Kant invented the conception of morality as autonomy. I use the notion of invention as Kant himself did in an early remark. “Leibniz thought up a simple substance” he said, “which had nothing but obscure representations, and called it a slumbering monad. This monad he had not explained, but merely invented; for the concept of it was not given to him but was rather created by him.” Autonomy, as Kant saw it, requires contracausal freedom; and he believed that in the unique experience of the moral ought we are “given” a “fact of reason” that unquestionably shows us that we possess such freedom as members of a noumenal realm. Readers who hold, as I do, that our experience of the moral ought shows us no such thing will think of his version of autonomy as an invention rather than an explanation. Those with different views on freedom and morality may wish that I had called this book The Discovery of Autonomy. We can probably agree that Kant's moral thought is as hard to understand as it is original and profound. Systematic studies from Paton and Beck to the present have greatly improved our critical grasp of his position. In this book I try to broaden our historical comprehension of Kant's moral philosophy by relating it to the earlier work to which it was a response.