May 30, 2011

Tonal preferences in music/speech

PLoS ONE 6(5): e20160. doi:10.1371/journal.pone.0020160

Co-Variation of Tonality in the Music and Speech of Different Cultures

Shui' er Han et al.

Whereas the use of discrete pitch intervals is characteristic of most musical traditions, the size of the intervals and the way in which they are used is culturally specific. Here we examine the hypothesis that these differences arise because of a link between the tonal characteristics of a culture's music and its speech. We tested this idea by comparing pitch intervals in the traditional music of three tone language cultures (Chinese, Thai and Vietnamese) and three non-tone language cultures (American, French and German) with pitch intervals between voiced speech segments. Changes in pitch direction occur more frequently and pitch intervals are larger in the music of tone compared to non-tone language cultures. More frequent changes in pitch direction and larger pitch intervals are also apparent in the speech of tone compared to non-tone language cultures. These observations suggest that the different tonal preferences apparent in music across cultures are closely related to the differences in the tonal characteristics of voiced speech.

Link

22 comments:

BilliardTraveler said...

Not sure about the musics of China, Thailand and Vietnam, but the music of France, Germany and America are so closely related in terms of their choice of discrete pitches and intervals as to be virtually identical. I'd have to see the details of what exactly what examined and how, but it seems pretty unlikely that you'd be able to get independant and identically distributed samples from populations that are so closely related in terms of musical evolution. What do you think?

Matt said...

seems pretty unlikely that you'd be able to get independant and identically distributed samples from populations that are so closely related in terms of musical evolution

Yes. For tonal languages, you're limited by their distribution (my understanding is that there are tonal languages in the South East Asian/China area and West Africa, but that the West African languages are tonal in quite a different manner), but it seem at first blush that it would have been more ideal to contrast against an areally separated set of non-tonal languages (e.g. Dravidian, Korean, Arabic for instance).

I expect this would be more due to limited resources amongst the authors than of any other intent though.

Charles Nydorf said...

A third class of languages using tonal patterns to signal syllable structure. Examples are Lithuanian, Norwegian and Swedish. Ancient Greek also belonged to this class. It would be interesting to see if the traditional music of peoples with this kind of language also shows characteristic features.

Anonymous said...

In contrast to other Slavic dialect groups, West South Slavic languages have largely retained the Proto-Slavic system of free and mobile tonal accent. Standard Neoshtokavian Serbo-Croatian has four accents, which involve either a rising or a falling tone on long or short vowels, with optional post-tonic lengths.

Andrew Oh-Willeke said...

Linguistically, the tone languages are much more different from each other than the non-tone languages in the sample. To the extent that tone is an areal rather than genetic feature of a language (as the evidence suggests to some extent) there may be even less independence in the samples.

I know little about the historic sources of musical culture in Asia, but the core Western musical tradition is even more unified historically than the languages of those people.

It would also be useful to know if the "American" musical tradition measured includes or excludes genres where influences other than Western classical and sacred music are known to be present (e.g. African-American traditional songs, Native American music, and Hispanic music showing indigenous American influences).

Jim said...

There isn't much connection between tones in tomes languages and pitch in music.

For one thing even in those tone languages that use level tones (pitch tone) these pitches are not absolute, at this or that frequency, but are relative to the speaker's voice, specifically to the other tones in the tone system. So they are not equivalent to musical pitch in any meaningful way.

Other tone languages use other tone systems - contour tone (where the tone of a syllable goes from on level to another, sometimes more that two levels) and register tone (where the tone involves the phonation of the vowel - various kinds of glottalization). I can't see how these have anything to do with musical pitch at all.

The article refers to "Chinese" as if it is one language. It is not, and specifically with regard to tone this matters. Mandarin is a contour tone langauge and Cantonese is a pitch tone language. If the argument is that this doesn't matter, what matters is the experience is that speakers attend to tone and this colors their music, okay, but that requires more than a simple assertion.

Moreover the article goes on to contrast tone in Chinese against "American". In fact Mandarin and American English coincidentally happen to have very similar tones - it is fairly easy to mistake either of these languages for the other at a listening distance - except that in Mandarin is lexical whereas in English it has a syntactic function, sort of an oral punctuation. (Apparently they don't teach this feature of Englsih in India, and Indian languages seem not to have it either, which makes it so frustration to try to communicate with an Indian call center.)

I wonder why the authors did not choose a set of tonal and non-tonal languages from a single cultural areas to compare. Southern Mexico/Guatemala comes to mind. It could not have been so hard to recruit test subjects, or to find samples of music.

Anonymous said...

(Apparently they don't teach this feature of Englsih in India, and Indian languages seem not to have it either, which makes it so frustration to try to communicate with an Indian call center.)

Indian languages are the only perfect phonetic languages in the world. you seem to have some bias.

