Let me quote Ennio De Giorgi on this: "If you can't prove your theorem, keep shifting parts of the conclusion to the assumptions, until you can"