1 Introduction

As moral agents, we appear to have the ability to exercise substantial control over our own normative situations by the use of what can be called normative powers. At first blush, the class of normative powers can be described as the class of abilities agents have to effect changes to their reasons, permissions and rights by the means of communicative actions whose function is to effect just those changes. Take the promissory power. In being able to give valid, binding promises, we are apparently afforded the opportunity to shape our normative situation in substantial ways—we are able to create obligations to perform the promised action that normatively constrain our choice independently from any prior independent reasons for performing it. Like other normative powers such as the power to consent or the power to transfer property, the power to undertake promissory bonds plays a prominent role in our moral lives.

Though it is certainly not uncontroversial, there are some compelling reasons to accept that we genuinely have such powers.Footnote 1 But, assuming such a view, why do we have the abilities to influence our normative situation in a select number of specific ways, yet not in others? Why, for example, can we create obligations for ourselves through promises, but not likewise create permissions for ourselves through an equivalent act? A natural answer to this question consists in pointing to specific values or human interests that are being served by us having certain normative powers, but not others. Philosophers such as David Owens and Joseph Raz have attempted to show precisely how the availability of a promissory power is good for us, with the explicit intent of thereby contributing to an explanation and vindication of promissory normativity.Footnote 2

While the move of offering such a value-based grounding story in conjunction with a normative power view of promising has intuitive appeal, it gives rise to a serious problem. Tracing the bindingness of promissory normativity back to value in this way invites a charge of wishful thinking. Establishing that it would be good for us if we could cure all cancer clearly does not seem to contribute any justification to the belief that we can, in fact, cure all cancer (Heuer, 2012b: 845). Why should it be any different with regards to the power to change the normative situation by giving valid promises, for example?

While this objection has been raised in the literature before, it has so far not benefited from the kind of in-depth discussion it deserves.Footnote 3 In this paper, I lay out the challenge in detail, and then offer a defence of value-based accounts of normative powers against it, focusing on the case of promising as a paradigmatic case. As I shall show, there are in fact crucial differences between the power to cure cancer and the power to create obligations through promises, a proper understanding of which makes clear why these kinds of worries need not affect a value-based grounding story for the latter.

My defence against the objection comes in two main steps. First, I make a distinction between what I call the normative and material components of our power to give promises. I show that the claim at the heart of the project of supplementing a normative power view with a value-based grounding story is not one about the material component of our promissory power, but rather its normative component. And with regards to this latter, I hold, defenders of a normative power view can simply stand their ground and deny the problem. If what we are talking about are purely normative principles at a high level of abstraction (as I will show that we are, in the case of the project at issue), and as long as we are not principally opposed to value-based grounding in the realm of the normative, then the kind of wishful thinking involved in such value-based accounts of normative powers turns out to be benign. I support this claim by drawing on other cases of value-based explanations, in particular ones related to rights of autonomy. I conclude by sketching the prospects of accounts that attempt to give a deeper explanation of why forms of argument that would usually smack of wishful thinking have their place in the domain of the normative.

2 A normative power view of promissory normativity

Let me begin by laying out, in the broadest strokes, why I believe pursuing a value-based grounding story of normative powers to be worthwhile. I will do so focusing on the example of the power to create obligations by giving promises, but the basic structural points are easily transposed to other normative powers, such as legitimate orders or consent.Footnote 4 While I will not be able to spell out all the arguments in favour of such a view, a brief overview over the dialectic will help us understand what is at issue and how exactly the wishful thinking objection comes into play.

First, I will clarify what exactly I take a genuine normative power view of promising (or NPP, as I shall call it from now on) to amount to. At its core lies a straightforward and intuitive account of what transpires when a valid promise is given. In David Owens’ words, promising “wears its function on its sleeve” (Owens, 2012: 144): by promising, we voluntarily undertake an obligation to perform the promised act, making it the case that we have an obligation to do as promised.Footnote 5 In giving the promise, we are communicating that we intend to undertake this obligation, one that is furthermore owed specifically to the promisee.Footnote 6 If, for no good reason, we do not keep our promise, then what we are doing is wrong simply because we are not acting as we had promised. On this way of viewing things, promissory obligation is a sui generis, non-reducible form of obligation.Footnote 7

