3.1 Bootstrapping Objection 1: Conflicting Normativity

Let’s return to Philippa. She ought to get out of bed and do some work. Her desire to become a great philosopher is one of the things she wants the most. But she still also has the desire to stay in bed – it is a cold morning and her bed is warm. Her desire to stay in bed means (according to a simplistic version of our desire-based theory of practical normativity)Footnote ¹⁰ that she also ought to stay in bed. Here we have our apparently conflicting conclusions:

Both of those conditions are true, so Philippa both ought to get out of bed and ought not get out of bed.Footnote ¹¹

3.2 Bootstrapping Objection 2: Normativity from the Wrong Kind of Source

The second version of the bootstrapping objection is that oughts can be generated by mental states that should not be able to generate them. This is the kind of objection that Holton seems to refer to when he says:

And Matthew Smith refers to this version of the bootstrapping argument as the ‘wizardry argument’ (Smith, Reference Smith2016, p. 2261), since it is concerned with generating normativity out of thin air. As if by magic!

Some philosophers like Holton (and Kiesewetter) are worried about intentions, rather than desires. But as with the previous objection there are several different states that might give rise to this kind of worry. Not just intentions (which can be formed as a result of a weak will), but beliefs (which can be incorrect or unjustified), and – most importantly for this paper – desires (which can be akratic, unfounded, or otherwise criticisable).Footnote ¹² This indicates the kinds of theory that this objection targets: any theory of normativity that relies on any of these kinds of non-factive mental state can lead to bootstrapping objections.

We can also see this version of the bootstrapping problem in Philippa’s case above: the desires that she recognises as less important, as manifestations of her weak will, are still generating oughts.

Furthermore, this objection can also come in the form of what I call ‘moral bootstrapping’. This particular problem occurs when we have desires which make it the case that we ought to do immoral things. Let us look at some examples:

Such an immoral desire should not – according to the objection – make it the case that agents ought to do the immoral things: kick the puppy, steal the cake, and tell the joke. But a desire-based account of practical normativity does not give us an easy way to distinguish between the kinds of desires we would want to give rise to oughts and the kinds that we do not.Footnote ¹³

Objective theories of normativity can escape bootstrapping objections because they do not need to rely on possibly flawed aspects of human psychology, such as our problematic desires. So they do not need to worry about normativity coming from those sources, or from those sources causing the normativity to conflict.

Before moving on, I’ll briefly say something about how these two objections relate to one another. What makes the ‘conflicting normativity’ objection and the ‘normativity from the wrong kind of source’ objection both instances of bootstrapping? The main similarity is that they’re both instances of the ‘force’ of normativity supposedly arising in ways that it can’t. When a person pulls on their shoelaces, they don’t generate the right kind of force to pull themselves up, only to pull up their shoelaces. The kind of force they apply isn’t the appropriate kind for if they wanted their whole bodies to move upwards into the air. So too, as the objection goes, with these oughts. They supposedly aren’t generating the right kind of normative force in the right kind of way.

4.1 Solution to Objection 1

Firstly, I’ll explain how this approach gives the desire-theorist a response to Objection 1, by showing us that it is plausible to think of agents having oughts that pull them in different directions, to both do a thing and to not do the thing. This is because they are different oughts, which are relativized to either a particular desire or a particular set of desires. They could be described as such:

Here, ought(A) and ought(B) have different desires at their source, and so are different oughts.

This solves the confliction worry because there is nothing more unusual about being pulled in different directions by two different oughts than there is to say that agents have conflicting desires to begin with, because we understand the way in which the different oughts are limited. And given the broad understanding of desire that we began with it would be almost impossible not to have desires that conflict to some extent.

Philippa both wants to stay in bed, and wants things that can only be achieved by getting out of bed. It might be true, given the first desire, that she ought to stay in bed, but it is also true given a certain set of her standing desires that she ought to get out of bed. No conflict here.