Onur Dincer said...

Indian languages are the only perfect phonetic languages in the world. you seem to have some bias.

There is an unusually high degree of correspondence between the phonemes of English and South Asian languages (Indo-Aryan and Dravidian), especially in vowels (maybe even more than between English and East European/South European/West Asian/Central Asian languages). That is probably why South Asians are fairly good at speaking English with a perfect English pronunciation.

Jim said...

"Indian languages are the only perfect phonetic languages in the world. you seem to have some bias."

The only phonetic perfection is pure silence.

The Russians think Russian is the most complex, the most expressive, the most evocative....yadda yadda. The Arabs think God speaks Arabic. The Chinese just think Chinese is the only really natural language, that little kids will naturally understand it no matter what, and all the foreigners are just being unnecessarily stupid. And so on.

Since Indian languages vary phonetically, which one is the phonetically perfect one? Are there bleachers to watch the riot from?

Andrew Oh-Willeke said...

"Mandarin and American English coincidentally happen to have very similar tones - it is fairly easy to mistake either of these languages for the other at a listening distance - except that in Mandarin is lexical whereas in English it has a syntactic function, sort of an oral punctuation. (Apparently they don't teach this feature of Englsih in India, and Indian languages seem not to have it either, which makes it so frustration to try to communicate with an Indian call center.)"

Interesting observation. There are indeed a lot of phonetic and tonal aspects of American English that native speakers know intuitively but that even sophisticated and educated speakers of American English are not consciously aware of because they are not noted in the written language or any widely disseminated formal treatments of the language.

Anonymous said...

"Since Indian languages vary phonetically, which one is the phonetically perfect one? Are there bleachers to watch the riot from?"

This is highly uneducated statement. does it deserve look from bleachers? Naa

go to nearest university on linguistics and find out.

When International Phonetic Association came up with standardized representation of the sounds of spoken languages it is nothing but a copy of Indian vowels and consonants.
100% for vowels and 50-75% for consonants.

http://en.wikipedia.org/wiki/International_Phonetic_Alphabet

Jim said...

Okay, a Hindutva chauvinist. I thought so. Not worth talking to.

That last bit about the IPA
"When International Phonetic Association came up with standardized representation of the sounds of spoken languages it is nothing but a copy of Indian vowels and consonants.
100% for vowels and 50-75% for consonants."
is too ridiculous to answer. You clearly have no iea what you are talking about.

Dienekes,
""Interesting observation."
Just my subjective impression.

"There are indeed a lot of phonetic and tonal aspects of American English that native speakers know intuitively but that even sophisticated and educated speakers of American English are not consciously aware of because they are not noted in the written language or any widely disseminated formal treatments of the language."

That's common with native speakers of any language. It's generally worse if those speakers have any level of education, because then they think they know more than they do. In English that means they tend to deny that the language has and uses glottal stop, they misunderstand the phonolgy of aspirated stops and it gets even worse with regard to grammatical rules. There you get all kinds of misinformed just-so crap they picked up in the 8th grade.

Anonymous said...

Jim,

You got to be Chinese obviously.

I will always be anti chauvinistic whether it is Hindu,Chinese,Greek,English, Black etc.

again.

go to nearest university on linguistics and phonetics to find out what is phonetics.

Jim said...

"You got to be Chinese obviously."

Guess again.

You have a point about the *form* of the IPA being taken form Indian modesl. That is quite treu, or if it was not copying, the fact remains that Panini came up with that matrix first. (Which is not the same thing as what you asserted above.)

But that is the Indian genius anyway - systematization. It carries over into describing and formalizing grammar, obviously, but also in philsophy and of course mathematics.

But back to phonetics - no language is more perfect than another on phonetics. Any phonetic system with flaws gets corrected immediately by the speakers into a balanced and sufficient system. It's analogous to selective pressure. And it doesn't matter whether a system has only 12 consonants and five vowels and super long morphemes (Hawaiian, Pawnee and others) or 55 consonants and two vowels (North caucasian langugaes) or something in the middle and tones to distinguish morphemes [they all generally end up with the same number of contrasts, and that is what matters in a natural language.

And a correlary to this is that two languages are equally different form each other. So where a speaker of say Tamil is going to have difficulty learning the prosodic rules of English, an English-speaker is likewise going to have a problem with listening comprehension in Tamil if he is looking for prosodic rules to guide him.

Onur Dincer said...

they all generally end up with the same number of contrasts, and that is what matters in a natural language.

They don't generally end up with the same number of phonetic constrasts; the number of phonetic contrasts can and often do greatly vary across languages.

Jim said...

