On 11/07/08 04:25:12, John Vertical wrote:
> On Thu, 10 Jul 2008 17:45:46 +1000, Tristan McLeay wrote:
> >> countour tone
> >
> >Well, level/contour tones can develop into register tones (e.g. a 
> low
> >tone may become creaky voice).
> I've suspected as much - got any examples?

I will try and find them later today; I'm already late for work. I 
think Danish is one though? (The stød --- I don't really know what it 
is, but I think it's glottalisation --- corresponds largely though not 
entirely, I think, to the low tone in the other continental North 
Germanic languages.)
> >Stress, however, is I think a poor analogy for
> >tones; a better one would be length. Length and tone can both
> indicate
> >stress, but they can both be phonemic features on their own.
> Interesting angle. I was just going with "prosodic suprasegmental". I
> suppose it would be possible to consider length such too in some
> circumstances ("suprasegmental length" gets 16800 Ghits) but that's
> still
> not too commonly seen.

Why? If length only affects one of consonants and vowels, then the 
other tends to try to make up for short syllables by being longer (or 
for long syllables by being shorter); at least, this is true of North 
Germanic languages, Italian, and intuitively (Australian) English.

Other features frequently described as segmental are not always 
segmental either; in Turkic languages the vowel harmonic features last 
for the entire word (e.g. in a rounded vowel, the consonants are 
rounded; in a back word, /k g l/ are retracted/dark compared to a front 
word). Considering the boundary between segmental and suprasegmental 
aspects to be fixed or even just clear is I think too simplistic.

> Also to consider: tone is typically orthogonal to vowel quality,
> stress and
> length less so.

Ah, yes, but that's just because stress in the languages in question 
comes with extra vowel length. You find vowel quality differences in 
languages like English where stressed syllables are longer than the 
equivalent unstressed one, but not in languages like Japanese which 
have a pitch-accent system.

You would not however expect tone to cause changes in vowel quality, 
but whatever it does change/affect/get affected by/etc you would expect 
it to happen regardless of whether it's independent or just an aspect 
of stress.

> Apparently, I simply managed to mkae teh smae tpyo twice. :)