Abstract
In this dissertation, I develop a non-representational approach to metaethics that avoids the ontological and epistemic problems that traditional realist views face: accounting for the nature of moral facts and our knowledge of these facts in a way that comports with naturalism. According to expressivism—the most popular non-representational account in contemporary metaethics—moral claims don’t aim to represent moral facts or properties, but instead function to express our non-cognitive attitudes like approval or disapproval. This is a promising approach, but the view has substantial difficulties accounting for important features of moral discourse, like objectivity and complicated moral reasoning. I develop an alternative to expressivism: metaethical inferentialism, which gives us a new way to deal with the problems expressivism faces. If we understand the meaning of moral concepts by way of the inferential roles they have instead of the attitudes they express, we can see why these concepts have meanings that stay constant even in contexts where they aren’t used in the typical way, and in doing so solve the infamous Frege-Geach problem. This approach offers simple answers to at least two problems that have vexed contemporary expressivists—the “problem of permissions” and the commitment to “mentalism”, both of which are problems readily solved by an inferentialist approach. I also argue that moral objectivity that the objectivity of morality is a function of the inferential roles that we should expect moral concepts to have, given their purpose—the coordination of social behavior.