Having oughts relativized to different sets of desires isn’t so counter to our everyday understanding of them. We might tell someone that legally speaking they ought to do one thing, but morally speaking they ought to do another. Or that someone ought to fulfil a promise when it comes to thinking about their friend, but that they ought to break the promise in order to do what’s best for themselves. Sometimes I play turn-based board games online, taking turns slowly over a number of days. If I have twenty different games underway at once, there are at least twenty different moves I ought to be making in order to (fulfil my desire to) win the different games. There are also things I ought to do to make a good dinner, to enjoy my evening, to relax, etc.

One mistake being made by the objectors here might be to think that everything we ought to do is something we ultimately all-things-considered ought to do.Footnote ¹⁵ That each of these oughts are telling us the one, single, important thing that we ought to be doing in light of all of the relevant considerations. But the oughts here are not trying to be all-things-considered oughts. They are just what one ought to do given a certain, limited picture.Footnote ¹⁶ Philippa ought(B) to stay in bed, given her desire to stay in bed. And if it were the case that the desire (or set of desires) in question were the only ones she had, then it would be the case that, all-things-considered, she ought to stay in bed. The reality is much more complicated, given the sheer number of desires that any person will have.

Such an account of oughts would not prevent the agent from acting. Some oughts will be normatively stronger than others, and there are a number of ways that proponents of desire-based normativity could spell this out. They might, for example, tell a story about how some desires will result in stronger normativity because they are stronger desires, or because they are more internally consistent, or that they align better with some greater cause. Most likely the fact that Philippa ought to get out of bed will be normatively weightier than the fact she ought to stay in bed. After all, when all things are considered, her desire to be a respectable philosopher is more important to her than her desire to stay in bed. The way the normativity is grounded in the desire is a strength of the view – not something that makes it susceptible to the bootstrapping objection.

What about the second framing of the problem, where the objector’s worry was more explicitly about having any kind of normative ought contradict what an agent overall ought to do? Again, I want to argue that this is unproblematic. Suppose we continue with the comparison with physical forces. Many different forces can be exerted on an object at once. I’m pulled down by gravity from the Earth and pulled up a very small amount by the gravity of the other celestial bodies above me. My hat is also very slightly pushing me down, and a slight breeze from the window is pushing me to the left, and my feet are propelling me forward. Overall, I might only be moving in one direction: forward, across the room, where I’m walking. We can talk about how the different forces combined together and the force propelling me forwards won out. This doesn’t mean that the other forces aren’t still there, even though some of them pull in the opposite direction to the one I’m moving in. Something similar is happening with normativity. There are still oughts that apply to an agent even when, overall, they ought to do something incompatible with them.

Finally, I would like to say something briefly to explain the different phenomenology agents can experience when they have normative conflicts. Sometimes, when we really want to do something that we feel like we shouldn’t (in some other sense) do, it can really feel like our desires are in conflict with some external forces, such as morality, the law, or something else about what’s best overall. It’s important that desire-based theories of normativity can account for that experience. But this is exactly the kind of thing that the theory is well-placed to explain. It makes sense that different kinds of normativity that issue from the different sets of desires will also feel different. And sometimes, when the source of the normativity is our desire to do what’s best overall, or best morally speaking, etc, then it makes sense why that might feel like it has a certain kind of external pull on us. And in some cases, why it might feel alienating.

So in the case of the first kind of bootstrapping objection, we have a clear response available to the subjective theories of practical rationality. I will now develop a similar response to the second version of the objection, where I will also show that desire-based views have a particularly good case to make, in virtue of their role in explaining normativity.

