Re: Latin c- > Romance g-, any explanation?

--- In cybalist@yahoogroups.com, "Marco Moretti" <marcomoretti69@...> wrote:

>
> > Is there any explanation for the trend *c->g- present in some Romance words?
> > 1- Latin cattus > Romance *gattu > Portuguese gato
> > 2- Latin crupta (<crypta) > *grupta > Portuguese gruta
> > 3- Latina crassa > *grassa > Portuguese graxa
> >
> > Joao SL
>
> Something similar is found extensively in Romance languages and
> also in Italian:
>
> Latin /cattus/ > Italian /gatto/
> Latin /crassus/ > Italian /grasso/
> Latin /crupta/ > Italian /grotta/
> Latin /cavea/ > Italian /gabbia/
>
> and also /pr-/ sometimes > /br-/
>
> Latin /prui:na/ > Italian /brina/
> Latin /pra:vus/ > Italian /bravo/ (with semantic
> shift "cruel", "fierce" > "brave" > "able", and cfr. also the old
> meaning of "hired assassin").
>
> Perhaps ancient dialectal variants?
>

IMHO this is due to substrate influence on Vulgar Latin. In languages which had a tense/lax contrast on plosives rather than voiceless/voiced one, Latin voiceless plosives would have been assimilated to the native lax ones. This has happened, for example, in Latin loanwords to Basque (e.g. castellu- > gaztelu).

This lenition process has actually happened to INTERVOCALIC stops in Western Romance, and has been explained (see for example A. Martinet: Economie des changements phonétiques, 1955) by a Celtic substratum. However, in these languages Latin voiceless plosives have been kept at word initial in most cases. This could be explained assuming they were realized as tense like intervocallic geminates (in fact, this is what happened to the rhotic r-, which is throughly pronounced as [r:]).