Research

Bart Streumer argues for an error theory about all normative judgements, but argues that we cannot believe this error theory, since believing a proposition is incompatible with at the same time believing that there are no normative reasons for believing it. Rather than this being a problem for the error theory, Streumer argues that it makes it more likely to be true by undermining several objections and revisionary alternatives to it and making it harder to reject his arguments. In my paper, I argue that we can distinguish two ways in which the error theory is unbelievable: a thing can be unbelievable in that it is strictly impossible to believe, and/or it can be unbelievable in the sense that it is simply very unintuitive. I argue that the latter sense of the word carries the ways in which Streumer thinks the error theory is strengthened by its unbelievability. My distinction strengthens the error theory's case by making it resistant to counterarguments resting on revisionary notions of belief, and it means that even if the unbelievability thesis fails, the error theory might still be plausible.

Draft available upon request

Error theories across a range of domains share a common structure: claims about a subject matter K essentially ascribe a property P; nothing instantiates P; thus, all K-claims are in error. In debates about error theories of propositional attitudes and responsibility, critics have pointed out that this inference presupposes a substantive account of how the relevant discourse secures reference: unless the property identified in the conceptual claim is genuinely reference-fixing, the failure of that property to exist may motivate revision of the discourse rather than elimination. This paper argues that an analogous "reference problem" has not been explicitly appreciated in discussions of moral error theory. I examine several candidate accounts of reference—meaning-driven, theory-driven, context-driven, and reflective equilibrium approaches—and argue that the meaning-driven strategy often relied upon by moral error theorists is difficult to defend. I also consider the possibility that reference is pluralistic or indeterminate, and outline several ways in which error theorists might attempt to avoid the problem. The aim is to show that standard arguments for moral error theory require a more explicit defence of their underlying theory of reference than they have typically received.

Draft available upon request