4.2 Solution to Objection 2

The second bootstrapping problem is also alleviated. This is because the oughts that come from different mental states (such as desires or intentions) have no more authority than those mental states gave them, or that we confer on them based on the mental states in question. Sometimes, in the cases of our strongest or more important desires, for example, this would be a lot of authority, and sometimes, in the case of whims and fleeting inclinations, it would be much less. In this section I’ll go over this response, and show a variety of ways in which desires can still be thought to have different levels of (e.g.,) social, moral, or personal importance. To see this in action, let us return to one of the examples:

Ought(B) is the more problematic of the two; it arises as a result of a desire that the agent would, overall, rather not listen to. The desire is fleeting, even if it feels particularly strong to the agent at the time. But because of the flaws in the agent’s desire here, the ought that it generates is just as weak, and comes with just the same problems. The normativity should be given no more importance, no more authority, than the desire itself. Sure, the desire-based account tells me, Philippa ought to stay in bed given a certain set of her desires. But she does not hold very much stock in those particular desires, and the authority that the normativity comes with is correspondingly weak. Philippa’s desire to be a respectable philosopher, on the other hand, is one that she holds with a much higher regard, and so the ought that it generates will be normatively stronger. Just as the desire itself comes with more normative oomph, so it makes sense that the oughts that it generates will be the same.

Similarly, when an agent has immoral desires, we should not expect the oughts that follow to be moral ones. When we understand them this way, it does not seem to be such a problem that these sources give rise to oughts – they are not gaining any extra authority beyond what comes with the desire already. Let us look back at a couple of moral cases:

We need not pretend that both (C) and (D) generate the same kind of normativity, just because they are both oughts.Footnote ¹⁷ They are very unlike each other, and this is exactly what we should expect from two such starkly different sets of desires.

The first desire is the kind that might qualify as a moral desire. What exactly a moral desire is, of course, is likely to depend on other facts about morality. They might turn out to be those desires which will lead to the agent flourishing, to the agent helping others to flourish, or to the agent respecting the rights of others, for example. Whichever view of moral desires turns to be right doesn’t matter for the desire-based view of normativity – as long as there’s nothing circular in the explanation of what makes them moral desires.

Beadie’s desire, on the other hand, is an immoral one. The basic normative structure is the same, but we should still expect the normativity that comes out of it to be very different to that which comes out of Brianna’s desire. Again, there are a range of possible explanations for what might make desires important that don’t necessarily boil down to them just being the desires agents ought to have. We can evaluate desires based on whether they’re compatible with an objective moral code, on social norms, or against other sets of external rules. Even thinking only in subjective terms, the agent can assess a desire in terms of how strongly they feel it, how often it recurs, how compatible it is with their other desires, the extent to which they identify with it, etc. The desire-account doesn’t absolve normativity of this kind of criticism, it just means that the oughts and reasons that come out of it are also liable to be assessed in the same way. Immoral desires will generate immoral oughts, irrational desires will generate irrational ones, etc.

Why should anyone think there are immoral oughts at all? Again, this should only seem counter-intuitive when we’re mistaking the ‘oughts’ in question as being oughts of a certain kind of normativity. Immoral desires don’t generate moral oughts, and (depending on how you understand ‘overall’ normativity) they might not generate overall oughts either. But that’s not to say they don’t generate some form of ought.

Take the example of Beadie’s desire to hurt others. We might criticise this desire in a number of ways. We might say that it goes against objective moral rules, or that it is harmful to society. We might also say it’s harmful to Beadie herself, and conflicts with other things she values (recognition as a good person, self-respect, a less cruel world, etc.). When these criticisms appeal to something subjective like her other desires, then we can point to these other things she ought to bring about and persuade her that, overall, she ought not kick puppies, regardless of what she ought to do in order to fulfil that particular desire. When the criticisms appeal to something objective, like a social or moral norm, then we might not have a good way to reason with her, but we still have our own standards by which to denounce those oughts.Footnote ¹⁸ In this latter case we’re not criticising the fact that the desires lead to normativity, just the normativity itself.

