Interpreting Bernard Williams’s ethical philosophy is not easy. His style is deceptively conversational; apparently direct, yet argumentatively inexplicit and allusive. He is moreover committed to evading ready-made philosophical “-isms.” All this reinforces the already distinct impression that the structure of his philosophy is a web of interrelated commitments where none has unique priority. Against this impression, however, I will venture that the contours of his philosophy become clearest if one considers that there is a single, unchanging root conviction from which his ethical philosophy grows. Despite the perpetual motion of his philosophical thought—its erudition, originality, range, and unceasing forward momentum—still, I contend, there is something unchanging at the heart of it. I will show this by reference to three signature theses: internal reasons, the relativism of distance, and the porous borders of philosophy and history. I will argue that the root conviction of which these are the fruits is the conviction that the constraints of universal rationality seriously underdetermine how one should live. This, I believe, is the vision of the human ethical condition that constitutes the largely inexplicit yet utterly fundamental presupposition beneath Williams’s ethical philosophy taken as a whole. I label the object of this root conviction ethical freedom, and thus portray Williams as a philosopher of ethical freedom.

Our ideas about forgiveness seem to oscillate between idealization and scepticism. How should we make sense of this apparent conflict? This paper argues that we should learn something from each, seeing these views as representing opposing moments in a perennial and well-grounded moral ambivalence towards forgiveness. Once we are correctly positioned, we shall see an aspect of forgiveness that recommends precisely this ambivalence. For what will come into view will be certain key psychological mechanisms of moral-epistemic influence – other-addressed and self-addressed mechanisms of moral social construction – that enable forgiveness to function well when it is well-functioning, but which are also intrinsically prone to deterioration into one or another form of bad faith. Thus forgiveness is revealed as necessarily containing seeds of its own corruption, showing ambivalence to be a generically appropriate attitude. Moreover, it is emphasized that where forgiver and forgiven are relating to one another in the context of asymmetries of social power, the practice of forgiveness is likely to be further compromised, notably increasing the risk of negative influence on the moral-epistemic states of either the forgiver or the forgiven, or both.

Our understanding of social experiences is central to our social understanding more generally. But this sphere of epistemic practice can be structurally prejudiced by unequal relations of power, so that some groups suffer a distinctive kind of epistemic injustice—hermeneutical injustice. I aim to achieve a clear conception of this epistemicethical phenomenon, so that we have a workable definition and a proper understanding of the wrong that it inflicts.

In this paper I respond to three commentaries on Epistemic Injustice: Power and the Ethics of Knowing. In response to Alcoff, I primarily defend my conception of how an individual hearer might develop virtues of epistemic justice. I do this partly by drawing on empirical social psychological evidence supporting the possibility of reflective self-regulation for prejudice in our judgements. I also emphasize the fact that individual virtue is only part of the solution – structural mechanisms also have an essential role in combating epistemic injustice. My response to Goldberg principally concerns my perceptual account of the epistemology of testimony, which I defend as being both well-motivated and best categorized as a species of non-inferentialism. I also explain its relation to the reductionism/non-reductionism contrast, and defend my resistance to casting it as any kind of default view. In response to Hookway, I contrast discriminatory with distributive forms of epistemic injustice, and defend the basic taxonomy I present in the book, which casts testimonial and hermeneutical injustice as the two fundamental discriminatory forms of epistemic injustice.

We are perhaps used to the idea that there are various species of oppression: political, economic, or sexual, for instance. But where there is the phenomenon that Nancy Hartsock picks out in saying that the world is “structured” by the powerful to the detriment of the powerless, there is another species of oppression at work, one that has not been registered in mainstream epistemology: epistemic oppression. The word ‘structured’ may be read materially, so as to imply that social institutions and practices favour the powerful, or ontologically, so as to imply that the powerful somehow constitute the world. But for present purposes I am interested only in an epistemological reading, which implies that the powerful have some sort of unfair advantage in “structuring” our understandings of the social world. I will try to present an account of what this initially vague idea involves. I hope thereby to explain an exact sense in which the powerful can have a kind of epistemic advantage that means the powerless are epistemically oppressed.