Whistling

[sucaeyl]

I found myself whistling merrily this morning, and realized that there was no language that I knew of that had labial whistling as a phoneme. This puzzled me, seeing how easy (for me anyway) it is to whistle, even within a word-like context. I would've thought that these sounds would show up somewhere, but after a little research, I found only whistled sibilants and whistled languages (that is, entirely whistled). What's going on?

For practice, try the following hypothetical words, with ‹ʍ› representing the whistle:

/kʍː˧/
/a.tʍː˦˨.ma/
/ɛ.ʃʍː˥.ksʍ˦˨n.ta/
/aːʍ˧.ka/

Easy, right?

[Aurora Rossa]

I suppose part of the difficulty is explaining how such a sound would originate diachronically from other more typical speech sounds. Perhaps it could develop from rounded vowels caught in environments where they devoice, providing the necessary lip rounding or something.

[Nooj]

Maybe it doesn't have to originate from other sounds, speakers could conlang it. One of the hypotheses for Damin, which is the only language outside Africa to use click consonants, was that it was a conlang. If click consonants can be invented, why not whistle consonants?

[Astraios]

Easy, right?

No, because a lot of people just can't whistle.

[Gulliver]

I found myself whistling merrily this morning, and realized that there was no language that I knew of that had labial whistling as a phoneme. This puzzled me, seeing how easy (for me anyway) it is to whistle, even within a word-like context. I would've thought that these sounds would show up somewhere, but after a little research, I found only whistled sibilants and whistled languages (that is, entirely whistled). What's going on?

For practice, try the following hypothetical words, with ‹ʍ› representing the whistle:

/kʍː˧/
/a.tʍː˦˨.ma/
/ɛ.ʃʍː˥.ksʍ˦˨n.ta/
/aːʍ˧.ka/

Easy, right?

Actually, there are languages which only use whistled phonemes. I remember seeing a bit of one on an advert for local news when I was in Tenerife on holiday years ago.

[sucaeyl]

I suppose part of the difficulty is explaining how such a sound would originate diachronically from other more typical speech sounds. Perhaps it could develop from rounded vowels caught in environments where they devoice, providing the necessary lip rounding or something.

Or syllabic /ɸ/...

Easy, right?

No, because a lot of people just can't whistle.

Nor can a lot of people make /ʢ/. It still's phonemic though. Also, speakers are lazy and it is likely, if such a language with whistled phonemes existed, that they wouldn't always be crisp, in-tune whistles worthy of song. A vaguely resonant bilabial sputter would suffice in most speech situations.

Actually, there are languages which only use whistled phonemes. I remember seeing a bit of one on an advert for local news when I was in Tenerife on holiday years ago.

I've heard about those. It's interesting how many spoken languages can be whistled entirely.

[Miekko]

I've heard about those. It's interesting how many spoken languages can be whistled entirely.

I'd guess that's the result of some tonal languages having pretty a larger amount of information in the actual tonal contours of the speech than say English has in the CV-part of any syllable. Considering that afaict, whistled languages are spoken as a secondary language derived from some spoken language simply by dropping everything but the tones. Whistling languages are probably not even spoken by everyone in the community, simply because some might not learn how to whistle.

[Xephyr]

Wouldn't it be likely for a labial whistle phoneme to pattern with fricatives or sonorants, rather than as syllable nuclei?

[sirdanilot]

Shona does have whistled sibilants, though I suppose you already knew. A pure 'whistle' would technically be [ɸ*] (I use the wildcard * for 'whistle' since I don't know of a proper IPA symbol for it). I don't know if everyone can do it but a voiced whistle can also be produced and has quite a marked sound: [β*]. I find these easiest to pronounce at the beginning of a word, and with a glottal stop in front of it: [ʔɸ*aːt]; in contrast, I hardly get any whistle at all at the end of a word.
They could also be syllabic, allowing all sorts of tonal contours: [tɸ*t].

Now, why does this perfectly possible phoneme not exist? I think the answer falls in the same category as 'why does English not have clicks? In fact it does (doesn't everyone do a dental click when they're annoyed?) but not as a productive phoneme. In English, one whistles when an attractive lady passes by, or when someone shows his new car, etc. The diachronics for a whistle phoneme were perhaps too complicated to have happened. But if you want this for a conlang, you could go like:

tupat > tuɸat > tɸat > tɸta > tɸ*ta

There should be some labial environment for the ɸ to turn into a whistle, according to my intuition.

[finlay]

Easy, right?

No, because a lot of people just can't whistle.

Also, whistles use a different mechanism to make sound – it's to do with some kind of vibration around the lips or whichever articulator is making the whistling noise, rather than vibration around the glottis as in a normal voiced pulmonic sound. You can prove this empirically by making an ingressive whistling noise: it's easy to get it to the same volume and pitch as your egressive whistle. Right? Now try saying a pulmonic egressive voiced vowel (for the sake of example, cardinal 4 will do, but it doesn't really matter). OK? Now try cardinal 4 ingressively.

The result of the effectively different airstream/voicing mechanism is that you have to turn off the pulmonic voice in order to make a whistle, and thus it's not actually that easy, even besides the people that just can't make the noise in the first place.

[sirdanilot]

^^ That is a nonsense argument. Many non-european languages have non-pulmonic sounds: khoisan clicks, salish and caucasian ejectives, sindhi implosives... Clicks are the best example here because they are the most versatile. Despite being non-pulmonic, it's perfectly possible to make voiced vs. unvoiced clicks (it is in fact a phonemic feature in many languages). The only reason that, for example, voiceless implosives are rare (but not entirely non-existant I believe) is that they are hard to articulate, not because it's theoretically impossible. Only voiced ejectives are theoretically impossible, which is why we say that ejectives are not specified for voice (unvoiced rather than voiceless).

