Re: Help on Sound Changes
From: | Roger Mills <rfmilly@...> |
Date: | Tuesday, November 28, 2006, 18:39 |
Santiago Matías Feldman wrote:
>
> I need some advice on my sound changes from Latin into
> my Romlang "Laturslav".
>
> In the inventory of phonemes, I decided to include the
> fricatives /S/, /Z/ and the affricates /tS/ and /ts/.
> No /dZ/ there.
>
> So, I chose a usual sound change in Romanian (I think)
> and in many Slavic languages: /t/ ---> /ts/ before
> /j/, and /k/ ----> /ts/ before 'e' and 'i':
>
> informatio ---> informacya (where c is /ts/)
> caelum ----> cel (idem)
>
> But then, I discovered that if I applied those sound
> changes, Laturslav would be full of /ts/'s and I don't
> want that to happen.
How about *t > ts before j (_and_ i too?)
And the more common *k > tS before e and i etc.-- both in keeping with
Slavic changes IIRC.
OTOH the first change including __i might also produce too many /ts/s; you'd
get tsitul- (= Sp. título), but also maybe *petitione- > petsitsiV...,
depending on how you treat intervocalic voiceless stops (voiced as in Span?
retained as in Italian/Romanian IIRC) as well as lots of /ts/ in the
I-conjugation (assuming it survives). You might have to be careful in your
choice of roots-- avoiding too many that have "-ti-"; just having the change
in the ending *-tion- might not produce _enough_ /ts/s.
We need a little more info. I gather from the name that this is going to be
a Slavicized Romance lang?
> Therefore, I decided to replace those sound changes
> with these:
>
> /t/ ---> /s/
Then don't do it, keep it /ts/ :-))
> /k/ ---> /s/
K > tS > s would be a possibility
Another possibility: keep k < Lat. qu(e,i)-, but have k > tS < Lat. c(e,i)-
???; that would (more like Span.) create /ker-/ < quaer(ere), and /ki/ < qui
etc., but /tSit-/ < cit(are) (Span. citar)
Replies