On Tue, 25 Aug 2009 23:10:43 -0400, Daniel Demski <[log in to unmask]> wrote:

>On Tue, Aug 25, 2009 at 9:28 PM, Paul Roser<[log in to unmask]> wrote:
>> I can readily produce coarticulated uvular trills with back vowels, but
>> my efforts at front vowels seem to be clearly central vowels, so I'd
>> think front vowels probably can't be coarticulated iwth a uvular trill.
>I can produce some sort of trill coarticulate with [i]. It sounds
>different than a straight uvular trill (it aquires a bit of a buzz,
>but is still more trill-y than a velar fricative) but I believe it's
>uvular. Try trilling and saying 'why' slowly- it helps to switch to

If I try to maintain true uvular trilling the /i/ retracts to central
(barred-i) and if I do as you suggest, and try to maintain the front
vowel, what I get is a rattly post-velar/pre-uvular fricative rather
than a true uvular trill - which fits in with what most languages 
experience: front vowels either retract or lower adjacent to 

>> Could you be a bit more specific on the nature of the whistle? I can
>> produce a whistled labial-palatal approximant*, or perhaps it's a fricative,
>That is an interesting sound. But what I'm doing is a labiovelar
>approximant with a ("normal", ie bilabial rounded) whistle, well
>really a labiovelar stop with a whistle of course. It behaves much
>like kĘ·. But I can try and make it palatalized instead and still get a
>whistle; that is, a palatalized velar stop followed by the sound you
>were referring to, I believe.
>> it's unclear to me which. There's no regular IPA symbol, though extIPA
>> has a small up arrow subscript that's used to indicate whistled
>> sibilants. <>
>They mean a different whistle, of course. :) I was just trying to
>produce an [s] with a bilabial rounded whistle, and realized there is
>such a thing as an alveolar rounded fricative and postalveolar rounded
>fricative. ([s_0], [S_0])
>> * I can also produce retroflex whistles and trilled whistles - the latter
>> of which I am inclined to transcribe as voiceless rounded trills [r^w_0]
>> (there's probably greater constriction/frication than in a normal trill).
>It seems I can produce a whole range of whistles from what I'm tempted
>to call whistled [s] and whistled [S] to the retroflex whistle, going
>from high pitch to low. I can also vary the pitch with my lips'
>Or did you mean the bilabial ("normal" still feels necessary to say)
>whistle but with a retroflex tongue?

I think that bilabial whistles are probably the norm for most people.
There's a wikipedia article on whistled speech with links to the research
of Julian Meyer. From what I've read, the whistles in whistled speech 
are essentially vowels produced with a whistle airstream instead of 
voicing. The vowel articulation seems to be primarily responsible for 
producing the variation in pitch (though presumably variations in lip shape
and lip tension could also alter it).

Whistled sibilants and trills are outside his research, though I think that 
trilled whistles (and perhaps lateral whistles, though I haven't tried to 
produce those) are roughly analogous to whistled vowels. 

Shona and Tswa-Ronga contrast plain /s, z/ and whistled /s*, z*/ - 
written <s, z; sv, zv> - at least one dialect of Shona produces the
whistled sibilants with rounding, but apparently most dialects don't
require the rounding to get the whistling quality.

>I can add a velar trill to a whistle too, breathing in or out.

Do you mean uvular trill? Technically one can't have a velar trill since 
there's nothing to trill, though a trill-like sound can be produced if one 
can generate enough saliva.