On 9/7/07, John Vertical <johnvertical@...> wrote:
> >I have the following:
> >
> >- _t_ = /s/
> >- _tt_ = /ht/
> >- _d_ = /t/
> >- _dd_ = /?t/
> >- _k_ = /x/
> >- _kk_ = /hk/ (thanks to some other rules)
> >- _g_ = /k/
> >- _gg_ = /?k/
> >- _p_ = /f/
> >- _pp_ = /hp/
> >- _b_ = /p/
> >- _bb_ = /?p/
> >- _h_ = /s/ before stops
> >- _s_ = /S/ before velars
> >
> >Does this seem like a plausible variant/outcome of an Icelandic-like
> >process under Alg (and Inuktitut) influence?
> >
> >Paul
>
> h ([x]?) > s doesn't seem to fit in, if you're allowing preaspirates and
> have s > S / _k. If there are [S x h] in the system, I'd expect h to to end
> up as one of them, rather than > s.
You're suggesting the following changes to the changes?
_tt_ = /ht/
_pp_ = /hp/
_kk_ = /hk/
_h_ = /S/ before stops
The only problem I see in that is a purely-orthographic one: how do I
write /S/ immediately before a stop if /s/+stop clusters may also
exist? Is re-using 'h' sensible? What if /h/+stop clusters develop
later? What about transcribing /S/+not-stop in borrowings? More
irregularity than I'm comfortable with starts seeping in.
> And what's the "other rules" part with _kk_? It looks perfectly regular...
Ah. That would be my fault. I originally had the final two rules (for
_h_ and _s_) rolled into the rest of the rules, and I guess I forgot
to un-annotate _kk_ when I broke those two rules out.
> Other than that, it's looking very cool :)
Thanks!
Paul