There was a scientific American article on the evolution of feathers (last year sometime I think).
One idea is that they started as heat regulating devices.
Once feathers are there they can act against the air to various degrees. There has been further work on that. Namely allowing an animal to run up a slope better. It seems the kinematics works out well.
Those are a couple of the idea that allow a wing to get there in stages.
I think, but don't know, that feather fossils show a series of developmental steps to the complex, asymetrical feathers needed for real flight.