Re: [tied] "b" versus "g"

I ask myself about the fallowing phenomenon:
"bi" is very distant from "ghi"
Normaly, after some criterias, if form a primordial word in
our case, called "xxx" in two languages we find children of
this word with the forms:
langauge1 = "bixxx"
language 2="ghixxx"
then we will speak about separat evolution of the word. From
this example, if we work on a bigger scala, we can say that we
will have two separated language.
In this short example I can say is the "esence" of the
indoeuropean ideea of a common langauge with actual
descendants.
My question is:
I have in romanian "bine"= good.
But I have in sub dialect of romanians the same word whichi is
pronounced "ghini"
And here I wonder. I should understud if tehre should have
been a "vini" a "pini" a labial one .
But ghi? How can be such kinds of divagations explained in the
same language at all?