One more possibility that I’d like to address is when seemingly moral desires might give rise to non-moral actions. Think, for example, about a parent who wants the best for their child.Footnote ¹⁹ It’s likely to be the case that some decidedly immoral actions would follow from this desire, such as actions that put the welfare of one’s own child ahead of many other people’s. In these cases, the criticism wouldn’t be about the immorality of that desire in particular. Instead, it might be that we criticise a lack of other moral desires, or the way the agent’s desires are balanced – such as the desire for equality or general well-being of others.

There might be other objections against the desire-based account of normativity, and I will not have time to provide a full defence of such a view in this paper. But what I have aimed to do is to demonstrate that for two kinds of bootstrapping objection, the desire account has good responses. Doing so, I hope, has emphasised some of the most appealing features of the account: its ability to provide a clear explanation for normativity, its structure, and its different guises.

Although I have given a response on behalf of subjective accounts, I said above that this response to the second bootstrapping objection also teaches us something about why a desire-based view of practical reasoning in particular is so appealing. This is because of the key role that desire plays in explaining normativity. As I discussed in Section 1, the desire-based accounts of normativity are appealing to many because of the way that the desire explains why the ought applies to that agent. Em’s desire for hot coffee explains why she ought to warm the jug, and if the desire goes away then the ought goes away too.

This strength of desire-based accounts shines through when we look at how they respond to bootstrapping objections. This explanatory role that desires play shows us how the normative force can be so variable, depending on the desire that gives rise to it. So although this response to bootstrapping objections works for a number of mental states, it is particularly convincing when we think of desire, and as an apt tool of defence for those who believe in desire-based theories of normativity (at least in part) because of its explanatory power.

5.1 Too Many Oughts

One potential objection to this desire-based account of oughts is that it means we have too many different oughts.Footnote ²⁰ After all, in my responses to the bootstrapping objection I talked about a different ought for each desire or set of desires. But my opponents might worry that this is not a plausible way to understand oughts at all. It does not seem like there are lots of different things going on when an agent considers what to do in light of their various desires. Rather, agents seem to be subject to one kind of normative phenomenon, and they try their best to work out where it is ultimately pointing.

The reason that this is not a problem for the desire-account is that although the oughts are different in one respect (in that they are each related to a different desire or set of desires), they are still the same kind of thing, that is, an ought that is related to a desire. The oughts themselves will have differences depending on the desires in question: some will be more socially, morally, or subjectively important, for example, and others will be less so. Some oughts we’ll value more, others we’ll value less. But they all share the same basic structure, and they are all still oughts. Although there are differences between them, it shouldn’t necessarily feel like they’re completely different normative phenomena.

One particular token of this worry comes from Kiesewetter when he discusses ‘subjective oughts’. He says,

This is not quite the same as the worry that there are too many oughts, but rather that one of the oughts is not doing the right kind of thing; it is not performing the same function that is necessary, that the others perform. Perhaps Kiesewetter’s worry here is that subjective oughts do not seem to be the same kind of thing as oughts that matter for deliberation. We might think here of some of the weaker kind in particular: the fact that Philippa ought to stay in bed because she wants to, or the fact that Beadie ought to kick puppies because she wants to.

But these are all still the kinds of oughts that do matter for deliberation, just as in the metaphor with the forces all of the different forces will be acting on the object even when it only moves in one direction. After all, the things we deliberate between are directions in which our different desires pull us. This is still the case when a desire pulls us but not very much, or a desire is very easily overruled, or immoral, or something we (upon reflection) would want to reject. And that is not a problem for the account, that these weaker oughts are still what Kiesewetter calls ‘deliberative oughts’. After all, not all desires will give rise to motivation in proportion to their strength or importance, not unless the strength of a desire is the amount that it motivates you, so it will not be the case that the only things agents ought to do are the things they end up doing.

