Tones (was: Mandarin pronouns (ta1) [Was: a question about names])
|From:||Roger Mills <rfmilly@...>|
|Date:||Friday, October 1, 2004, 5:38|
David Peterson wrote:
>The voicing of different coda consonants affects the length of the vowel,and certainly certain types of codas affect the quality of the vowel moreso
than the onset usually does...>
Hmm, I hadn't taken the voiced:length correlation into account, and in fact
have been of two minds about what to do with voiced finals-- (1) devoice
them (with preceding V change e.g. -áb > -op, later -o?; depending on what
the preceding syllable was, that would produce, probably, high-falling tone)
or (2) lenit them, as they do in medial position:
*kábat > kaw@t > kawt > kaw? HF
*kabát > k@wat > kwat > kwa? (probably MID)
OTOH *katáb > **k@taw > taw MID? HI?
while *kátab simply > **kat(0) > ka? HI? or maybe this would be lowered to
MID by the old vd. final, to contrast with ka? H < ká(ptk)V(ptk)
Here's one I like: *(tk)Vdáy >**(tk)0y& > [tS]& (not sure of tone, either H
or M) but note: if *tíday, > **tiy& also > tS& HF
or: *táday > tay& > t&: probably HF
>(note: uvulars, IMO, affect the following vowelmore than the previous vowel,
Well, *q lowers the stressed vowel, whether preceding or following, and also
lowers the tone:
*píqV > peq > pe? Mid vs. pí(tk)V > pit > pi? Hi
*pVqí > .... qe Mid vs. *pV(tk)í > (tk)i Hi
Initial q causes a following p,k,s > f x S in the case of *qvCV (where v=
unstressed, V= stressed); and in *qv(bdg)V > (ptk)V LO (this change
obviously ordered before lenition).
Shout if anything looks unreasonable!!
>(1) Vowels followed by voiced codas tend to be longer than vowels followedby voiceless codas.
(2) Volume and intonation go down as an utterance progresses.
(3) If a vowel's going to be longer, then there's a better chance of its
being lower than a short vowel.
(4) Therefore, low tones could arise from words with voiced codas, and high
tones could arise from words with voiceless codas.>
Yes to all of the above; But also, voiced initials produce low tone (the C's
remain voiced in BDG, in many natlangs they devoice). I'm not sure what
initial/medial/final nasals do, except they likely can't be high tones.
Fortunately the vocab at the moment is still small, so errors or changes of
mind can be accommodated.
The real challenge is deriving 8 vowels from original 3 :-))))