Chapter 6 turns to metaethics and explains how the results in normative ethics might lead to a skeptical challenge. I start by revisiting Duhem and Quine and how they thought underdetermination should impact our view of science. I then explain why I understand the major challenge that moral underdetermination poses to be an epistemological, not a semantic, one. This leads me to sketch out the general lines of a skeptical argument to the effect that we should withhold belief in the explanatory claims of moral theories. I propose a standard form of the skeptical argument in science and show how it can be adapted to ethics.Finishing up, I address two possible objections. The first concerns the question of whether so-called theoretical virtues might tip the ballance between two deontically equivalent theories. The second questions whether withholding of belief is indeed the correct reaction to equally believable moral theories. Iargue that the skeptical argument in ethics is at least as plausible as, if not more plausible than, its counterpart in the philosophy of science.

Chapter 7 discussesa novel position in metaethics that the skeptical argument might give rise to: constructive deonticism. This position is structurally analogous to one of the most discussed anti-realist positions in science, Bas van Fraassen’s constructive empiricism. I start with an overview of van Fraassen’s view, discussing both its most important features and how it relates to its main rivals, scientific realism and logical positivism. Next, I flesh out the new position in ethics, focussing on the pragmatic understanding of moral explanation that it entails. Following this, I discuss how constructive deonticism should be classified as a metaethical position. I show that it is clearly not part of the expressivist family and at most half error theory (or fictionalism). However, the position is arguably not what realists are looking for either, instead prompting us to rethink how the metaethical realism debate has been framed. Finally, I point out what I consider to be the two most important challenges, i.e., whether we can draw such a sharp distinction bewteen deontic and explanatory claims and whether the position is a stable one.