THEORY: voiceless laterals ( was: h huffnpuffery)

At 7:14 pm +0100 7/2/00, BP Jonsson wrote:
[...]
>
>Those Tibetan "lh"s I've heard didn't differ appreciably from the Icelandic
>sound, and it is a fact that when I used voiceless lateral approximants in
>Icelandic words like "hleypa" Icelanders told me I was pronouncing the
>sound too weak.  According to them it "came out as /h/".  I think that sums
>up pretty well why phonemic voiceless laterals tend to be fricative.  BTW
>the sound written "hr" in Tibetan is actually a voiceless retroflex
>sibilant, so /hl/ being fricative makes sense systemically.
Thanks  :)

I thought that was probably the case.  So 'twould seem Tibetan 'lh',
Icelandic & Nguni 'hl' and Welsh 'll' are the same animal: a voiceless
lateral fricative.

I have a theory that while unvoiced lateral approximants can happily exist
as allophones of the voiced approximant, as in English, if they find
themselves achieving phonemic status they then become unstable and either
(a) simply get voiced and merge with [l] as, e.g. Old English 'hla:f' -->
'loaf', or (b) become fully fledged voiceless fricatives as in the
languages mentioned above.

Ray.

=========================================
A mind which thinks at its own expense
will always interfere with language.
                   [J.G. Hamann 1760]
=========================================