(Title redacted for anonymous peer review)

According to a standard picture, normative reasons do not extend beyond the boundaries of agency. If something isn’t an agent – if it can’t do rudimentary practical reasoning, at the very least – then there can’t be normative reasons for it to do one thing rather than another.

I reject the standard picture. In this paper, I argue that there are reasons for non-agents to behave in some ways and not others. If my kitchen is full of smoke, then there is a reason for my smoke detector to start alarming. If my detector doesn’t go off, then there is something wrong with it: it is failing to do what there is good reason for it to do.

This has important consequences for theories of normative reasons. For one thing, we should not analyse what it is to be a reason by appealing to distinctively agential properties or capacities, like the standards of good practical reasoning. Instead, I suggest, all one needs to get reason-ascriptions going is some kind of teleology. There’s a reason for my alarm to go off in the presence of smoke because that’s what smoke alarms do when they’re functioning well.  

(Title redacted for anonymous peer review)

Take a look at the following two popular and plausible principles. First principle: facts about what you ought to do are tightly bound-up with facts about what you have reason to do. For example, if you ought to watch Oppenheimer, then there must be more reason for you to watch Oppenheimer than for you not to watch it. Second principle: ought is ‘upwards monotonic’ – if it’s true that you ought to φ in some specific way, then you ought to φ. If you ought to wear red socks today, then you ought to wear socks today. This second principle is not only plausible on its face, but also follows from the widely-endorsed, standard quantificational semantics for ‘ought’.

I argue that these two independently plausible principles are, in fact, incompatible. Respecting the connections between reasons and ought requires giving up monotonicity, and so requires giving up the standard semantics for deontic modals. We need something new. I sketch a non-monotonic meaning for ‘ought’ which builds the ought/reason connections right into the semantics itself, but also explains why monotonicity looked so attractive in the first place.

Sometimes we use ‘reason’ as a count noun, to pick out facts in the world. For example: “the fact that you like Greta Gerwig is a reason for you to watch Barbie”. Other times, we use ‘reason’ as a mass noun, to pick out something we can have more or less of. For example: “there is not much reason for you to watch Oppenheimer” or “there is more reason for you to watch Barbie than Oppenheimer”.

It’s natural to think that the count sense of ‘reason’ comes first – to think that the meaning of a comparative like ‘there’s more reason to watch Barbie than Oppenheimer’ is comparing the weightiness of the particular reasons to watch one movie with the weightiness of the particular reasons to watch the other. I argue that this ‘count fundamentalist’ semantics is untenable: when we talk about what there’s reason to do, we’re not talking about the reasons that there are.

I use this result to motivate a semantics for mass and count ‘reason’ which inverts the standard priority of reasons over reason. The central notion on this semantics isn’t there being a reason for S to φ, it’s there being some amount of reason for S to φ. I use this semantics to build an argument against the Reasons First programme (the enduringly popular idea that count noun reasons are normatively fundamental).