The fact that it takes promissory obligation to be non-reductive sets NPP apart from opposing reductive views. On these theories, it is in some sense also true that we have a power to create obligations by giving valid promises. As theories of the normativity of promising, they do not wish to deny that performing a promissory speech act typically results in the incurring of an obligation, or at least some reason to act as promised. As reductive theories, however, they do deny that the mere fact that a valid promise was given can ever be the full explanation for the creation of an obligation to keep a promise. Instead, they claim that the wrong-base for an individual breach of promise to φ includes further facts beside the fact that φing was promised. These might, for example, be facts about the expectations raised in the promisee (Scanlon, 1998), or more general truths pertaining to the morality of free-riding on useful conventions (Rawls, 1971). In this way, reductive views hold that the power to create promises turns out to not be radically different from the power to incur obligations to apologize by wilfully hurting others (Chang, 2020: 276), or from the power to incur an obligation to buy a ticket by entering public transport. In these latter cases, we are able to effect a change to our normative situation, but only in virtue of some external fact that has independent normative relevance.

In contrast, NPP is committed to the idea that the full explanation for any given promissory obligation to φ is that a valid promise to φ has been given, and nothing beyond that. NPP enjoys considerable prima-facie appeal, which is reflected in the range of philosophical defenders it has had over the years.Footnote 8 Besides several other advantages, NPP most crucially is afforded with an extensionally adequate account of which promises bind “for free”. Since all that is necessary for a promise to be binding on NPP is to be given under conditions of validity, the view automatically returns the intuitively correct result that all valid promises give rise to obligations. This is a crucial advantage of NPP over its reductive competitors, which face a notoriously difficult task in getting their theories to square up with our intuitions about the extension of promissory normativity.Footnote 9

Notwithstanding its advantages, NPP has fallen out of favour with many philosophers. Though a full argument to this end is beyond the scope of this paper, I believe that this ultimately has to do with its status as a non-reductive theory. Unlike reductive theories, which, in specifying a given reduction base for promissory obligation, automatically provide a deeper explanation of why and how promises bind, non-reductive theories like NPP leave open the question of the grounds of promissory normativity. To put it squarely, NPP is the statement that promises give rise to obligations, not why they do so (Watson, 2009:156). This lack of a grounding explanation can be seen as a problem for several reasons.

First, the promissory power may appear in need of some sort of external vindication. A grounding story can help make the case against the claim that we are just substantially mistaken about the normativity of promises. Think of the problem this way: How can we be sure that the promissory power is a genuine normative power that successfully creates obligation, but other performative speech acts that at the surface appear to exhibit much the same structure are not?Footnote 10 Take a challenge to a duel. Until relatively recently, many would have likely thought that such a challenge, issued under the right circumstances, can create an obligation for the recipient to answer in an appropriate fashion, requiring him or her to engage in potentially lethal sword- or gunplay. Without a story about how promissory obligation is grounded (i.e. one that entails that it is, unlike the power to challenge for a duel, well-grounded), we seem to have little more to go by here than our bare surface intuitions if we want to set these two cases of putative powers apart. And these intuitions, given their historical mutability, as displayed by the duel example, may be found to provide insufficient support.

Secondly, apart from questions of vindication, it may seem that a lack of a grounding story leaves promissory obligation understood along the lines of NPP irredeemably mysterious.Footnote 11 How exactly to flesh out the notion of mystery is a substantial, and, as it turns out, not particularly easy question.Footnote 12 For now, let me highlight one feature of promissory obligation that centrally contributes to the sense of mystery. This feature is the, to use the words of Joseph Raz, “content-independent” nature of promissory obligation, and the distinctive independence from value that results from it (Raz, 1972: Volume 95).

This feature becomes most clearly apparent when we focus on the fact that promises seem able to obligate even in situations where no good would come from their fulfilment. Take the following example, which goes back to Hume (1738/1978: 480f). Imagine that I ask a friend to lend me a certain amount of money, and that I promise to pay it back in half a year’s time. But by the time the repayment date has come around, my friend has unexpectedly come into a lot of money, having won the lottery in the meantime. He is no longer in need of the money, nor is it likely that he would put it to any good use. Nonetheless, my promise seems to be binding still. My having promised gives me a reason to pay back the amount that I borrowed, and this reason is independent of the desirability of my doing so. At least in some cases, promises thus appear to be able to render acts obligatory even when performing these acts would not be desirable in any way. In this way, the justification for kee** one’s promises is “content-independent”.

There will of course be many cases where having promised to φ will cause φing to become a more valuable option when compared to not φing. A promise will, for example, generally lead the promisee to rely on the performance of the promised act, meaning that non-fulfilment of the promise would stand to negatively affect her by frustrating her expectations. But this is not a necessary result of giving a promise, and of course not part of the explanation of promissory normativity on NPP.

