Chinese and the Brain

[amego]

dear geraldc i agree with you

thanks for the info!

[Altair]

Did some digging into tonal vs pitch-accent, just so I know what's really going on. Seems like no-one disputes the division between contour tonal (Mandarin) and pitch-accent (Japanese), but some linguist classify pitch-accent in with tonal languages whereas some don't.

I think there is a continuum between accent and tone and quite a variety among these. There are some languages where words have discrete pitch levels, and this feature is obvious to listeners. Cantonese seems to me an example of this. For Cantonese speakers who tend to use a high level tone rather than a highfalling tone for the 阴平 tone, 芬份坟 are all pronounced alike (like English "fun") except for the pitch. 史 and 市 have the same rising contour, but the former is a mid rise (like the Mandarin second tone), while the latter is a low rise (almost like the Mandarin 3rd tone). Since such words can only be distinguished by pitch, I think speakers tend to pronounce the words in isolation with the same pitch. I would guess that this is the basis for developing perfect pitch.

In Cantonese, if you say a word with a mid level pitch (阴去声), but do not make it high enough, you do not leave enough room in your voice register to say words with low level or low falling pitch. On the other hand, if you make a mid level pitch too high, it begins to sound like a high level pitch. You need to have something like a flatted musical fifth between top and bottom and so must make keep your pitch levels within certain fairly specific boundaries.

I think that in Mandarin, the contour of a pitch is more important than in Cantonese; however, I still think that pitch is the most important feature of tone. Although the overall pitch of words seems to vary more than in Cantonese, I have never heard someone pronoune a 1st tone word with a low pitch or a 2nd tone word with a low rise. Again, I think this would help in developing perfect pitch.

There are also langagues like Hausa (spoken in northern Nigeria and neighboring countries by 20-30 million people) which have "downstep." From what I recall reading, the language has high and low syllables, but the pitch of these syllables is highly dependent on where they fall in a sentence and what the pitches of neighboring syllables are. Because of this, a low pitch near the beginning of a sentence is often higher in pitch than a high syllable occurring later in the sentence. Since this feature is probably not relevent for the citation of individual words, I would guess that speakers of this language would still have a strong sense of pitch.

I don't know enough about Shanghainese to venture a classification, but I would guess that it would fall here as a language where pitch alone can distinguish minimum syllables ad where speakers need a strong sense of pitch.

After Shanghainese, I would put a language like Japanese, but with an important difference. The Japanese pitch accent is somewhat of a virtual accent. It never affects the syllable that carries the feature, but can only affect the following syllable under certain circumstances. In a great many cases, the pitch accent has no effect at all on pronunciation. Many accented words show no trace of an accent in their citation forms: e.g., 月 (tsuki) and 手 (te). Under these circumstances, I would doubt that Japanese speakers would grow up with a strong sense of absolute pitch.

After Japanese, I would put Swedish. Swedish has very many words that have a clear pitch contour, such as kvinna ("woman") and every first-conjugation verb (e.g., tala "to speak"). (The first syllable is accented with a mid-falling contour, and the second syllable is unaccented with a high-falling contour.) Despite the high frequencey of such patterns, there are almost no words that differ only according to this feature. It has been many, many years since I knew any Sweidsh, but I can recall only one such case, which was something like anden. Said one way, I think it means "the duck." Said with another pitch pattern, it means "the spirit." One can speak excellent Swedish while totally ignoring the issue of pitch contours. Because pitch has such little importance in distinguishing meaning, I would think that Swedes would not grow up with much of an automatic pitch sense, despite speaking a language with quite marked pitch patterns.

At the end, I would put languages like English, German, French, and Spanish. In these language, pitch is only a secondary optional feature of accent and there is no concept whatsoever of pitch levels. I would doubt that speakers of such langague would tend to grow up with a strong sense of pitch.

[carlo]

It all depends what you mean by a 'strong sense of pitch'. Yes, certain syllables are regularly found at a fairly high pitch levels, in Cantonese (high level tone) and in English (stressed syllables). Yet what is the connection with perfect pitch?

You may have heard of Zhao Yuanren description of intonation in Chinese as a 橡皮带, the idea is that intonation is like an elastic rubber band, and pitch contours are like regular shapes drawn on it. So when a certain syllable is said with particular emphasis (say), the elastic band is stretched thin, i.e. the actual pitch range for that syllable is greater, and the contour shape is somehow distorted. In actual speech, the same pitch contour can be twisted and transposed in all kinds of ways by much more than a semitone. Even in sentences with all level tones, each level tone realization is different. Try saying:

西安加工波音737飞机机身.

Children with perfect pitch often dislike hearing pieces of transposed music, because they associate a given key with a melody. Obviously if you associate an absolute pitch value with a syllable of spoken Chinese, you'll never understand connected speech: this strategy simply doesn't work.

Even as children., we spend an overwhelming amount of time listening to and producing connected speech, and we babble intonation before we can even speak.

I agree with you than when pronouncing words in isolation, as adults, we might follow strategies such as those you mention above, so a high level tone in Cantonese in isolation (for the same speaker) will be realized as approximately the same constant pitch. But isn't this the same in all languages? Say 'approximately' in English like you would reading a dictionary entry (no attitudinal intonation etc), isn't the pitch contour approximately the same every time?

As I said above, I believe the association tonal language = perfect pitch (or relative pitch) is an urban myth. But I'm willing to hear contrary arguments.

[ala]

Say 'approximately' in English like you would reading a dictionary entry (no attitudinal intonation etc), isn't the pitch contour approximately the same every time?

One difference, in English it's not semantically distinctive. You would think I sound funny if I screw up my stress accents, but you wouldn't think I was saying "duck."

A music analogy can be made like this: If I first played an in-tune "A" note and then play for you a one-whole-step off-tune "A" note on my violin, you would think the second note I played was "off-tune" compared to the first, but you wouldn't know it was "G" note. In the case with say Cantonese, it's potentially possible that you know the off-tune syllable si to mean "city" as opposed to "poem", whereas in English all you know is that "approximately" was mispronounced.

[amego]

ala, that was a very cool analogy, I like it

[carlo]

Ala, I agree with your first paragraph, this is why Cantonese is tonal and English is not. I'm not disputing that.

However your analogy is misleading. What we semantically associate are not individual notes, but sequences of notes (more like motifs in a complex melody, if you like). The semantic association is invariant under transformations such as transposition (of the whole phrase) or multiplication by a constant (the 'stretching' of the pitch register). There is no association with either absolute pitch values or constant interval relationships.

So how is speaking a tonal language useful in ear training? Sorry if I insist, but it seems I'm either not explaining myself well, or missing something obvious.

[snarfer]

I saw somewhere that BLCU is experimenting with using karaoke as a teaching tool. Does any of what we read in the preceding thread suggest that learning to sing in Chinese would be of especial benefit to language learners?