If I may, I think you're being a bit too dismissive of Russell's paradox.
We start with the observation that some sets aren't members of themselves: the set of stars in the Milky Way galaxy isn't itself a star in the Milky Way galaxy; the set of regular polyhedra isn't itself a regular polyhedron; and so on. It seems that we've easily found two items that answer to the well-defined predicate
S: is a set that isn't a member of itself.
Naively, we might assume that a set exists for every well-defined predicate. (For some of those predicates, it will be the empty set.) But what about the set corresponding to the predicate S? This question doesn't seem, on the face of things, to be nonsensical or ungrammatical. But the question shows that our naive assumption implies a contradiction, and therefore our naive assumption can't possibly be true.