When learning any new sufficiently abstract typeclass (whether it is Monad, Applicative, Functor, Monoid, Arrow, or whatever), the best way to learn it is in tw

When learning any new sufficiently abstract typeclass (whether it is Monad, Applicative, Functor, Monoid, Arrow, or whatever), the best way to learn it is in two phases: 1) What is the mathematical meaning of the typeclass? 2) How can the typeclass be applied (what are the instances) and how do they map to reality? It is very tempting to try to jump quickly to 2 - but it helps to force yourself to avoid thinking about applications for a bit and understand the applications. I think the article does achieve this.