5.2 Reasons and Oughts

The second objection that I want to defend against is the worry that my account is an account of reasons, but not of oughts. As I mentioned at the beginning, this paper defends a desire-based account of practical normativity, and this includes both what agents have reason to do and what they ought to do. But some of my opponents might think I have only justified the former, and not the latter, because the former is normatively weaker.Footnote ²¹ An agent having a reason to perform an action doesn’t seem as strong as if they ought to do it.

But I want to deny that this is the case. Once again, I do not think that everything we ought to do is something that we overall, all-things-considered, ought to do. And to say so is not to confuse oughts with reasons. Firstly, because this still tracks the way that we talk about what we ought to do. It is perfectly common to talk about what we ought to do in one sense while thinking about the way that this competes with other things that we ought to do. I can think about the fact that I really ought to keep a promise to Sophie (given that I desire to keep my promises, particularly to her!), but that in another sense I really ought to break it, because of stronger competing desires, and a stronger competitor for something I ought to do. I take it that such thoughts are not just the product of my spending too much time with other philosophers.

Secondly, I concede that we often talk about what we ought to do and mean what we overall ought to do. But this does not mean that any ought is an overall ought. It just means that when we use our language we often use a broader term as a shorthand for a more specific term. This should be evident enough from our ordinary talk of ‘reasons’ – just because our everyday talk tends to be about just ‘reasons’ generally, that does not mean that there are not a range of different specific phenomena that we pick up with that talk, such as epistemic reasons, normative reasons, motivating reasons, etc.

Thirdly, the variety of oughts that I discuss are similar enough, structurally, to phenomena that do seem to be oughts, that it makes sense to understand them as different types of the same phenomenon. I mentioned this briefly in response to the ‘too many oughts’ worry above, but it’s relevant here, too. Take one of the least controversial examples of an ought that I’ve argued for: the fact that Brianna (if she wants to help her community) ought to attend the protest. According to the views I’ve been defending, this ought applies to her because of her desire to help her community, and if that desire is taken away then the ought would only still apply to her if the action would fulfil other desires that she has, such as the desire to promote equality or for the improved well-being of her friends who are members of oppressed groups. Without any desires, it would be left unsupported. It’ll be a subjectively strong ought if Brianna cares a lot about her community, and endorses this about herself, because she likes being the kind of person who has these desires, and so the kind of person who ends up going to protests. There will also be plenty of other senses in which this ought is objectively important – it’s the kind of moral ought (and the kinds of moral desires) that society generally values, and ones that likely match up to an external moral standard.

Structurally, this ought seems to match the myriad of other oughts that my account argues for. Beadie ought to (given her desire to hurt others) kick the puppies, and that ‘ought’ would be left unsupported if that desire of hers went away. Although I’ve argued that there are a variety of ways in which this ought is likely to have different importance or strength to it, those features are explained in the same way as they’re explained in the case of Brianna’s ought. Where Brianna’s desire and ought matched an external moral and social standard, Beadie’s contravene it. And perhaps, if we’re lucky, Beadie’s desires also might be subjectively weaker. None of the differences between these two types of ought seem to be differences in their structure.

Fourthly, I’ll say something about what the difference between oughts and reasons still is, on this account. I have argued that there are a lot of oughts – oughts that can match any desire or set of desires that an agent has. But there’s still a sense in which an ought is the thing to do, even though it’s the thing to do relativized to one of those desire sets. We can suppose for each of these desire sets, there’s one thing that will best fulfil them, and that’s the thing that they ought to do. But there will still be even more things that they have reason to do. Brianna desires to help her community, and so she ought to attend the protest, because we can suppose that is the thing she can do to best help her community at the moment. But that desire also gives her plenty of (outweighed) reasons to do other things: to pick up litter, to organise a film screening, to cover up an ugly wall with some beautiful graffiti. There are plenty of reasons that each desire favours, but perhaps only one thing that a desire means we ought to do.

