one had ever said they didn't understand me. Basically I suppose I just say "si" a bit instead of "shi", and other typical stuff like that.

I came to Shanghai to work and in one company which interviewed me, the boss (Shanghainese) said I "had a thick Taiwanese accent" and he was VERY clear that I was expected to lose it. It was not a pleasant experience and he was pretty rude about it.

To those of you who will be employed in the future on the Mainland I would say consider learning a Beijing or a neutral accent.

[leosmith]

The method they discuss in that paper, i.e. an entire class repeating phrases in chorus, was the traditional way that Chinese was taught. It was the way my dad learnt to read 50 years ago etc. However I don't think adult foreign learners are happy with that method of learning.

Agreed. I wouldn't want to learn that way for however many years it takes to learn the language. But an hour a day for 4 to 6 weeks, as he suggests in his paper, sounds tolerable. I won't be able to do true chorus repetitions myself, because I'm not in an area that's condusive to learning Chinese, so I'll probably have a tutor say a phrase, then repeat it with her/him.

One of the things I really liked about the article was his theory that all the pronunciation one needs to learn can be contained on a single sheet of paper. Another thing I liked was his suggestion that the teacher say the repetition phrases differently, to allow the student to learn the acceptable range. Good stuff.

However I don't agree that listening for 400 hrs before you start talking is silly. You are in fact not sitting idle, you are practicing your listening comprehension. Both the natural method and Olle Kjellins are bascially about the same thing. To build a solid image of the intonation of the language in the learners mind. Also note that even with Olle Kjellins method the learners first listen to the sentence a couple of times before saying it.

Sorry, let me clarify. I agree with you that listening to a sentence a couple times before you try to repeat it makes perfect sense. Sometimes I need even more than that just to get it clear in my mind. What I find silly is the 400 hours. 40 seconds maybe, but 400 hours? Imagine what you could accomplish in 400 hours. You can finish Olle's method in 20 to 40 hours, by my calculations.

I came to Shanghai to work and in one company which interviewed me, the boss (Shanghainese) said I "had a thick Taiwanese accent" and he was VERY clear that I was expected to lose it. It was not a pleasant experience and he was pretty rude about it.

To those of you who will be employed in the future on the Mainland I would say consider learning a Beijing or a neutral accent.

Wow, thanks for the tip Rose. Is that guy still your boss? Hope he chills out a little.

[Ferno]

Sorry, let me clarify. I agree with you that listening to a sentence a couple times before you try to repeat it makes perfect sense. Sometimes I need even more than that just to get it clear in my mind. What I find silly is the 400 hours. 40 seconds maybe, but 400 hours? Imagine what you could accomplish in 400 hours. You can finish Olle's method in 20 to 40 hours, by my calculations.

you're misunderstanding. You're not memorizing a Biology textbook or a shopping list. You're learning a language - different brain functions. memorizing anything = bad

children do not start speaking until they have a high % of comprehension due to listening, allowing the sounds to map into their brain. Adults learning a second language start speaking right away, and build on the pronounciation and grammar of their native language. They get permanent accents and bad grammar and have to think about what they say.

note that the natural approach is at least 600 hours, and involves interaction, tone of voice, body language, physical objects - recreating that specific type of learning.

[bomaci]

Sorry, let me clarify. I agree with you that listening to a sentence a couple times before you try to repeat it makes perfect sense. Sometimes I need even more than that just to get it clear in my mind. What I find silly is the 400 hours. 40 seconds maybe, but 400 hours? Imagine what you could accomplish in 400 hours. You can finish Olle's method in 20 to 40 hours, by my calculations.

Well actually listening is the foundation of speaking. If you spend a lot of time listening in the beginning you will definetely get pretty good at speaking. Once you begin speaking you may still have to work on some sounds in the language to get them correctly but your intonation will be perfect. And intonation is excatly where most adult learners of chinese fail. Da Shan is not famous because he pronounces zh, ch ,sh ,c ,z perfectly, it is because his intonation when speaking mandarin is extremely native like. This can only be accomplished by alot of listening to native speakers.

