I think that the original monad papers explain things better than the myriad of short tutorials tutorials around the interwebs. In particular, I like this one f

I think that the original monad papers explain things better than the myriad of short tutorials tutorials around the interwebs. In particular, I like this one from Simon Peyton Jones because it gives some historical perspective: http://research.microsoft.com/en-us/um/people/simonpj/papers/marktoberdorf/ http://research.microsoft.com/en-us/um/people/simonpj/papers... The TLDR: Haskell uses lazy evaluation, meaning expressions are evaluated in an umpredictable order. If expression evaluation had side effects, programs would become unmanageable, so in Haskell, al computations must be pure. Monads are a neat trick where you use a specially crafted abstract (opaque) interface that lets you model side-effecting computations, specifying what order things should run in. Of course, initially the biggest reason for using monads so prominently is the IO but the monad interface is much more general and an also be used for other things (List, Maybe, etc)