Lisp isn't semantically different from other languages, only syntactically - sure things are expressed as `map` instead of a for loop[0], but what's important is the memory layout of your data and whether or not it'd actually be faster to parallelize it, rather than if you can force it to work that way. So as long as the program works the same way it still can't/shouldn't be parallelized.

[0] except when Lisp programmers brag about how cool their complicated `loop` macro is