But…
But there are two problems with this, too
- The “magic” approach to representing recursion without recursion falls apart for applicative-order evaluation; a special reduction rule for recursion must be introduced
- It isn’t always faster to evaluate