But this poses an enormous problem, of which Kant appears not have been aware. What is the relationship between the empirical self and the self in itself, between the phenomenal self and the noumenal self? Since the noumenal self is the moral self, the self that abides by the Moral Law, it is clear that the noumenal self is aware of itself. It is, Kant sometimes says, Practical Reason [thus identifying the bearer of a faculty with the faculty.] The empirical self, which experiences temptation, desire, pleasure and pain, is the appearance in the realm of phenomena of the noumenal self. I cannot recall that Kant ever says this straight out, but there is really no other possible answer to the question.
Let me try to clarify this by means of an analogy. Suppose that I decide to write a story about my family when I was a boy. In the story are a number of characters -- my father, my mother, my sister, my aunts and uncles, my cousins, and my grandparents. One of the characters in the story is called "Rob." [Since my big sister, Barbara, had already appropriated "Bobs" as her nickname I couldn't really be called "Bob," so I became "Rob."] All the characters in the story bear a relationship to me as the author of the story, but the character called "Rob" bears a special, a privileged, relationship to me as author. That character is the appearance of me in the story. No other character in the story can have that relationship to me as author. Now, my sister might decide to write a story about the family as well, and she might [though it is unlikely] write word for word the same story that I have written. In her story, there will be a character named "Bobs" who will have the same privileged relationship to her as author of her story that the character named "Rob" has to me as author in my story. There will of course also be a character named "Bobs" in my story and a character named "Rob" on her story, but neither of those characters will bear a privileged relationship to the author of its story.
This is, roughly speaking, the situation Kant is describing. I synthesize my manifold of sensuous intuition according to the rules for such synthesis, and produce thereby a story, if you will, that I call "nature." In that story are many objects -- the sun and the moon, the earth and all that is on it, and a number of human beings, among whom there is one bearing the name "Robert Paul Wolff" born in 1933 in New York City, New York. That human being is conscious -- indeed, he is self-conscious -- and he bears a privileged relationship to the noumenal self that has synthesized the manifold and has by so doing acted as "the law-giver to nature." That noumenal self is a moral agent, and it has certain obligations, says Kant, to other moral agents.
But how on earth can this moral agent ever encounter other moral agents in the realm of appearance that it has synthesized from a diversity of sense contents?
I want you just to think about this for a moment. According to Kant, I have moral obligations -- duties -- to other moral agents. Not to rocks or to trees or to horses or to human beings understood merely as natural things in the realm of appearance, but to other moral agents, which is to say other noumenal selves. But it is as a phenomenon, as an appearance, that I speak, make promises, tell lies, borrow money, kill other human beings, and do all the things that the Moral Law tells me I must do or must not do. And leaving aside the duties to myself, which Kant is sure to make a place for in his ethical theory but which do not really count for all that much, all my duties are to other noumenal selves. For the life of me, I cannot see how, on Kant's theory of a priori synthesis, I could ever even hypothetically encounter another noumenal agent in the realm of appearances.
I mean, we cannot each of us be telling numerically the same story, any more than -- to use an old philosophical example -- we can all be sneezing the numerically same sneeze. We might be telling qualitatively identical stories, by a sort of Leibnizean pre-established harmony, rather like two year olds engaged in what child psychologists call "parallel play," but Kant's very first foray into what became the Critical Philosophy, the Dissertation, decisively rejected the theory of pre-established harmony, and besides, even though it might somehow save physics, it cannot save morality. If I can never encounter another moral agent, then Kant's ethical theory is vacuous.
Well, no doubt you will along about now be waiting for me to tell you how this tangle can be unraveled, concluding with a triumphant reconciliation of Kant's ethical theory with his epistemology. But I cannot. No more can I reconstruct out of the text of the Groundwork a satisfactory demonstration of the validity a priori of a substantive fundamental principle of morality. Lord knows, it is not for not trying. I spent seven years puzzling over these questions. In the end, I wrote a book about Kant's ethical theory and a number of scholarly articles. But none of those writings contains a satisfactory resolution of this problem. Indeed, I do not think that it has a resolution. This has profound consequences for ethical theory, whether you are sympathetic to Kant's approach to the subject or not. But that is a story for another day. This "Introduction" has no gone on for more than thirty thousand words, and it is time to bring it to a close. I hope it has proved useful. I shall edit it slightly and post it on box.net, so that anyone who is interested can read it at his or her leisure.