"You mustn't know too much," he says, describing how he captures "the decisive moment" on film. "There's nothing to know. Cats know more than human beings on the subject. Cats sniffle: sniff-sniff. Intuition. People use brains too much. Brains are not used for making love."
Henri Cartier Bresson when interviewed by David Friend.
That does sound to me as the closes you can get to create a novel image. Forget thinking too much and explore the medium. When you explore you make mistakes, and some of them are by far more interesting than the average good shot. It's all about the interesting mistakes and the things that are a bit different. Once you have localized those best mistakes you can follow them, and create a new imaginery out of those new sets of ideas. Bad pictures don't follow the rules, great pictures break them.
"Being a (good) photographer is about taking good pictures, IMO. The bad ones don't matter; you learn from them and edit them out. If you don't allow yourself to risk taking a bad picture, your photographs will be staid, safe, boring. Simply trying not to fail is no way to succeed. It sounds suspiciously like a way of finding comfort by lowering the bar."
I read that here. Makes sense. Unless you make a living out of a formula. Shoot, get it wrong, explore. Later you'll have time to edit out and develop. My next toast will be to anarchy.