Note that this won't actually solve the original problem; Haskell is an implementation of a strongly typed lambda calculus, not of number theory, and Haskell collections cannot (easily) contain elements of different types --- so the empty set is not an element of a set, and the empty list is not an element of a list.