Now that Tim has done all the hard work of describing the Seiberg-Witten equations, I'll try to motivate the structure of the 3+1 dimensional "TQFT". I'll start by reviewing Morse theory on a closed manifold, the finite-dimensional model for Floer theory. I'll then discuss the right way to adapt this story to a manifold with boundary. The latter serves as the finite-dimensional model for monopole Floer homology, leading to three versions of the theory.