What is this based on, Vrylakas? From what I've heard, languages of sedentary people tend to develop a bit differently from the clear way Indo-European langauges have developed; you get more of a "web" or "mix" of languages than a straight-line genetic relationship.
This isn't a new idea; Colin Renfrew proposed it in the late 1980s as a reaction to the highly romanticised picture of burning-and-slashing barbarian ancestors who, as true Aryans, swept across Eurasia leaving fragmented forms of their language wherever they went. His chief point was that the Kossinian connection between language, ethnicity and the archaeological record just doesn't add up. Almost assuredly there have been instances of Indo-European speakers conquering others, but to explain how a language family became so widespread, how it came to be represented across such a massive landarea and across such diverse topography, one needs much more. The spread and development of language is more complex than simply wild barbarians riding out of the hills and slaughtering all.
One aspect that has been ignored until very recently is the roll of indigeonous peoples - the folks who were already here when the historical peoples we know about showed up. We also know next to nothing about trading relationships and how far widespread they were in the pre-Classical and even immediate post-Classical era, for instance. One of Renfrew's more radical (though plausible) thoughts is whether or not the Celts were truly ever a single people. Have modern archaeologists and paleo-linguists manufactured a single ethnicity out of what may be many scattered and unrelated groups who were bound together by trade, and came (mistakenly) to be by moderns described as a people because of the wildfire spread of iron age technologies?
Bottom line: In the late 19th and early 20th centuries we thought we had the Indo-Europeans pretty much all worked out. There were even some linguists who were claiming to be able to reconstruct the original Indo-European language through etymological studies. Sure, there were a few loose strings, but the big picture was all worked out.
With advances in ethnography and archaeology in the 20th century, however, we've come to see increasingly that we know much less than we thought and many of the earlier assumptions were skewered, wrong or just politically-motivated.
Speaking of Celts:
Who were the people they conquered, and how much did that effect the structure of language?
That's a question that only recently linguists have started asking. To date most scholars looked at the pre-Indo-European peoples as cultureless, empty vessels just waiting to be either slaughtered or assimilated by our linguistic ancestors. The problem of course is that we are increasingly understanding that they didn't just go away, and the question naturally follows that if they didn't just go away, then what kind of cultural or linguistic imprint did they leave? The short answer: We just don't know yet. Stay tuned.
If they didn't conquer but mearly lived or mixed with those already there, how do they know that its indo european languages that are being spoken and not that of the original inhabitants, or a mix of both?
The easiest initial method is to find if any given word in one language has identifiable matche sin other Indo-European languages. For instance, the word "father" is fairly evenly spread through most of the Indo-European languages ("pater", "Vater", etc.) though few words are universal: "father" in the Slavic languages is a variation of "ojciec". The problem is that no language develops in a vaccuum; from beginning to end they are all influenced by other languages, and of course languages evolve. If you want to study Latin, you need to specify what era of Latin you want to study, because even after its formal death variations of it developed across Europe. Medieval European Latin is not the same that Caesar spoke. Modern Bulgarian is labelled a Slavic language and as such a member of the Indo-European family but it was originally a Turkic language spoken by a people who eventually conquered and ruled over the Slavs of the lower Balkans, and gradually fused (unintentionally) Old Bulgar with Slavic to make modern Bulgarian. If you study Slavic philology, you very commonly hear the phrase, "This is common to all the Slavic languages...except Bulgarian..."
How did they trace it back to the caucasus mountains just based on language?
First of all, there is no single theory that is fully accepted by the linguistic community about where the Indo-Europeans first began speaking their language. In fact, some theories claim that there never was a single Indo-European language, but rather several different linguistic fountains that inadvertantly spread the language through trade and technologies. It was once popular to try to trace the "ethnoi-genesis" of a people through their language to some definite geographical location, but again, much of the ideas behind that have since been called into question.
For tracing though, as I recall the Hungarians (a non-Indfo-European speaking people, although a very substantial proportion of their vocabulary is derived from the Indo-European lexicon - i.e., Slavic, German, Latin) traced the footsteps of their linguistic ancestors by tracing back the names of trees, birch trees in particular. This is a popular method. Basically, you categorize the names of trees or plants in a given language, then try to find cognates in other surrounding languages, meanwhile trying to find the geographical location of some of these plants. It is fascinating work, though inexact.
That seems like an almost impossible task.
It is. We just have to accept that there is much we will never know. It is tempting for some, sometimes for nationalist or political reasons, to push the evidence farther than it really can take us, but in truth much will remain a mystery.
One trait that is thought to be an old Indo-European trait is the old adjectivial suffix form "-ski". By now most perople think of this as a Slavic, and particularly Polish trait, but it was once nearly universal. Many modern Polish surnames are actually adjectives, which is why they end in "-ski", "-ska" or "-sko"; but other Indo-European languages also preserve this form: the guy who built the dome over the Florentine cathedral was named "Brunolescchi", a pure Italian whose name was pronounced "brunoleski". I've seen many Italian names with this ending. It also shows up in French, though archaically so, and through French in English as well: Kafk-esque, grot-esque, etc.
If you ever want to study the Indo-European languages, you will find yourself studying Lithuanian. It is, so far as we're able to determine, the language that has preserved the most of the original traits of the Indo-European language family. It is amazing to see; it fills up the page as if Lithuanians were determined to use every possible diacritic mark available to the Latin alphabet.