The fact that promises are independent from value in the way just sketched leads to a worry: How is the property of being promised able to make such a substantial difference to the normative status of an act, while not having any impact of this act’s value, or impacting a value in some other way? This difficulty, of course, plays into the worries about the vindication of our promissory power laid out above. It can also go a long way to explain worries about bootstrap**, which for a long time have played an important role in the criticism of non-reductive theories of promissory normativity.Footnote 13 Finally, it also appears that promissory obligation, when understood along the lines of NPP, ends up being incompatible with any value-based theory of the normative.The central postulate of such a theory is a principle that we, loosely following Joseph Raz, may call the Value-Reason Nexus (Raz 2001: 5).

(VRN) Whenever we have a reason to do something, this is ultimately explicable in terms of value.

Notice that although VRN clearly follows from any consequentialist theory in normative ethics, it is not a principle that should be attractive only attractive to consequentialists.Footnote 14 The principle does not state that the normative status of an act can be determined exclusively by considering whether or not it would maximize or even just promote a value. It is perfectly consistent, for example, with the claim that one ought to refrain from killing one person, even if doing so would allow one to prevent some other person from killing five people in the future. All that VRN entails is that, if this were true, our reasons to refrain from killing the one must be ultimately explicable in terms of value, e.g. because refraining would respect the value of inviolability of life, protect the valuable individual life of the one, or something similar along these lines. Though it is of course not uncontroversial, value-based accounts of the normative endorsing something like VRN have a growing share of defenders.Footnote 15 If, on NPP, promissory obligation were to violate this principle, this would considerably reduce the attractiveness of the view.

As noted, these problems of NPP ultimately have to do with its status as a non-reductive theory. Unlike reductive theories, the explanation of promissory normativity inherent in NPP entails no grounding story that establishes the desired link between value and obligation. But that no grounding story is inherent in the proposal does not mean that no grounding story is compatible with it. This situation makes it very attractive to supplement the basic theoretical framework of NPP with a value-based grounding story. I will now briefly outline what appears to me the best way to go about such a project.

3 A two-level grounding story for normative powers

How, then, would a value-based grounding story have to look like if it were to retain the central tenets of NPP? Most crucially, it could not be an account providing value-based grounds of the individual obligations created through the promissory power. For these obligations are, NPP holds, grounded in the fact that a valid promise has been given, a fact that itself has no necessary evaluative relevance. The idea then has to be, instead, to provide an explanation of why a given normative power has a genuine impact on the normative situation by making recourse to a special value that having this power has, or differently put, a special human interest that is served by its existence.Footnote 16 And if this value is to explain the normative power as a capacity to change the normative situation, it must be irretrievably connected with the conception of the normative power as just that. That is, the value served by a given normative power must lie in the very obtaining of some normative truth or other.Footnote 17 To apply this to the case of promising: if we understand promising as the power to voluntarily bind ourselves to a certain course of action by undertaking a (directed) obligation to the promisee, then the value that underlies this power must be the value of having the ability to bring about precisely this type of obligation. Differently put, it must be the case that there is some good for us, as agents, in being afforded a kind of normative control—the ability to obligate ourselves in the way NPP holds we can through promising.Footnote 18

It is important to again stress that, on the proposed account, an agent’s reason to keep a validly given promise is thus not directly derived from the value of having normative control. Nor is it to be traced back to reasons to uphold, establish, or do one’s fair share in maintaining a convention that allows us to exert such a control.Footnote 19 Instead, the explanation of promissory obligation inherent in the proposal is better understood as having a two-level-structure (Raz, 1977: 219). A way to make this structure clearer is by conceptualising the existence of a normative power through a focus on principles and their grounds. Think of it as there being a general moral norm requiring us to perform those actions which we have sincerely and validly promised to perform. Call this the promissory principle.

(PP): If you make a valid promise to φ, then you have an obligation (to the promisee) to φ because you have made a valid promise to φ.Footnote 20

On a first level, the individual obligations to keep one’s promises are explicable in terms of the promissory principle. If I break a promise, what I do is wrong because I have made a valid promise and am subject to a norm that holds that promises must be kept—that is the full wrong-making explanation. This bald answer is a commitment of NPP, which takes promissory obligation as non-reducible.

