If you were to ask how far back we can track continuous developing cultures living in the same location, then I can answer your question! The answer is a very long time. For example the excavations at Namu (central coast of BC) suggest unbroken habitation reaching back around 10,000 years, with no evidence of the community being replaced (no layer where things suddenly change). This jives with oral history that seems to be accurate in the area going back beyond this time, with stories of villages, events, landmarks, that have recently been associated with some excavations something like 14,000 years old.

So in other words, people's children have been raising their children in the same places on the central coast for a very long time.

In the prairies on the other hand, and most other places in North America I know a lot less about concrete dating of older sites, or for that matter about site continuity, but I will be able to reference these areas in the next section of my answer, which is this:

cultures, even physical culture, doesn't equal language! Today, the central coast is a potlatching culture, politically linked across at least four language families. Although there are markers that might let you distinguish a Nuxalk village from a Heiltsuk village at different times, culturally they might be very similar, and you often have two villages living almost together speaking the different languages, and until recently there were a lot of places where being bilingual or trilingual was the norm. Going back in time, there's pretty good arguments that the gradual drift of languages can only be tracked back about 10,000 years, beyond which it's a game of telephone just too far gone (strangely enough ,stories seem to change less than languages!). So remember that village that has been continuously inhabited for ten millenia? just because it was the same people, doesn't mean their language didn't change many times.

Nuxalk for example, is a mixture of root words from different origins. Proto-Salish, coast Salish, Interior Salish, Proto-Penutian, with a liberal does of several different Wakashan languages and a bit of Proto-Athapascan. While there are many different possibilities for how it ended up that way, a likely model (given what I know or have surmised about the history of the area) is that there was a mixture of communities in the valley, probably with some of the oldest being Penutian speakers, and some of the younger ones (but still several millennia ago) being Salish, and eventually the Salishan speakers ended up as the common language, with the vocabulary of the other languages remaining in the new communities. As a result there are names in stories that don't translate well into Nuxalk names, not to mention names that come from many other languages. Later on the Wakashan people came up the coast, and marriage and contact added more in.

The result is this language as it is today, representing a culture that is continuous, but whose ancestors may at one time have spoken several different languages at the same time in close proximity to each other - with their political identity being determined through marriage ties, location, potlatching patterns and such, rather than language.

In the prairies fairly recently we have several good examples of similar things going on - the Nehiyaw Pwat was an alliance of 4 peoples who spoke at least six different languages, yet politically were very closely tied to each other for well over a hundred years, some going back far further. There is no way to know that similar alliances and connections did not take place going back further into history, without necessarily having an impact on material culture.

One clear example of this is in perhaps deeper pre-history - the clovis culture. We have a culture that spanned North America prior to the end of glaciation, yet the oral histories of multiple languages families suggest that they came up the central coast at the end of glaciation, already speaking these varied languages. Looking at the landscape at the end of glaciation we have a lot of small biomes, rapidly changing as the climate adjusts to the gradual retreat of the ice. Swamp, tundra, prairie, forest, all were present in the lives of most people, as the biomes weren't yet large enough for specialization, and a physically similar culture could exist between languages.

The above is definitely only my own conjecture - I see evidence of multiple language families existing within a single material culture, based on my reading of oral history, backed up by what I observe in the recent past, of cultural and political affiliations being more important to identity than linguistic ones.

Now - I'll finish by saying this - look at Asia - we have Korea, Japan and China, with many other small nations in the area as well, all using a large amount of Chinese vocabulary despite no blood connections. We see the French and Spanish speaking Romance languages probably again more as a remnant of political and cultural affiliation and prestige than one of blood, and we see Jewish people speaking Germanic languages, Roma people speaking Basque mixolects, and much more - so while things are common, they are not the rule, and going back in history looking for continuity is always tricky.