JOURNAL OF PHILOSOPHICAL RESEARCH VOLUME 36, 2011 I HOW TO BE A KANTIAN AND A NATURALIST ABOUT HUMAN KNOWLEDGE: SELLARS’S MIDDLE WAY JAMES R. O’SHEA UNIVERSITY COLLEGE DUBLIN ABSTRACT: The contention in this paper is that central to Sellars’s famous attempt to fuse the “manifest image” and the “scientiic image” of the human being in the world was an at- tempt to marry a particularly strong form of scientiic naturalism with various modiied Kantian a priori principles about the unity of the self and the structure of human knowledge. The modiied Kantian aspects of Sellars’s view have been emphasized by current “left wing” Sellarsians, while the scientiic naturalist aspects have been championed by “right wing” Sellarsians, the latter including William Rottschaefer’s constructive criticisms of my own reconciling interpretation of Sellars. In this paper I focus irst on how (1) Sellars’s Kantian conception of the neces- sary a priori unity of the thinking self does not conlict with his ideal scientiic naturalist conception of persons as “bundles” or pluralities of scientiically postulated processes. This then prepares the way for a more comprehensive discussion of how (2) Sellars’s modiied Kantian account of the substantive a priori principles that make possible any conceptualized knowledge of a world does not conlict with his simultaneous demand for an ideal scientiic explanation and evolutionary account of those same conceptual capacities. Sellars’s own attempted via media synthesis—what I call his “Kantian scientiic naturalism”— merits another look from both the left and the right. I. INTRODUCTION t is no surprise that Kantians have often been hostile to comprehensive forms of philosophical naturalism, whether in theoretical or practical philosophy.