To underscore George's point about the importance of evidence in science, ask yourself this question: would Kant have been doing worse astronomy had his picture been wrong? Why?