There was a research project named Wyvern (not related to the current programming language Wyvern) that was intended to invent what would probably be called a "lexically scoped paradigmatic programming language."
Programming languages are built on paradigms. In programming language theory, a "paradigm" is a set of ideas that form a mutually exclusive way of expressing the meaning of a program. The imperative paradigm treats programming mechanistically: "do this, then do this, then do this"; the functional paradigm treats programming mathematically: "if you have these different transformations of the data, eventually you get the answer you want."
The notion of lexically scoped paradigmatic programming is back, and I still believe it's a bad idea. Jinnzest writes in favor of the idea: "[I]f a language, on the contrary, provides means to split code into separate paradigm we will have to examine only one of them at once abstracting away all other."
The idea of lexically scoped paradigmatic programming is simple: using some kind of syntax, we tell the compiler "Okay, now we're gonna go all-functional, but we're gonna glue it together with imperative," or "Everything in here should be assumed to be async and event-driven," and so on.
... and we're back to Lisp. Lisp in Small Pieces, specifically. (Seriously, anyone who is working with a programming language who has not worked their way through LiSP is missing out.) We're back to a world where reading someone else code is basically a "guess what's going on here" and "your idea of the environment, stack, and so on are laughable." Lisp is a great language, but only if your people are super-disciplined about using it one way and one way only; otherwise you end up with all the paradigms all mixed up in a bucket.
This is one of the reasons I recommend Go for a commercial venture that's trying to move away from Python and Node. Go has only one paradigm: imperative. It's not even a very good event-driven language; goroutines and message passing devolve into imperative code very easily.
On the other hand, it's a vile, ugly language, and that's why I like to write in Haskell for learning, Rust for performance and Lisp for elegance. So, you know, not in production.