>In Portuguese the palatalization of s before
consonants and in final syllabes occurs
> in Portugal, but it's not general in
Brazilian Portuguese
Exactly palatalized "s" is the marked
member of no-palatal/palatal oposition therefore it is not strange that
phonetic change: s' [palatalized] > s [dental] actually had
occurred in some regions of Brazil (people q emigrated to Brazil did
not have languages with s', the substratum of these pre-Brazilian immigrants
favors the change!)
If in Portugal Latin s [dental] were
replaced by noticeable member of the oposition we must look for an
explanation. But the phenomenon is repeated in the other romanic
languages of the Iberian Peninsula, therefore I find
reasonable look for something in the substratum that conditions such
changes [the Iberian has two non-voiced sibilants: s and
s'; Basque has three: s (dental), z (alveo-palatal), x (palatal)
].
David Sánchez