The Whistled Language of Northern Turkey | Hacker News Reader

The Whistled Language of Northern Turkey

newyorker.com

29 pointsshiplet10 comments

It seems obvious, even to those unfamiliar with phonetics, how tonal languages are transformed into whistled languages by whistling the underlying tone contours.

However, even languages with no phonemic tonality are able to be whistled because we distinguish vowels almost entirely by the volume of the harmonic frequencies produced by changing the shape of the resonating chamber of the mouth. Musicians often call this "color" or "timbre". Jaw aperture roughly corresponds to the volume of the first formant produced, and how the tongue is raised or retracted within the mouth roughly corresponds to the second formant produced.

Even the surrounding consonants leave their mark on the surrounding vowels. For instance, sounds produced with the lips, (like English "b"/"p"), tend to lower the formants of the surrounding vowels. Consonants produced with the back of the tongue against the soft palate (c.f. English "k"/"g") tend to "pinch" the second and third formants of the preceding vowel closer together. Collapsing and merging consonant clusters while bringing the tonal variation into the fundamental pitch is one of the most common theories of tonogenesis, or how tonality evolves in languages.

Whistled languages based on non-tonal languages tend to whistle the patterns of the second formant, which captures subtle information about what the vowel and surrounding consonants would have been in the underlying spoken language.

Three reasons I love this:

1.) It's Aristotelian mimesis made manifest AND practical. It does diverge from his linguistic ideas about syllable construction, etc., but that's fine because it's not about him.

2.) It's about communicating, clearly, across huge distances chock-full of interference. If you watch the video at the end of the article, the two guys are probably a quarter mile apart and having a perfect conversation: in my experience, not possible with just the human voice. I have a hard enough time talking to someone across the table in a noisy restaurant - and I've (at least socially) attributed that to the fact that my voice resonates at about the same frequency as background chatter. I either have to seriously amp up my volume, or raise the pitch of my voice, neither of which are comfortable for extended periods.

3.) It's weird and it's beautiful and it's man-made. It's like realizing the power of Lisp macros when you've only ever written JavaScript, or learning FP when you've only ever worked with OOP. It breaks down ideas of what a "proper" language is, and reconstructs them in a way that conveys an entirely different ideological purpose.

I love it.

Since Turkish is a highly agglutinative language, copulas are rendered as suffixes, albeit with a few exceptions See: https://en.wikipedia.org/wiki/Turkish_copula and https://en.wikibooks.org/wiki/Turkish/Present_Tense

@shiplet I know turkish, but I was still amazed that I mostly understood what they whistled after reading your post (not sure why). I heard out some customizations to spoken turkish that I think are there to allow for a better distinction between similar sounding words. But I suspect that these are domain specific and have to be learned in order to fully speak and understand the whistled language.

Whistle-speaking in english is much harder, because you don't have enough syllables. One can't easily concatenate "meaning" to a word like in Turkish. That's why whistled-english words would sounds too similar and not harmonic enough to be distinguished.

For example try to say this loud: "I won't go to the school today." Now speak it in your mind, but with syllables. And now try to whistle that sentence (as if you were speaking, but without actually speaking). There are only single-syllable words and these sound all too similar when whistled.

See https://en.wikipedia.org/wiki/Whistled_language for more instances of this phenomenon and a good discussion. My understanding is that whistled languages often arise on top of tonal languages where tones carry a lot of information. The Turkish example is interesting since Turkish is not a tonal language but they still figured out how to do it.

The tonal aspect makes a lot of sense, and in the case of Turks applying it to a non-tonal language, I'd love to learn how they did it. Maybe it'd be possible to apply to English?

It sounds similar to Chinese.

As a native Turkish speaker: It actually sounds.. Turkish... They replaced clear words with their corresponding tones. I think with a little training, a native speaker can understand what he hear. It feels like listening a karaoke version of a well known song.

In James Gleick's The Information, the first chapter is about a West African culture, turning phonemes, stressed and unstressed, into drumbeats, for conveyance over longer distances than one could shout, and needing to compensate for the information-loss of sound into drum pulse by using longer and longer "phrases".

This is so cool. I was trying to correlate the sounds I heard to the shapes of the words, but I've got limited exposure to Turkish which made it more difficult. Fascinating to know that it [sounds] Turkish as well.

Cool, thanks!

What I meant was that this language changes intonation a lot and this is similar to Chinese.

Would it be possible to apply this to some other language?

Reply on news.ycombinator.com