However, we may go beyond the mere appeal to this principle by providing, on a second level, a story on why (PP) obtains. The principle is fully grounded, so the idea goes, in the value of a certain kind of normative control. This control is provided to each agent that is subject to the principle by the very fact that the principle obtains. What is valuable then is a certain kind of dependence—the dependence of the normative status of an action as obligatory on the fact of us having performed a communicative action with the intent to effect a certain normative change. The fact that this dependence obtains, i.e. the fact that the promissory principle holds, is part and parcel of the agent having the kind of normative control in question—a kind of control that, we may hold, is independently valuable.

This two-level structure allows us to give a response to the aforementioned worries while retaining NPP’s commitment to promissory obligation as a sui generis obligation that is created by every valid act of promise-giving. Though the individual reasons to keep promises are independent from the value of the promised act, we can nonetheless offer a justification for their obtaining that ultimately traces back to questions of value.

4 Wishful thinking and an important distinction

So far, I have laid out the central structural features of a value-based grounding project for normative power-accounts such as NPP, as well as sketching the main motivations for finding such a view attractive. When kee** these features in mind, it is easy to see how worries about wishful thinking could be raised in response to such a project.

What I have laid out is a view which holds that, ultimately, we have the power to give valid promises because having such a power would be good for us. Here, one is easily tempted to see bad reasoning. In general, there seems something fishy with explanations of the form [it would be good if p, therefore p]. This seems true whether we fill in the formula with facts about other abilities (it would be good if I could cure cancer, therefore I can cure cancer) or more basic descriptive facts (it would be good if I was tall, therefore I am tall). Why should things be so radically different when the p at question is, for example, our having the power to give valid promises?

I believe that things are in fact radically different when it comes to value-based grounding stories of normative powers. To explain why this is so, I will begin by making an important distinction. When we talk about the normative power of promising, for example, being grounded in the value of its obtaining, we must make sure to keep apart two distinct parts to our having that power. On the one hand, certain normative facts must obtain for promises to be normatively binding. Call this the normative component of our promissory power. On the other hand, further non-normative conditions also must be met for us to be actually able to successfully exercise any promissory power. We can call these the material component of our promissory power.

To clarify the distinction, it is helpful to go back to the interpretation of NPP in terms of the truth of the non-reducible principle (PP). The normative component of the promissory power simply is the truth of (PP)—as long as it obtains, the normative preconditions for our successfully entering into promissory bonds are in place. However, this is not sufficient for our having a normative power. We are only able to enter such bonds as specified in the consequent of the conditional if we furthermore have the ability to render true the antecedent. Whether or not we are able to do so depends on certain descriptive matters of fact. To be precise, we have to be able to perform a communicative act that is understood by the promisee as being of the kind characteristic of promising, i.e. as expressing the intention to undertake a directed obligation to perform the promised act. Only if the promisee is able to understand us as doing so and, thus, able to properly accept the promise, will our promise be valid and the antecedent of (PP) triggered.

When focusing just on our ability to undertake obligations using the specific locution “I promise”, it becomes clear that a whole background of conventional facts must be in place for us to have it. There is, of course, nothing magical about the specific words “I promise”, and the fact that this particular locution has the normative impact that it does is only intelligible against the background of our actual, historically developed, promissory convention. We may say, borrowing the words of Erin Taylor (2013), that our ability to enter promissory bonds is “conventionally mediated”, even if it is not ultimately grounded in convention.

It may be that the required understanding of the promisor as expressing the intention to undertake an obligation to the promisee to perform the promised act is only possible against the background of a promissory convention that fixes its distinctive features to a specific communicative marker. Though this is possible, it can equally well be put into question. Locke (1689/1960: 277), who defends an account of promises as a normative power that is in the spirit of NPP, directly challenges the idea that any type of conventionality is necessary for promises to be binding. Whether or not it can only be given against the background of a fully-fledged convention, some sort of shared understanding of a given communicative act as an exercise of a normative power, and a basic willingness to accept it as such, are thus in any case required for an exercise of a normative power to take effect. This is what I call the material component of the normative power.

