Nuclear family was the place where original language was created. This is also the place where language is learned.
As family grows, nuclear family becomes extended family, then clan, then tribe, then nation. All along the same language is used between members of the this growing family to keep communication going. Family is preserved through communication and cooperation and for that you need a common language.
If all these people are linked genetically, through father to son inheritance, I believe that we can talk about genetic languages. R1a languages, R1b languages, I2a languages...As genetic tribes mix, genetic languages mix...If genetic tribes disappear genetic language might disappear as well, unless it was passed onto in full or in part to some other genetic tribe. Today's languages are evolved mix of old languages. This is why we can find ancient language structures and words in modern languages. Serbian language is a mix of R1a, I2a, E1b, R1b...languages. Irish language is a mix of R1b, I2a, I1, R1a...languages. But Russian is a mix of R1a and N languages. Spanish is a mix of R1b, E1b...languages. Basque is pretty much pure R1b language. This is why we can find certain linguistic traits in Serbian and Irish but not in other Slavic languages or Atlantic (or as you know them Celtic) languages. Because both Serbian and Irish share I2a language traits not present in say Welsh. This is why certain words are found in some Germanic and Slavic languages but not in French or other Germanic languages. Because the languages who have these common characteristics share common R1a language..
This image, explains what I mean:
This is why I believe that we can still find traces of old languages and cultures in modern languages and cultures. Hopefully the picture is worth thousand words.
The modern languages which are most interesting and which can give us the most information about the old European languages, are the languages spoken by populations which are genetic mix of old races. If these populations still show distinct geographical separation of these genetic subgroups, and if this genetic separation matches dialect separation, we have clear proof of genetic family language mixing, resulting in mixed dialects. These types of languages usually have multiple unrelated words for the same thing, each used within a particular dialect linked to a particular genetic subgroup within the population. Some of those words belong to the core language spoken by the dominant genetic family, but the others are obscure and have no roots in the dominant genetic family language. So what these kind of languages can give us is the translation of the words from multiple old languages into a dominant genetic family language. This can give us a key for unlocking the old genetic family languages of the minority genetic subgroups. If we compare these dialects with dialects related to the same genetic family groups from other parts of the world, and we find the overlap, it is most likely that we have found parts of the old genetic family language of the particular genetic family group.
One of such mixed languages is Serbian.
Genetic, archaeological, linguistic, ethnological and other data point at very important fact that in Serbian language and culture we have unbroken cultural, linguistic and genetic continuity through R1a population, lasting at least 12,000 years. This is why we can use Serbian language and culture to answer some of the oldest cultural and linguistic questions that are still unanswered. But Serbians are not all R1a. Serbians are a genetically mixed population. We have pretty much every old Y haplogroup present in Serbia in significant numbers. Most importantly we have huge I2a population which is probably indigenous to the Balkans, and which has also preserved their own culture and language and have added it to the mix which is today Serbian culture and Language. We also have large old R1b population in Eastern Serbia which has also added their culture and language to the Serbian mix.
If Serbs were pure R1a race, then Serbian language and culture would not have been so interesting. It is the fact that Serbs are such a mix of races, and that Serbian culture and language are such a mix of cultures and languages, which makes Serbian (Croatian, Bosnian) language and culture almost like a living cultural Rosetta stone. People in Serbia, preserved some of the oldest cultural and linguistic traits which have been lost in other parts of Eurasia populated by R1a, R1b and I2 people. And they provided us with mutual translations of these cultural and linguistic traits, through thousands of interlinked local cultures and dialects.
We have the same situation in Irish culture and language. It also hides some of the oldest I2a, R1a, R1b cultural and linguistic layers still existing in the world. This is why when we cross reference Irish and Serbian culture and language we come up with almost like a code which we can use to decipher things from these ancient cultures.
The principle is universal and can be applied to any language and culture.
I just need to add at the end, that I believe that genetic genealogy is important tool that we can use to help us understand the past. Full stop. There are no better languages and cultures and worse languages and cultures. There are no superior languages and cultures and inferior languages and cultures. Just different, old and new, related and unrelated languages and cultures.
You can read more about the thought process that lead me to postulate this new theory of Eurasian languages, as opposed to Indoeuropean theory, here: