This is straight from Saussure, and then Chomsky. You can read his first course in linguistics [1].
It was highly disputed in the 1800s.
I'm not an expert, but:
> Uh, what? How do you think we got more than one language in the first place?
We would compete for resources, new languages would be created rapidly from old languages instead of the gradual change we're seeing in the last 4000 years.
> How do you think we have entire language families with mutually-unintelligable members?
All language, like I said, is composed of many -emes. Chomsky's big discovery was that you can't seaparate syntax and semantics when it comes to any language.
> Or what about dialectal continuums? If anything, I would posit that we are closer to your statement now than at any other time, because the printing press calcified spelling -- which could vary quite a bit in the time of scribes. And radio and TV and the Internet have started spreading linguistic innovation faster.
This is the standard intuitive view with when one has no evidence. There is strong evidence to suggest that it is the individuals who mispronounce written words who cause this rapid change in language--they are mispronounced and then subsequently miswritten, and then their semantics are subsequently misinterpreted, so their syntax and semantics slowly change in this fashion.
Saussure, the founder of the modern linguistic tradition, has this all outlined in his book [1]. He outlines your view, and then shows you that it is in fact not true.
[1] https://en.wikipedia.org/wiki/Course_in_General_Linguistics