They might say, "Clojure is a lisp," but when referring to "Lisp" (i.e. without an article "a" or "the") they're talking about CL.
All these languages are somewhat compatible with the original Lisp 1.
Then there is the wider Lisp family which includes Scheme, Dylan, Clojure, ECMAScript, Racket, SKILL, Hop, Logo, ...
Certainly Doug Hoyte, the author of this phenomenal ~400 page book on Common Lisp programming, is aware of that.
In any case, what I was responding to in your earlier comment was the implication (I thought) that Hoyte doesn't understand "lisp as a family of languages with many dialects." But since you've read LOL, that cannot have been your point...
Common Lisp hackers may use "Lisp" colloquially as a shorthand for the language they're working with ("Common Lisp" is a bit of a mouthful, after all), but I don't think many of them are confused about the existence and nature of other lisps, or that they tend to disagree with academics on this.
The complexities surrounding state in CL are not unlike any other non-functional language. I find it hard to call a language functional if it doesn't strongly encourage immutability (or require it) and, preferably, have persistent data structures.
It's for these two reasons that I don't think Swift is functional, and I don't consider CL to be functional.