It's just a loust in the category of endosequences.
I think that was the article that made me actually try to understand "A monad is just a monoid in the category of endofunctors". Researching "endofunctor" helped significantly more than any of the analogies floating around at the time.