With the distinction between the material and normative components of normative powers at hand, let us return to the problem of wishful thinking. Is an explanation of the form [it would be good if p, therefore p] problematic when applied to the two components individually? As becomes apparent quickly, it is indeed problematic when applied to the material component. The fact that it would be good if a given group of agents shared some sort of understanding of a given communicative act as an exercise of a normative power does not, at least not by itself, serve to show that the group has this sort of understanding. As what we are interested in here are straightforwardly descriptive questions, [it would be good if p, therefore p] is clearly wishful thinking of an objectionable variety.Footnote 21

Crucially, however, a supplemented NPP-account simply need not be able to prove that the material component of a given normative power obtains. After all, the abovementioned considerations requiring NPP to provide a value-based grounding story are all squarely on the normative side. The challenge is thus best conceived as one that demands a story of how a non-reductive form of promissory normativity could be grounded in value, given that our promissory exchanges take the form that they currently do, shared understanding and all. As far as the defender of NPP is concerned, the historical circumstances that led to the development of the promissory practice enabling our promissory power could have been wholly contingent, a lucky cosmic fluke. What matters is that for us, as beings who find themselves in the circumstances where the material component for a promissory power is given, it is good to be subject to a certain conditional norm requiring the performance of promised actions.

The key question that needs to be answered thus is whether explanations of the form [it would be good if p, therefore p] are problematic when focusing only on the normative component of the promissory power, i.e. the obtaining of the abstract promissory principle. I am now going to argue that they need not be, and that there are even some good reasons for believing they never are when brought to bear of normative facts at a sufficient level of abstraction.

5 Wishful thinking and pure normative facts

When considering the problematicity of wishful thinking for the normative component of the promissory power, it is important to again highlight the nature of the claims made by NPP. What is supposed to be grounded in the value of its own obtaining is only the promissory principle, a normative principle at a high level of abstraction:

(PP): If you make a valid promise to φ, then you have an obligation (to the promisee) to φ because you have made a valid promise to φ.

The explanation is not supposed to extend to individual instances of promissory obligation, i.e. facts of the kind [This person is obligated to keep his promise to show up punctually to his friend’s party on the weekend.] As explained above, these individual obligations are justified simply by reference to the truth of (PP) and the fact that a valid promise has been made. The remaining question then is whether it can ever be good reasoning to assume the truth of such high-level abstract principles because their obtaining would be in some way good. In what follows, I will argue that, at least if we commit to value-based explanations in the domain of the normative, we simply must assume as much if we are to capture moral principles that are near and dear to us. Most crucially, these types of consideration are necessary for a full value-based explanation of our rights to autonomy over our own bodies and personal property.

Let me set out with an assumption that is rightly widely shared: it is wrong to use the personal property of others or their body without their consent. Theft and (especially) rape are almost universally recognised as grave forms of wronging. To put the same idea differently, we all clearly seem to be bearers of important moral claim-rights protecting our personal property and bodily integrity from any unwanted outside interference. But why do we enjoy these special protections? There is a venerable line of argument, reaching back to John Stuart Mill, which attempts to explain these rights of autonomy through a basic interest all humans share: an interest in being able to make their own decisions in the pursuit of their lives. The value is one to be found, as Mill himself says, in each of our own’s “mode of laying out [his or her] existence” (Mill, 1859/1974: 133). Joel Feinberg expresses a similar thought when he calls our most fundamental right to autonomy the right to “decide how one is to live one’s life” (Feinberg, 1986: 54).

Crucially, appeals to this value seem fit to ground reasons not to violate autonomy-rights not only through the disvalue that violations can have for the pursuit of individual plans, but at least also through the value that having the right itself has for an agent.Footnote 22 We may plausibly think that the value of having my own life to lead can only be fully realised if there is a domain over which I not only have de facto control, i.e. the ability to actually decide without outside interference, but also an entitlement to this control, i.e. rights to this domain such that, if anyone were to attempt to forcefully take control of something in it, they would thereby be wronging me. This being true appears to me a strong candidate for a necessary condition that needs to be met if my life is to be called truly “mine to lead” in a meaningful way.

To illustrate this, compare two individuals, A and B, who both live a subsistence lifestyle on remote farms. While A is the owner of the farm she is working, B is only working “her” farm with the permission of a third-party owner C, who could revoke this permission at any time, leaving B in transgression of C’s property rights. Even if there is no way for C to enforce their claim against B and actually get her kicked off the farm if they decide to revoke permission, I take it to be hard to deny that, in this situation, A enjoys a level of autonomous self-determination that B lacks by the very fact that A enjoys a more robust entitlement to their means of subsistence than B.

In a very similar context, Arthur Ripstein (2006) helpfully speaks of a “sphere of individual sovereignty”, the very existence of which is crucial to my existence as an autonomous agent. The existence of this sphere of sovereignty, however, is of course nothing but the truth of some set of normative principles that forbid the unauthorised use of my body or personal property by third parties. On this line of thinking, then, these principles would be grounded in the value of their obtaining, just as the promissory principle on NPP.

