Re: Distribution of Front Rounded Vowels

Ben Poplawski wrote:

> While applying the sound changes, I noticed that there is an awful lot of
> the vowel [y]. The sound already appears in the parent language and is
> multiplied by the vowels [i] and [u] reciprocally affecting each other to
> [y].
>
> > And now my question: what is the normal distribution of front rounded
> vowels? Something tells me the overabundance of them in the daughter lang
> is
> unnatural.
>
In the sense that front+rounded is considered "highly marked", yes. I don't
know about the distribution in natlangs.  Perhaps your problem is the
reciprocality; at least in German, umlaut only works in one direction.

I've been finding the same process as I work on Gwr, except the outcome in
both cases is a compromise [1] (high central unround). I was also bothered
by the prevalence of [1]-- Your question suggests a way around the problem:

Perhaps u...i > y, fine (and I assume [ui] in hiatus does too)...then
perhaps i...u (and [iu]) might > [1] (barred i) or [M] (unround u) or even
[i] (deletion of second vowel without affect).

Another way:  early, stage 1: u...i, ui  > y, then this [y] and original *y
> something else, perhaps [i] (or [1]?  I happen to like [1] :-)))  ).
later, stage 2:  i...u, iu > y

Yet another way:  u...i > y, fine; i...u > y too, but the preceding
consonant is palatalized (in particular, td/kg > tS,dZ perhaps.  Perhaps
too, palatalization could override umlauting. Your rule order would be:
1. u...i umlaut
2. palatalization
3. second-vowel deletion-- this would remove the environments for i...u
umlaut.

Not sure I've worked this out; it's a complicated problem.....