Hallo conlangers!

On Tuesday 06 December 2011 11:57:42 Paul Bennett wrote:

> I've been thinking about vowels in Proto-Terzemian.
> 
> It should have the vowels /i/, /u/, /a/, /Q/, (i.e. very loosely [+/-
> high] and [+/- round]) plus length, and needs to get there from
> Proto-Indo-European.
> 
> I'm guessing the obvious route for /Q/ ~ /Q:/ is just to lower all *o(:),
> but what about *e(:)? Merge them all into /a/ ~ /a:/?

Several IE languages have merged *a(:) and *o(:), so a good
solution would be merging them into /Q(:)/, while *e(:)
becomes /a(:)/, except where raised to /i(:)/, see below.

> Split them somehow
> between /i/ ~ /i:/ and /a/ ~ /a:/? If I split them, should I apply the
> same logic to splitting *o(:) into /u/ ~ /u:/ and /Q/ ~ /Q:/? What are
> sensible criteria for splitting?

A sensible criterion would be that *e(:) is raised before
voiceless consonants.  You could raise *o(:) under the same
condition if you like to.

>        Can I make do with the conceit that
> post-PIE vowels influenced by laryngeals went (down/up) while "native"
> vowels went (up/down)?

I'd rather not.  No IE language, AFAIK, treats these differently.

>        Should it (being related to the *Indo-Slavic
> metabranch) do the same as Proto-Slavic (whatever that was) to get to the
> same or a similar place?

Yes.  Slavic merges *a with *o and *a: with *o:, as I have
suggested above.  I think there are also some raisings in
Slavic, but I don't recall now the conditions.
 
> Your considered thoughts, if you please. Sorry for the mess: I'm kinda
> making it all up as I go along.

Never mind.  Your questions are welcome.

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Êm, a Êm atha cvanthal a cvanth atha Êmel." - SiM 1:1