Re: [tied] Latin pinso etc. (was: *(H-)p/bh[-r/l-] again again)

> why should be reduced the "ns" group to "s" when the "nC" group ( if C=
> stop or siflant) is very stable,
> This is not usually in Rom. and it must have happened already in Latin
> (ns > s) but I wonder under which circumstances.

The variation is very common in Latin. -Vns- > -V:s- (with compensatory
lengthening) before the classical period, and the -n- was later
re-inserted. That is why every vowel in Latin before -ns- is long.
Something similar happened before -nf-.

Spellings vary, and late Latin often shows -s- for classical -ns-.

Peter