[I’m unsure how much sense these paragraphs make without the preceding 15 pages or so. First draft material.]

Allow me to reiterate my thinking about the three illusions in PI As I have said, I think that transcendental illusions are central to PI, central to understanding its conception of philosophical problems. But I do not mean to deny that the other two types of illusion make appearances in PI. They do. If they didn’t, the other two readings of PI I have discussed would probably never have tempted anyone. But I do not think either of those other two illusions is central in the way transcendental illusion is central. Even more, although I do not have time to detail this now, I also think that the other two types of illusion can metastasize into transcendental illusions, and that what seems merely an empirical or logical illusion can itself be revealed to be or involve a transcendental illusion.

Still, I suspect at this point that you may be less interested in this nicety of housekeeping and more interested in the spirit-sinking midwinter bleakness of the picture of philosophical problems that I have been presenting. Haven’t I said that philosophical problems don’t get solved? —Yes. Haven’t I said that the best we can hope for is to cope with them, and that only temporarily? —Yes. Isn’t that an incredible downer? —Yes. —No. —Yes and no.

Let me revert to Kant once more, while also recalling my opening remarks. At least from his early 20’s, Kant is vexed by the uncanny fact that for so long and so often philosophers of equal education and gifts, each deeply serious and sincere, persist in apparently irreconcilable conflict. Kant regards this as a disgrace to reason. We can think of Kant’s entire philosophical career as driven by his passionate concern to settle conflicts in philosophy—not by entering a judgment in favor of either of the conflicting philosophers, but rather by entering a judgment on the conflict itself, by finding a way to end it. Kant first pictured these conflicts as between philosophers; but he later came to picture them as conflicts between arguments; and still later he came to picture them as reason in conflict with itself, as reason having fallen out of agreement with itself. This became Kant’s new picture of the uncanny fact: reason can fall into apparently irreconcilable conflict with itself.

Wittgenstein we can think of as accepting and, in a sense, extending this movement of Kant’s. He comes to see philosophical problems as manifesting my having fallen out of agreement with myself. It is this uncanny fact about me—and about you too of course—that is the mainspring of PI. Philosophical problems are all at once conceptual, i.e., invulnerable to the empirical, and also deeply personal, personal in the sense that I am at stake in them in various ways, personal in the sense that the problems encroach upon me. My desires and longings, my needs and fears, –my whole affective being potentially is part of the problem. Because it is, we cannot take a simply objective approach to the problems. We must instead always and everywhere approach the problem subjectively, attending to our mode of involvement in it. And there is always already one of those whenever we reflect on a philosophical problem. (Reflecting on a philosophical problem is not an empty holding of something, of the problem.) I am a part of, involved in, every philosophical problem that I take up, like it or not. You too are part of every philosophical problem you take up. The mode of involvement in the problem matters and is, in fact, itself a part of the problem. When we see this, we are also position so as to see what PI is showing each of us: I am susceptible to transcendental illusion. I can, with practice, develop a discipline of response to such illusions. I can teach myself—-with the aid of PI—-to push back against the pressures of transcendental illusion, to stick to ordinary realities. But learning this discipline neither ends philosophy itself for me nor ends any philosophical problem permanently. I have to recognize that I am tempted and I need to find a way of responding to the temptation other than yielding to it. And I have to recognize that there is no way simply to end temptation. I should expect to be tempted until my last breath. But none of this makes much sense if we think that philosophical problems can be independent of modes of reception. To care about illusion as Wittgenstein does is to care about modes of involvement and vice-versa: To use `illusion’ as a term of criticism as Wittgenstein does is to bring the mode of involvement in a philosophical problem into work on the problem.

Philosophy, as Wittgenstein teaches it to us in PI, is transcendental dialectic, but Wittgenstein’s transcendental dialectic is self-critique. Philosophy is self-critique. Philosophy is a form of self-knowledge. Is that a downer? No: there is philosophy to be done. There is always philosophy to be done. Self-critique has no endgame. Even more, there is progress of a sort that can be made. I can become better at resisting temptation, better at recognizing its onset; I can discipline myself to push back harder and for longer and with better focus. That is, I can come better to know myself in such a way that I become better at controlling myself. I cannot make myself impassible, invulnerable to temptation. But I can respond to it better, yield to it less often or less easily. Still, isn’t that a downer? Yes: since such self-knowledge is, face it, almost always bitter. (What I discover will not be things of which I have been ignorant, simply epistemically blank, but things I have refused to know, to acknowledge.) We would all rather not do philosophy as self-examination, self-critique. We would rather do it as distinct from the self, as objective investigation in which my self does not count. We would like to be the investigators of the investigation but not the investigated. We want to be neither shaken nor stirred.

One quick parting comment: We can understand the similarity and difference between Kant and Wittgenstein and their transcendental dialectics in this memorable way, even if it may be slightly misleading: for Kant, we see things aright philosophically when we come to see that the solutions to philosophical problems must be transcendental idealist solutions. For Wittgenstein, we become what we need to be philosophically when we become `transcendental idealists’ about the problems themselves. Wittgenstein’s Copernican Revolution (PI 108) revolves the problems—and, thus revolved, we can see past the apparent need for a solution to them. So, since I have been talking about downers, let me end with a hurrah: Vive la revolution!