I don't know if everyone can do it, as I said, but I can easily make a voiced whistle myself. It sounds a bit weird (the whistle in combination with the pulse from the glottis gives some funky resonance/interference/whatever) but it is theoretically possible. I have observed my father making that sound when mimicking a steam train, for example ([kɑdɪɡɑˈdɪŋ ʔβ*ː ʔβ*ː] where *is a whistle) , so it even occurs in some languages (though certainly not as a productive phoneme).

Also, even if some people can't whistle, it doesn't mean that the phoneme cannot exist. As long as the majority can pronounce a whistle it's fine. Many Dutch people can't properly pronounce [r], for example. They usually use [ʀ] or an approximant or a flap or somethinɡ. Usage of [ʀ] is, in my dialect, pejoratively called 'brouwen', but in other regions this usage is fully accepted.

[Rui]

Despite being non-pulmonic, it's perfectly possible to make voiced vs. unvoiced clicks (it is in fact a phonemic feature in many languages).

Because clicks are coarticulated with a pulmonic consonant (/k g/ in the case of voice contrasts). By definition, a purely non-pulmonic consonant can't have a voicing distinction because the vocal folds play no role in their articulation. In the case of clicks, anyway.

[sirdanilot]

^^ Matter of definition. Sure, the click itself is not voiced, but the entire thing (which is parsed as one phoneme) is certainly voiced because of the co-articulation.

The voiced whistle I talk about is also just that, a whistle co-articulated with a [β].

[Miekko]

Also, even if some people can't whistle, it doesn't mean that the phoneme cannot exist. As long as the majority can pronounce a whistle it's fine. Many Dutch people can't properly pronounce [r], for example. They usually use [ʀ] or an approximant or a flap or somethinɡ. Usage of [ʀ] is, in my dialect, pejoratively called 'brouwen', but in other regions this usage is fully accepted.

By what authority is [r] the proper way?

[----]

I read about whistled speech in Pirahã and apparently it's the sole type of speech they use when hunting, which makes sense because whistling would be a lot less likely to startle an animal such as a bird than human speech. It's entirely possible (as a couple people earlier in the thread mentioned) that a people would conlang a feature like this in.

[Aurora Rossa]

I read about whistled speech in Pirahã and apparently it's the sole type of speech they use when hunting, which makes sense because whistling would be a lot less likely to startle an animal such as a bird than human speech. It's entirely possible (as a couple people earlier in the thread mentioned) that a people would conlang a feature like this in.

I have heard that this works because it has a very small phoneme inventory but a rather rich tonal system, which means the phonemes carry much less information than European languages. Something else to consider, I suppose, when designing whistled languages.

[richard1631978]

I did once hear about one whilstled language that came about because of a local custom of cutting out tounge as a punishment.

[sucaeyl]

I did once hear about one whilstled language that came about because of a local custom of cutting out tounge as a punishment.

I'm curious where you heard this, seeing as you need a tongue to whistle.

[Aurora Rossa]

I'm curious where you heard this, seeing as you need a tongue to whistle.

I mostly whistle with my lips, although I may be doing it wrong.

[Zwap]

Start whistling, then move your tongue towards the back of your mouth. The pitch will get lower and lower, and eventually you'll stop whistling altogether. My conclusion from this experiment is that yes, you do need your tongue to whistle.

[sucaeyl]

I'm curious where you heard this, seeing as you need a tongue to whistle.

I mostly whistle with my lips, although I may be doing it wrong.

Your lips are equally important; however, in order to produce any tone, the tongue must be positioned at a certain minimum height. Try whistling whilst your tongue is in the position of /ɑ/. It just sounds like a bilabial fricative, right? No whistle?

EDIT: Just saw Zwap's post, refuted Eddy's unsubstantiated claim first, even included my same little experiment!

[sirdanilot]

I can whistle with my tongue in any position, except in retroflex position because it blocks the airflow too much. I am not sure, of course, but I wouldn't be surprised if I could still whistle after my tongue were removed.

By what authority is [r] the proper way?

Read my post again. In my dialect (certainly not all dialects of Dutch !), it is socially less acceptable to use a uvular trill. There are pejorative terms for it (such as brouwen ) and I have even heard of some mothers sending their children to a speech therapist (not for [ʀ] but for a child with some sort of weird epiɡlottal flap for /r/. if that's even at all possible for a normal person to do...).

Remember that language is determined by the majority. If the majority thinks you suck when you say [ʀ] instead of [r], it means [r] is the norm. In most other dialects of Dutch, pronoucniation of /r/ varies anyway. Some have an approximant in coda position, some drop the r in consonant clusters, some have a velar or uvular fricative for /r/, some have a uvular approximant...

[Qwynegold]

I found myself whistling merrily this morning, and realized that there was no language that I knew of that had labial whistling as a phoneme. This puzzled me, seeing how easy (for me anyway) it is to whistle, even within a word-like context. I would've thought that these sounds would show up somewhere, but after a little research, I found only whistled sibilants and whistled languages (that is, entirely whistled). What's going on?

For practice, try the following hypothetical words, with ‹ʍ› representing the whistle:

/kʍː˧/
/a.tʍː˦˨.ma/
/ɛ.ʃʍː˥.ksʍ˦˨n.ta/
/aːʍ˧.ka/

Easy, right?

Heh, I was just making [tʍui] sounds the other day, thinking that that would be a nice phone of some sort.