|From:||Roger Mills <rfmilly@...>|
|Date:||Wednesday, October 20, 2004, 17:37|
Vehke/Yann Kiraly wrote:
> On Wed, Oct 20, 2004 at 05:20:39AM -0400, Yann Kiraly wrote:
> > Thanks for the reply! Would the justification, that f->v and z replaced
> > s
> > in some words, be satisfying?
> Yes, I would say so. If you want, you could work out the exact
> conditions in which these sound changes take place. For example, /s/
> becomes [z] between vowels (intervocally), and stays [s] elsewhere.
> I'm sure it's possible to justify the complete loss of [f] by stating
> that /f/ simply shifted to [v] in *all* positions, whereas the shift
> [s] -> [z] was halted half-way.
>Consider these natlang changes from Fijian:
Fij. /v/ descends from Oceanic *p; Fij. and all the Polynesian langs.
descend from a common subordinate source (so we believe), and all PN langs.
have /f/ (or some further development like h or 0). Fij. seems to have had a
voicing rule for vl. frics. somewhere along the way...
Oceanic had *s and something resembling *ns: for *s, Fij. has /D/, for the
*ns /s/, but they are very mixed up; PN langs. have s/h/0 (mixed up) for
The development was probably: Fij/PN **s > early Fij. **/T/, then /D/ by
the voicing rule (while all PN developed **s > s/h/0)
So presumably Fijian had an early contrast **T::s, but the **s was not
affected by the voicing rule.
Not unlike the situation Yann K. is trying to develop.