Good post. I tend to think of Lisp/Scheme logical and conditional operators taking an implicitly defined monad for the type of "useful value" or "nil" (#f in S

Good post. I tend to think of Lisp/Scheme logical and conditional operators taking an implicitly defined monad for the type of "useful value" or "nil" (#f in Scheme). It's just that: 1. You can't add new instances to the type class. 2. The language doesn't check you handle the failure case. If there was a monad type class you could create a type "useful value" or "exception" (called Validation or Either in Haskell and Scala) and have the same concise code. However, without a static type system this kind of system just gets unmanageable (I know, I've tried it). Proponents of dynamic typing often wax lyrical about the benefits of escaping the straight-jacket of static typing. They often don't appreciate the kinds of programs that are impossible to write without a modern type system. NB: this post is not one of them.