[YG Conlang Archives] > [romconlang group] > messages [Date Index] [Thread Index] >


[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: De-palatisation in Northern Romance (WAS: Northern Romance chronology and



--- In romconlang@yahoogroups.com, "Benct Philip Jonsson" <bpj@...> 
wrote:

[...]

>
> >| CL          -- WRom / GRom -- NRom
> >|
> >| kj ki ke    -- ts tsi tse  -- T Ti Te ( #s #si #se )? ^2
> 
> Check, but I can't see why initial instances of [ts'] should
> get a different shave.  In fact we *want* initial _th/dh > d_
> in particular, no?

For sure. /T/ in all positions then.

[...]

> >| #dj         -- #dZ         -- #D ^3
> >| dj          -- 0           -- 0
> 
> I'm much more inclined to #dj > j (possibly #dj > d > t) and
> definitely -dj- > dd.

At the time the WRom palatisation is being undone, there is no /d/ in 
our target phonology, only /D/. This /D/ does become /d/ shortly 
afterwards though.  However, I think I prefer #/j/.  So toss a coin, 
do we want:

#dj > #Dj > #D > #d (we need a very fast change here! - feasible?)
#dj > #j

dj > (Dj >) dj > dd for certain.

[...]

> >| #ga         -- #dZa        -- #Da ^3
> 
> As I said before ka > tSa/ga > dZa is of limited validity even
> in Gaul.  I'd rather think it simply won't happen in Germano-
Romance.

OK, scratched.

> >| sj          -- z           -- s ^4
> 
> No, definitely sj > S > x > h. Old French has Vsj > Vis in all
> instances, which probably developed from [S] or rather [s\].
> Remember there was no /S/ in OF, only /tS/ < k / _a!

This is problematic. I want /S/ > /x/... BUT: There is no /S/ in OF, 
and there is no /S/ in OHG either. So we have no /S/ anywhere. I 
don't see how if we have no /S/ where we start or finish, we can have 
it in the middle!?

The Gmc had [s_a], and I think this is the closest phoneme to /sj/. 
Rom /sj/ also seems to have shifted to /z/ ([z_a]?) in many 
instances, which again would be [s_a] in Gmc.

That must mean we are stuck with sj > z > s ??

[...]

> >| rj          -- r           -- r
> 
> - Old French had rj > ir!
> - W.Gmc. simply *preserved* /rj/ (e.g. *nazjan > OE _nerian_)

No change for rj then.

[...]

Pete