Abstract

I explore dichotomous treatments of subjectivity and objectivity in accounts of knowledge of the world, others and the self. In Chapters 1-5 I argue that realism and relativism suffer forms of epistemic neurosis--each is undermined by its own account of objectivity. ;The realist sees the world and semantic notions, like truth and reference, as independent of our abilities to know about them. This implies the self-defeating result that we could be totally wrong about the meanings of our words, since meaning depends on our interaction with the world and with each other. ;The relativist sees rationality, truth and concepts as local to cultures, eras or persons. But this view defeats itself too, since any argument for it must use a non-relativist account of truth and rationality. It is tacitly realistic, treating other cultures and minds as independent of our epistemic capacities. Thinkers who accept the objective-subjective dualism can be driven to relativism by a distaste for scientism and a failure to see the contingent, historical character of the alliance of realism and scientism, which survives in varieties of Marxism and analytical philosophy. ;I avoid these problems by taking objectivity to imply that the world, other cultures and other minds are independent, not of what we could know, but only of our actual beliefs about them. Justification, truth and reference are normative--not objective, natural phenomena in the realist's sense, but not merely relative either. ;In Chapter 6 I consider the status of self-knowledge in the work of Descartes, Hume and Kant, and in Chapter 7 I examine recent efforts to explain the asymmetry in the justification of first- and other-person knowledge-claims. Linguistic competence, I argue, entails being able to articulate one's intentional attitudes, but not those of others. But this asymmetry is not absolute; it varies with the degree of experience and justifiable trust shared by particular persons. Finally, in Chapter 8 I argue that worries raised in radical theory about "unified subjects" do not rule out seeing the self as unified in its recursive capacity for self-description.