From: "Florian Rivoal" <[log in to unmask]>
>         Your answers makes it more clear. Yet a couple of questions
> remains. How can the case of english in USA, or latin in western
> europe be similar to PIE? Those two are conveyed by One single
> civilisation or empire. the whole area is influence by one culture.
> Was there such a Proto-indo-european empire reigning over an area
> as vast as the whole indo european area?

There is a lot of speculation about the nature of the Indo-European
people/community outside of the linguistic matters.  But the matter of the
language is somewhat useful in itself... Knowing the language had common roots
like *sneygwh- or *kwekwlo- or *dyews implies something about the people who
used such words ('snow', 'wheel', and a sky deity)

> I think it is hard to
> believe without a writing system or a more or less advanced
> technology, which is important for maintaining coherence on such a
> big area for a long time. I do not doubt some comunities had the
> colonial and dominating style at that time to, i just think their
> influence was restricted to smaller areas.

PIE is already constructed as having many major branches that split off as its
people moved around.   (So the Anatolian branch from people who lived in this
place, and the Greeks here, and the Indo-Iranians here, and Germano-Balto-Slavic
here, and Italo-Celtic here, and Tocharian here, etc.)

> And how can there be only one "winner" on such a wide area?
> Wouldn't it be more reasonable to think that this reconstructed
> root comes from one proto-indo-european language, and that other
> comes from a distinc language, potentialy non related?

It's not just the reconstructions of words that prove relation but their
morphology as well; also languages have characteristic root shapes...

If you reconstructed a word like *ataraSii or *kaeru you could recognize it
wasn't normal PIE word by the phonology (/S/, vowels in hiatus...)   If you
reconstructed a word like *wahine (sg.), *wa:hine (pl.), you know it's not
normal PIE by the morphology (PIE didn't form plurals that way).  But in general
you wouldn't find words like this, because the languages that were displaced
don't survive, and all the evidence of them is seen through their being
assimilated by PIE.  There _are_ words that PIE and its dialects borrowed that
are generally identified as being loanwords either by their strange forms or on
other evidence:  words like *abel- ("apple", strange for having *b, thought to
be from a European lang) or *septm= ("seven", oddly shaped altogether, thought
to be from Semitic).