>If I'm in my derivation of 'crassus', *grass- is the original form:
> --- In firstname.lastname@example.org, "Marco Moretti" <marcomoretti69@>
> > > 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.
> 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