On Fri, 29 May 2009 20:31:51 -0400, Herman Miller wrote:

> Well, back to the original comment, clicks aren't that unusual in this 
> kind of marginally-linguistic usage (English even has some). But not 
> many languages use clicks as actual phonemes (contrasting in the same 
> position with other linguistic sounds like stops or fricatives). That's 
> what I was trying to get at by saying "phonemic whistles". E.g. a 
> language might have "<whistle>ika" contrasting with "sika", or 
> "t<whistle>pa" contrasting with "tupa". I think whistles are easier as 
> consonants, but I wouldn't rule out some language having one as a vowel.

I find whistles to work quite well as syllable nuclei.

> At one point the Tirelat word for "whistle" was ['H_0yH_0y], which 
> wouldn't take much of a change to make actual whistles from those 
> voiceless labial-palatal approximants.

Why not?  I always perceived labial whistles to be somewhat like
front rounded vowels (or semivowels), and they could evolve from

... brought to you by the Weeping Elf