> --- In qalam@yahoogroups.com, "suzmccarth" <suzmccarth@...> wrote:
> I just want to be on the record as saying that the difference
> between vowel pairs in Cree (I am not sure about Tamil) is not
> really a difference in the length of the vowel but a difference in
> vowel quality. They are typically called short and long, but I
> don't think that is strictly true phonologically.

Phonologically? or phonetically?

Phonologically (as in the *system* contrasts) it would seem that long
vs short is valid, whereas the phonetic (surface) realizations are
manifested by differences in vowel quality. (Other languages that are
steretypically cited for this topic are: Hungarian, Cantonese, Thai,
other Indic languages, etc... unlike Japanese, which truly has a timing
difference in short/long distinctions.)

> In some ways that makes it more interesting that these systems were
> thought of as having pairs of vowels - why not just 7 vowels or 11
> vowels or whatever, depending on dialect. Economy I guess. English
> too. We like our vowels to multitask.

It's because of development, analyses parallel to other langauges,
orthography, etc...
Why not just say 7 vowels? or 11 vowels?
Largely because it's not a very succinct capture of how the vocalic
system works... when length distinctions play a role in how things
change back and forth, i.e. English "short/long" vowel distinctions and
morphology-induced vocalic differences, and native orthographies
indicate length distinctions, it's usually notable that if the natives
even realize there's length, it's probably salient (but not necessarily
true the other way around).

But, at least to pull it back to "Qalam" and writing systems...

The writing system makes the distinction between short vs long in most
cases... even if *modern* Tamil (or even current Cree) has some sort of
shifted differential mechanism. What it would point to is that prior to
some re-phonologization, length distinction was salient and distinctive.

cheers,
-Patrick (the hungry)