Finally, if the above has not convinced my opponent, then I will happily concede that, as far as those opponents are concerned, this paper is about reasons rather than about oughts. It is just a matter of terminology, after all, and sometimes different terminology is bound to be part of an explanation between apparent disagreement in philosophical accounts. I have been using ‘ought’ to mean a similar normative phenomenon as a reason. An ought being a conclusion of a set of reasons, a conclusion given certain limitations (certain desires or sets of desires).Footnote ²² But if others do not want to use the terms that way then they do not have to.

Before I move on I will briefly give some time to a similar worry. What if my opponent is not just worried about the terminology of oughts when some of them have different strengths of normativity to others, but they are worried about whether there is normativity at all with these oughts. Might it be the case that normativity is about what we overall, all-things-considered, ought to do? And that a reason to do anything else is not really a normative reason?

But I am confident that there is something meaningful, real, and shared between all the kinds of normative oughts and reasons that agents have under this account, and that such a thing is normativity. The connection with the agent’s desires gives these oughts and reasons exactly the kind of oomph that philosophers seem to mean when they try to explain what exactly normativity is. It explains why a reason or an ought counts in favour of an action. It does not necessarily correspond with what the agent actually does, but there is an important way in which it corresponds with what an agent might do, something that might factor into their deliberation, that might lead to motivation.

I find it helpful to think of Manne’s work in favour of reasons internalism when I think about what normativity is. She asks us to think about a number of ways in which we interact with each other:

Each of these behaviours stands a chance of success by either appealing to (or occasionally by trying to create) desires in the person we are interacting with. In contrast to these kinds of behaviours, when we act in ways that do not even try to appeal to anything related to a person’s desires (perhaps by restraining them, or casting them out) then we seem to be engaging with them in a different and non-normative way. The difference between the ways we interact when appealing to internal reasons (those related to the agents’ desires) and external ones (reasons without that necessary connection) says a lot to me about what normativity is. And it is definitely the kind of thing that can still exist at a variety of levels of strength, regardless of judgments about what an agent overall ought to do.

5.3. Proportionalism

The final worry that I will address is that the desire-account’s response to bootstrapping objections means we need to accept a kind of proportionalism. I will begin, of course, by explaining what this actually is.

Proportionalism in this sense is the view that the strength of an agent’s reasons correspond to the strength of their desires. If I have a strong reason to attend the protest, for example, then this is because I have a strong desire that would be fulfilled by that attendance. And it goes the other way too: if I have a strong desire to drink the coffee, then this will give me a strong reason to warm up the jug.Footnote ²³

I find proportionalism to be very persuasive – particularly given what I have described as a main strength of desire-based views: the fact that the desire explains the practical normativity, explains how it works and why it has its force. Proportionalism would mean that the desires are able to explain even more of the normativity, to reduce any metaphysical weirdness even further.

But the account of desire-based normativity that I have defended does not commit me to proportionalism – one could take what is appealing about the way the desire-based views respond to bootstrapping objections without taking the proportionalism as well.

What is important for desire-based views of practical normativity is that there must be some desire present in the agent for the normativity to get off the ground at all. And that there is then some correspondence between that desire and that normativity, that the weak-willed or immoral desires do very different normative jobs to, for example, Brianna’s desire to improve society. But the way that those desires affect the differences in normativity does not have to be through an explanation of the strength of the desire.

One example that might appeal to people distrustful of proportionalism is through some kind of explanation of the objective nature of those desires. Suppose, for example, we determined what kinds of desires counted as ‘moral’ desires. It might be an agent’s desire to help her fellow person, her desire to maximise their well-being, or her desire to exemplify particular virtues.Footnote ²⁴ Whichever those are, perhaps, it might be the case that their normativity is more important than others. Certainly in the eyes of the moral law. Such a person would have a kind of hybrid view: one that needs a desire to get the normativity off the ground, but that explains the differences in the resulting normativity in terms of objective facts about those desires.

It is possible to think there is something more to the source of normativity than our desires, something extra that adds to the normative force. Even for that view, the desire-account still has an answer.