On Sun, 21 Mar 2010 14:02:52 -0500, Patrick Dunn <[log in to unmask]> wrote:

>Except for /h/, which I had tentatively decided was [h] before a front
>vowel, /X/ before a back vowel.  But now I'm looking at that and
>thinking "eh, that doesn't make much sense.  Why would a back vowel
>condition a glottal fricative to become a velar fricative?"
>Am I right in thinking that one's a bit hinky?  I don't mind having
>some irregularity in the phonology, god knows, but I want it to look
>at least somewhat plausible.  

It's not the identical pattern, no.  I could see it happening -- /h/ is
always developing into a buccal fricative next to this vowel or that -- but
velarity is a significant measure less likely to spread than frontness or
rounding or lowness, yes.  

Or you might posit original [C] ~ [X] and then weaken [C] to [h].  That
would be a little strange too, having only [C] get affected, but it's doable.

>I'm also thinking of /t/ becoming [T] before a front vowel.  I think
>that's attested somewhere, isn't it?

That's more or less what happened in Spanish (only Castilian retaining it
now), and I think I've seen it resonstructed for pre-Proto-Semitic too. 
Anyway, there's a deaffrication step in there, if you do it that way.  

>Also, would it be heavily odd to have consonant mutation only in the
>pronoun system, and nowhere else?

Probably not, a pronoun system is a fine place to retain archaisms.  But it
very likely wouldn't still look ... phonological in that case

I liked the taste of the smidgen of Aerest in the street-names thread.  Is
there more of it written up anywhere?