One of the most popular objections to non-naturalist moral facts is the queerness objection given by Mackie. As Olson points out, there are four distinct queerness objections: supervenience, knowledge, motivation, and irreducible normativity (or categorical reasons as I’ll call it). In this post, I’ll skip the first three and focus on the queerness of categorical reasons.
Categorical reasons are reasons that do not depend on your goals/desires, while hypothetical reasons are reasons that do depend on your goals/desires. Many have rejected moral realism because of the metaphysical queerness of categorical reasons; these reasons are supposed to have binding force while being free-floating and unattached to desires. Hypothetical reasons, on the other hand, as attached to desires, are seen as perfectly natural and metaphysically acceptable. Consequently, moral anti-realists like Mackie and Joyce have opted for hypothetical reasons.
Bedke has argued that hypothetical reasons are also queer. If that’s right, one can’t accept hypothetical reasons and reject categorical reasons for being queer. To show that hypothetical reasons are queer, Bedke first analyzes what type of thing a reason is.
Symbolically, then, we can represent reasons as three-place predicates of the following form:
R(F, A, Φ) in C,
which we read ‘F counts in favour of (or disfavours) A’s Φing in circumstances C.’ This reveals the basic conceptual structure of reasons, and a core semantic commitment of reason claims – they purport to refer to relations.
So, for example, suppose that the fact that the baby is drowning is a reason for you to save her. F would be the fact that “the baby is drowning”; A is the agent; and Φing is the “action of saving”. There seems to be nothing queer about the relata F, A and Φ; the queerness is with the reason R, the favouring relation. Now consider this hypothetical reason: if the baby drowning would go against your desire, you have a reason for you to save her. Inserting this into the above formula for a reason, F would instead be “if the baby drowning would go against your desire”, while A and Φ would remain the same. The hypothetical reason does nothing to make R less queer. In both the categorical and hypothetical case there is nothing queer about the relata F, A and Φ; the queerness is with R. The upshot of the argument is that the queerness of all kinds of reasons—categorical, hypothetical, epistemic, etc.—stand or fall together, since all reasons have this structure R(F, A, Φ) in C.
One might think that hypothetical reasons are less queer than categorical reasons because there is a naturally acceptable story of how we can come to know hypothetical reasons, as opposed to this mysterious intuition posited for categorical reasons. (This is a route taken by evolutionary debunking arguments.) But notice that this is an epistemological queerness and not a metaphysical queerness. The queerness in question here is metaphysical.
I think the only way out of the queerness is to make the move that Olson does and reduce a reason to a descriptive notion. So the hypothetical reason above would be reduced to the descriptive fact that saving the drowning baby would satisfy your desire. Many will claim that in reducing reasons talk this way and stripping it of a favouring relation we have changed the topic. I think this is probably true, so instead of seeing it as a reduction it could be seen as an elimination.
What are the consequences of the elimination of reasons? Does that mean you have no epistemic reason to believe this elimination? To be consistent, the error theorist would have to say, yes, there is no reason to believe this theory in virtue of there being no reasons whatsoever; but, if you desire to believe the truth, and you see this as true, your desire will be satisfied if you believe there are no reasons.
In everyday parlance it seems natural to say we did such and such for some reason; but, once we step back, I don’t think we need to posit this queer favouring relation—called a ‘reason’—to explain our actions. This superfluous ‘reason’ would be something over and above our desires and the non-normative worldly facts. We don’t need a reason to explain our actions since we can explain our actions by pointing to our desires and our recognition of the worldly non-normative facts about how to satisfy those desires.
Irreducibly normative reasons realists are committed to a metaphysically distinct favouring relation that is supervenient on the worldly non-normative facts. An example may help clarify some ambiguities. Suppose someone explains, “I have a reason to go to the kitchen because I have a desire for food”. What is happening is that facts about my desires and facts about where food is located create this additional metaphysically distinct thing called a reason. Some may object to this analysis, for in ordinary parlance we say that the (perfectly natural) fact that the kitchen has food just is the reason for me to go to the kitchen; so no new metaphysically queer thing is created—we just have respectable natural facts. This latter way of thinking is what Olson calls the extensional fallacy: it is conflating the reason itself with the extension on which the reason supervenes. (The extension, in this case, is the non-normative facts about the kitchen having food and your desires.) This ordinary way of is-speaking (when we say a reason just is some natural facts) dupes us into thinking the reason is identical to the non-normative facts. But, irreducibly normative reasons, if they exist, would be supervenient on but not identical to non-normative facts. I take these irreducible reasons to be sui generis metaphysically, which is why it is open to queerness objections.
Bedke, Matthew. Might All Normativity Be Queer?
Olson, Jonas. Moral Error Theory History, Critique, Defence.
Olson, Jonas. Reaons and the New Nonnaturalism.