[stephanhodges]

I wrote Olle Kjellin and asked about the 4-6 weeks that was mentioned in the paper. He said:

We certainly switch frequently between chorus and individual practice from the first day on. What I mean is that by 3-4-5 weeks we will gradually need fewer and fewer repetitions per practice phrase or sentence, because if we have done it right from the beginning, the learners will typically have attained a very good, automatized command already, settled in their procedural memories.

I also asked him how many days per week of classtime.

I'm talking about "full time", meaning at least 4 hours daily, 5 days a week.

Lastly, I asked if he had any audio samples of his approach. He said:

In the Swedish start page I have a recording of a short phrase (the student's address) before practice and then the same student's first individual trials after 5-10 minutes of chorus practice. Find a Swedish speaking friend and get a comment on the difference!

[bomaci]

A contribution from your friendly neigbourhood Swedish speaking friend

I've attached a recording of myself (a native Swedish speaker) saying the sentence.

You can hear for yourself how native he sounds.

Here is his pronounciation before: http://www.olle-kjellin.com/SpeechDoctor/BahPre.au

And here is after:http://www.olle-kjellin.com/SpeechDoctor/BahPost.au

adress.wav

[okjhum]

Thanks to stephanhodges for introducing me to this forum. Long time ago I too studied some Chinese, but alas not enough.... I'm flattered that you have found my site and discussed the chorus method. Admittedly, this is a stone-age method, not an invention of mine, but it certainly has several benefits, many of which can be scientifically corroborated, and many of which I have only understood thanks to my medical training on top of a PhD in speech science. (I have listed 11 "heavy" benefits in my presentation when I give talks about this.)

The main effect why chorus practice is so efficient is due to brain anatomy and physiology:

***First, as most of you have surely experienced, it is quite difficult for untrained singers to sing in a different tune or pitch than the others in a choir. Similarly when you sing-along with friends in a "Happy birthday to you...", etc. And you must have felt the phenomenon when you are listening to someone with a creaky or hoarse voice, whereby you get a strange tickle and strong urge to clear your own throat! These experiences are due to two important neuroanatomical and neurophysiological facts that, of course, also apply to _speaking_ in chorus.

(i) In the vicinity of the speech area in the frontal lobe (Broca's area) there are networks of nerve cells called "imitation neurons", or "mirror neurons". They are important in perception processes by matching input data with our previous experiences ("memories") of saying the same or similar things. The best studied mirror neurons are involved in visual perception. Look it up in Google and Wikipedia, and you'll find amazing reports. (Giacomo Rizzolatti is a grand name.) The most amazing thing to note about these mirror neurons is that they are _motor_ neurons! I.e., they are active when the individual _speaks_. And, obviously, when the individual _hears_ speech, if this speech is in a comprehensible form, i.e., he or she has previous experience of saying like things. Otherwise they don't mirror the input. We all have noticed that as new beginners in a language class, haven't we?!

(ii) Direct neural connections from the auditory center in the temporal lobes to the auditory mirror neuron areas have been found. These connections don't themselves pass via the speech comprehension area (Wernicke's area), so the mirroring processes begin even before the listener is aware of them or of the meaning of the input. These mirroring processes in the brain's _motor_ area will directly influence the speech organs of the listener and enable, or even _enforce_, like magically, a very accurate pronunciation as he or she speaks along with the teacher. So even a "chorus" of 2 persons, or 1 cd + 1 learner, is possible, but I have found that 8 learners is a minimum number for maximal efficiency, and 7 learners is too few; a strange "step function" here. Maybe partly because the loud sound impact of the chorus triggers the Lombard voice reflex that makes us unconsciously speak louder. I don't know the upper limit. "Crazy-English" Li Yang may well be on the right track... At least for his revenues... ;-) http://www.china.org.cn/english/NM-e/83370.htm (I too ought to have been an English teacher in China! Hehe.)

