I would like to end by recapitulating the main claim of the transcendental interpretation divesting it of some of its Kantian garb. Though I have concentrated on Kant's text I think that the position it outlines should still be in the fore in contemporary discussions of epistemology. For this reason it may prove profitable to present it without commitments to which few modern Kantians would adhere. The transcendental interpretation reads the Appendix as offering a necessary condition of empirical knowledge. The Appendix claims that the mere physical event of sensing is no ground for claims of knowledge. A natural occurrence cannot justify a conceptual claim, nor can it supply a concept with meaning. It is through concepts that the world is given to us. And conceptual relations always receive their meaning through their place in a system of knowledge. It is through their place in a system that best makes sense of our encounters with reality that concepts gain meaning and empirical claims - truth. According to the transcendental reading, the Appendix adds a necessary condition of knowledge to the matrices of space and time and to the categories. But, significantly, the argument seems not to depend on commitment to these conditions. Thus, the argument is applicable to philosophical positions which are broadly speaking only Kantian. I have here in mind, first, Wittgenstein's claim that the human form of life is lived through language. Language, we might say, is for Wittgenstein the one transcendental condition of a human life. The argument is also applicable to McDowell's brand of Kantianism, influenced by Wittgenstein and Sellars, in which the supersensible and with it the categorial framework is discarded. Indeed, a version of the argument of the transcendental interpretation is explicitly endorsed by Sellars. Sellars holds that observation is the ultimate source of empirical knowledge. But he claims that "one couldn't have observational knowledge of any fact unless one knew many other things as well." Thus, the regulative idea of the systematic unity of knowledge may function as a single transcendental condition of empirical knowledge. It is significant too to note again that the argument does not depend on the specific shape Kant gives the system of knowledge he envisions. The argument is not committed to any particular view of what a comprehensive system of knowledge would look like. Finally, empirical knowledge of the world is gained under the necessary assumption that reality is made to be known, indeed is knowable through and through. The presupposition of the purposiveness of nature for our cognitive conceptual capacities is a necessary condition of knowledge. But it is conceivable that reality is not amenable to conceptual investigation. If the diversity of nature cannot ultimately be overcome then reality cannot be known by us. We are engaged in the relentless struggle to extend and refine our knowledge, and so, the world. The task seems reasonable and consistent. But, viewing it critically demands two conflicting standpoints. We must be aware of the limits of our knowledge, and so, of the bounds of intelligibility. We must believe - with no compelling reason - that these boundaries can be pushed forward to give us more knowledge and a greater world.