But the point is that a story is exciting because it has in it so strong an element of will, of what theology calls free-will. You cannot finish a sum how you like. But you can finish a story how you like. When somebody discovered the Differential Calculus there was only one Differential Calculus he could discover. But when Shakespeare killed Romeo he might have married him to Juliet's old nurse if he had felt inclined. And Christendom has excelled in the narrative romance exactly because it has insisted on the theological free-will. It is a large matter and too much to one side of the road to be discussed adequately here; but this is the real objection to that torrent of modern talk about treating crime as disease, about making a prison merely a hygienic environment like a hospital, of healing sin by slow scientific methods. The fallacy of the whole thing is that evil is a matter of active choice, whereas disease is not. If you say that you are going to cure a profligate as you cure an asthmatic, my cheap and obvious answer is, "Produce the people who want to be asthmatics as many people want to be Profligates." A man may lie still and be cured of a malady. But he must not lie still if he wants to be cured of a sin; on the contrary, he must get up and jump about violently. The whole point indeed is perfectly expressed in the very word which we use for a man in hospital; "patient" is in the passive mood; "sinner" is in the active. If a man is to be saved from influenza, he may be a patient. But if he is to be saved from forging, he must be not a patient but an _impatient_. He must be personally impatient with forgery. All moral reform must start in the active not the passive will. Here again we reach the same substantial conclusion. In so far as we desire the definite reconstructions and the dangerous revolutions which have distinguished European civilisation, we shall not discourage the thought of possible ruin; we shall rather encourage it. If we want, like the Eastern saints, merely to contemplate how right things are, of course we shall only say that they must go right. But if we particularly want to _make_ them go right, we must insist that they may go wrong.