Onur, strictly speaking you are right. I over-spoke - they all end up with the same number of cntrasts, but not all those contrasts will be phonetic. Extra-long morphemes are the method for compensating for lack of phonetic contrast. And then there is one more I forgot to mention - contextual contrast, for example in Mandarin where orphemes usually occur in compounds because there is so much homophony. way back when Zhao Yuanren did a little experiment by translating the Lord's Prayer or something into Cantonese, Mandarin and Shanghai - high to low phonetic contrast - and sure enough, the Cantonese text was shorter than the Mandarin, which was shorter than the Shanghai.

But what I meant was that where a language has a small number of consonant phonemes it may compensate with a lot of vowels, and where it has few consonants and vowels, it may compensate with tones, and those are all phonetic contrasts.

Onur Dincer said...

Jim, when you wrote above "they all generally end up with the same number of contrasts" it is obvious from the context of the statement that you only meant phonetic contrasts. Let's examine its context:

"But back to phonetics - no language is more perfect than another on phonetics. Any phonetic system with flaws gets corrected immediately by the speakers into a balanced and sufficient system. It's analogous to selective pressure. And it doesn't matter whether a system has only 12 consonants and five vowels and super long morphemes (Hawaiian, Pawnee and others) or 55 consonants and two vowels (North caucasian langugaes) or something in the middle and tones to distinguish morphemes [they all generally end up with the same number of contrasts, and that is what matters in a natural language."

You constantly mention phonetics and your examples are all phonetic examples, so it is obvious that you only meant phonetic contrasts there. So please do not twist your words and be honest to yourself and others.

Jim said...

"You constantly mention phonetics and your examples are all phonetic examples,..."

Onur, "all" my examples are phonetic? Why are you omitting these:
"I over-spoke - they all end up with the same number of cntrasts, but not all those contrasts will be phonetic. Extra-long morphemes are the method for compensating for lack of phonetic contrast. And then there is one more I forgot to mention - contextual contrast, for example in Mandarin "

And when I say "Onur, strictly speaking you are right." That is what I meant, that your observation that languages actually do vary in their amount of phonetic contrast.

So this from you "So please do not twist your words and be honest to yourself and others." is misdirected and insulting.

And a side note on courtesy - I do not know you cultural background, and perhaps your culture does not put much value on honesty, but mine does and your comment was insulting. That's just a piece of infomration you can heed or ignore as you wish.

Onur Dincer said...

Onur, "all" my examples are phonetic? Why are you omitting these:

All your examples in your comment that I quoted were phonetic and its context was about phonetics, so I objected to your statement "they all generally end up with the same number of contrasts, and that is what matters in a natural language". Then instead of openly accepting your error, you chose to downplay it. If you had said "I mis-spoke" instead of "I over-spoke", I wouldn't question your honesty. I apologize if my words sounded insulting to you.

Andrew Oh-Willeke said...

"That is probably why South Asians are fairly good at speaking English with a perfect English pronunciation."

I'd put my money instead on the fact that India spent a bit more than 400 years as a British colony, adopted it as a regional lingua franca in a subcontinent as linguistically diverse as Europe, and learn it from a young age, and hence have an opportunity to here native speakers of BBC English and American English at the time when they are still in the childhood formative language learning phase.

Also, of course, very few South Asians speak "English with a perfect English pronunciation." In fact, this is exceedingly rare. Most South Asians born in South Asia, not surprisingly, have a very distinctive South Asian accent that people who aren't used to it have a difficult time understanding. There are exceptions, and in a country of well over a billion people, those exceptions add up in raw numbers, but they are hardly the norm.

Jim said...

I see what you are saying. "Over-spoke" is vague - I meant to say that I had generalized beyond the evidence.

On the other - no harm, no foul. I didn't think you intended any offense.

Onur Dincer said...

I'd put my money instead on the fact that India spent a bit more than 400 years as a British colony, adopted it as a regional lingua franca in a subcontinent as linguistically diverse as Europe, and learn it from a young age, and hence have an opportunity to here native speakers of BBC English and American English at the time when they are still in the childhood formative language learning phase.

Also, of course, very few South Asians speak "English with a perfect English pronunciation." In fact, this is exceedingly rare. Most South Asians born in South Asia, not surprisingly, have a very distinctive South Asian accent that people who aren't used to it have a difficult time understanding. There are exceptions, and in a country of well over a billion people, those exceptions add up in raw numbers, but they are hardly the norm.


Thanks for that reply. I am well aware of that colonial history and that accent issue. I wrote more about the pronunciation of vowels than consonants, and in consonants South Asian languages certainly have many distinctive features. Sorry for my unclarity.

On the other - no harm, no foul. I didn't think you intended any offense.

Honesty is a very ambiguous concept and has many different usages, irrespective of the culture; also, the level of honesty in one specific situation does not entail the same or similar level of honesty in other situations. Moreover, we often make our mistakes in a very subconscious level. So I am again sorry if my words sounded offensive to you.