The conclusion from these first facts is that the learners will need lots of listening practice (in active, attentive silence) before they can have a fair chance to imitate. The teacher should solo-repeat the sample sentence (seldom single words) at least 7 or 8 times at a natural rate of speech before the learners are invited or permitted to speak along. Then when they start speaking in chorus, they will have to concentrate on the rhythm of speech, which has been shown to be one of the most important, if not THE most important, factor for perception and thus also for a listener-friendly pronunciation. NB: The teacher too keeps speaking in the chorus, dominating it to "enforce" a natural and correct rate and rhythm. The learners should be encouraged to begin and end synchronously. Initially also to cheat with any difficult consonants and vowels, if necessary. (This is how toddlers do when acquiring their first language! And they never get a "foreign accent" even if they can't pronounce the indivdual sounds; this is so because they have the correct rhythm almost from the very beginning of speech acquisition. Obviously a good idea also for adult learners.) Further on, the language class will alternate fairly frequently between teacher-solo and teacher+chorus. And when some confidence is built up, also learner-soloes. (At most in chunks of 3-4 solo repetitions in these individual-practice-chorus-practice alternations, to avoid embarrassment for those who need more practice time to reach the target.)

***Second, the chorus repetitions (with alternations as above) will have to go on for a great multitude of times. In all other practices for skilled performance, such as athletics, music, circus, type-writing, car driving, surgery etc., it goes without saying that one has to practice many, many times to acquire sufficient skill. Unfortunatly, this self-evident knowledge seems to have gone "out of fashion" in language education since decades ago, and now is almost forgotten. This is where I want to change the current routines. 20-30 repetitions is certainly too few. A hundred may be more appropriate. By the thousands must be better - but we have a limit to our patience and endurance, too, haven't we! But a toddler actually practices his speech much more than thousands of times, more closely to a zillion times. It does take some 5-6 years to acquire the very basic command of one's first language, doesn't it! Fortunately, adults can do it in a single year or less - if given appropriate instructions.

The 146 (!) speech muscles are no different than any other muscles in the way they are run by the nervous system. They have to be co-trained and co-ordinated, in our first language as well as in all our subsequent languages. In learning a new language and speech, we do need lots of repetitions. This is also due to neurophysiology, and specifically to the way neurons connect with one another with _synapses_, as they are called. Buds for new synapses are formed within seconds in a learning situation, and mature synapses within 10-15 minutes. If they are reinforced by repetitious stimuli. With too few repetitions, the buds may regress, and no synapses are permanented, and the time spent on that "pseudo-practice" will have been largely wasted. It has been shown that a new skill, e.g. one sample sentence, can be automated and permanented to perfection in about 15 minutes.

It will have to be repeated the next day, too, and the next, and the next.... But then! Unforgettable! Like we can never un-learn how to drive a car, or walk, or talk, once we have learnt it. In a simile, I say that learning and memory is like walking across a lawn: Paths will form where you walk sufficiently many times. And nowhere else! And all new paths have to be connected with older paths. I.e., we can't learn much without some pre-knowledge; every peice of new knowledge is a variation of old knowledge. Such as *speaking* - we are all super-experts on at least one language, aren't we! All human speech is based on the same fundamental principles, e.g., rhythm, melody, consonants, vowels, syllables, words, sentences, ..., etc., and we all have the same anatomy and neurology, etc. So just tweak it a bit - and voila - there's your new language! ;-)

Phew.... Please pardon my verbosity... ;-) I could go on and on for weeks, presenting one neurophonetic miracle after another! Let this suffice for now lest you flame me for occupying bandwidth, and good luck, everyone!

Olle in Sweden

[carlo]

There's one thing I've noticed, after spending some time listening attentively and repeatedly to a recorded voice (that of a storyteller, or an actor) I can somehow 'recall' that voice for a few hours afterwards, and can 'hear' it say whatever I like (not just the sentences I heard spoken) in my own mind. This seem to be related in some way to the way you pick up 'accents' in your native language as an adult (I do that a lot). Repeating (chorusing) doesn't even seem necessary.

[bomaci]

There's one thing I've noticed, after spending some time listening attentively and repeatedly to a recorded voice (that of a storyteller, or an actor) I can somehow 'recall' that voice for a few hours afterwards, and can 'hear' it say whatever I like (not just the sentences I heard spoken) in my own mind. This seem to be related in some way to the way you pick up 'accents' in your native language as an adult (I do that a lot). Repeating (chorusing) doesn't even seem necessary.

I have this exact same feeling. In fact I prefer not to repeat since I feel that this interferes with my listening. However chorusing may work better for people who are not very good at imitating or picking up accents. I enjoy imitating accents so I feel that chorusing is not very necessary.