Henrik Theiling wrote:
> Hi!
> 
> Is the sound shift /gl/ > /gr/ and/or /kl/ > /kr/ attested in Romance?
> 
> Or at least, would you think it's feasible like the /bl/ > /br/ in
> pt. 'branco' < BLANCU(M)?

Not that I know of, but IMO it is at least as feasible as Port.
'branco'. I think it's just an historical accident that Port. (AFAIK)
doesn't have any examples of gl --> gr - there aren't many Latin words
beginning gl- and Port. & Spanish didn't adopt VL *glacia for 'ice'.

As for /kl/ of course on the Iberian peninsular this developed to /L/

> I am thinking about 'grorie' < GLORIA(M) in a new toy romlang I
> started yesterday.

Wouldn't the tendency to dissimilate r....r sequences counteract that? cf.
Latin 'arbore(m) "tree" --> Span. árbol, Italian albero

Of course one could have 'grolie'    :)

> Also, what about /Nn/ > /Ngr/ like /mn/ > /mbr/ (sp. 'nombre' < NOM(I)NE(M))?

Sort of like Spanish _sangre_ <-- 'sang(ui)ne(m) ?


> Tere, I'm thinking of 'ringre' < RE:GNU(M).

Um - but the Spanish shift of /mn/ --> /mbr/ and /Ngn/ --> /Ngr/ is post 
VL and ....
=========================================
Alex Fink wrote:
 > On Fri, 23 Feb 2007 15:31:19 +0100, Benct Philip Jonsson
 > <[log in to unmask]> wrote:
[snip]
 >>The only snag is that [Nn] for _gn_ is AFAIK unattested in
 >>Romance, which IIANM has /J/ or /nn/ > /n/, in all the
 >>relevand words, e.g. Old French _renne_.

Yes, in western VL a velar before a dental shifted to a palatal, so in 
north Italy, Gaul & the Iberian peninsular we find [kt] --> [ct] --> 
[Ct], at which point [C] gains voicing from the preceding vowel with 
which it then forms a diphthong.

So /Nn/ --> /Jn/ which then generally became through assimilation /JJ/, 
in Gaul and Italy
agnellu(m) "lamb" --> Fr. agneau /aJo/, It. agnello /aJJello/
regnu(m) ---> Old French regne, It. regno (Mod. Fr. has _royaume_ from a 
different origin.

But Span. & Port. _reino_ presumably is from /Jn/ --> /jn/

 >But wait, Rumanian
 >>has LIGNU > _lemn_, and I guess all of /J/, /n(n)/, /mn/ may
 >>be derivable from [Nn] if one wants to.

Yep - just as _octo_ --> opt and nocte --> noapte

Velar + dental --> labial + dental in Eastern Romance

 >
 > I thought that Latin <GN> was generally understood to be [Nn], and 
Latin's

Precisely!! That's why Henrik was asking if it was realistic for this 
[Nn] to become [Ngr] is a derivative language.

 > Or are you suggesting this was a CL pronunciation that didn't survive 
into VL?

Well it didn't, did it? See above.

-- 
Ray
==================================
[log in to unmask]
http://www.carolandray.plus.com
==================================
Nid rhy hen neb i ddysgu.
There's none too old to learn.
[WELSH PROVERB}