More importantly, rights of autonomy are not only compatible with a value-based grounding story that proceeds from the value that their obtaining would have for individuals, they also seem to require it if they are to extend to all of the cases we intuitively take them so. Take cases of what we can call “harmless wrongdoing”. In Ripstein’s favourite case, a burglar with a meticulous sense of cleanliness breaks into your house and, taking all precautions to not leave any trace or damage, takes a nap in your bed while you are not there (Ripstein, 2006: 218). He thereby intrudes into your sphere of sovereignty but does not do you any harm (assume that you are certain to never find out). A more serious example of this type of violation is sexual trespass on a temporarily comatose person. Being groped by a stranger when comatose may not lead to any uncomfortable feelings, injury, or negative long-term effects, if it is done carefully and in secret. Nonetheless, we intuitively judge this to be a very serious violation of a moral obligation.

Now, if our rights of autonomy were only explained by the disvalue that violations have for us, i.e. through the harm that befalls us when others violate them, the intuitive wrongness of these types of action would be left crucially unexplained. There seems no harm to the victim involved in either of these cases—the cases are set up in a way that ensures they are not at any risk of being made worse-off by the action. Nonetheless, the unnoticed trespass and the sexual touching of the comatose seem wrong precisely as violations of autonomy-rights.

At this point, one might challenge the intuition that there is nothing bad in being subjected to actions that constitute violation of autonomy rights, such as the secret nap or the touching of the comatose person. In some sense, it may be objected, there clearly is a harm involved in these cases. After all, we feel sorry for those who are victims of such actions, and hope that these things never happen to us personally. I admit that the intuitions are perhaps not univocal here, and there is a case for the badness of these actions to be made. However, the kind of badness inherent in actions that constitute violation of autonomy rights, such as the secret nap or the touching of the comatose, can only be a type of deontic badness, which is not a suitable candidate to ground a prohibition against performing these actions. Here is an argument to bring out this point.

Consider a situation in which an agent expects to be temporarily comatose for a moderate period in the future, and has, in knowledge of this, consented to being sexually touched by a certain person while comatose (their partner, for example). In this case, is there any intuitive disvalue if the person is actually touched, compared to a counterfactual case in which, though permitted, the partner abstains? I submit that there is not—while consent is in place, it is very clearly the case that the action performed does not make the comatose person worse off in any way. But this is a noteworthy result, since it stands in contrast to most other cases of consent. If I have consented to your punching me, you may no longer wrong me if you plant your fist in my face. Nonetheless, it seems very clear that something bad is still happening to me if you make use of your newfound privilege. Similarly, if I consent to your taking my car over the weekend, I am worse off if you actually do so, insofar as that I am bereft of the opportunity to take it myself, in addition to any wear and tear that may occur. In general, it is an important feature of consent that it does not, at least not by itself, change the value of the consented-to action for the consenting party.Footnote 23 Instead, the effect that consent has is squarely on the deontic side of things, changing the status of an action only with regards to its permissibility.Footnote 24

This explains why in many cases of consent, the consented-to action will exhibit what we may call a remainder of badness, even though performance of it is fully permissible. Not so, however, in the cases of harmless wrongdoing under consideration. That the situation is radically different in these cases very strongly suggests that the value in question must be a type of deontic value from the start. Only if it was deontic value could we explain how the mere presence of consent, as something that by its nature only affects the deontic side of things, could make the difference that it does to our evaluation of the (dis-)value of the act in question. And if the value in question is deontic, it cannot fulfil the intended function of grounding our rights protecting us from these type of actions in the way that the badness of certain harmful actions usually grounds prohibitions against performing them. The wrongness of an action cannot depend on its badness if that badness itself depends on the wrongness of the action. If we want to provide a value-based grounding story for these rights, we therefore have no choice but to turn to one that proceeds from the value that their obtaining would have for individuals, instead of a more traditional one that proceeds from the disvalue their violation would have.

If what I have argued above is correct, it seems to me we have a strong argument against the claim that explanations of the form [it would be good if p, therefore p] are always problematic in the normative. There is a class of normative principles, those requiring respect for personal autonomy, where a venerable line of argument, proceeding from the value of leading one’s own life, can be very plausibly interpreted as involving an explanation of just this kind. Furthermore, an explanation of this type appears to be required to capture the intuitively correct extension of these principles through a value-based grounding story of them.

Of course, the considerations just presented may equally well be read as a modus tollens against value-based explanations in general. Friends of the wishful thinking objection might respond as follows: If we can only square a value-based explanation of autonomy rights with our intuitions about which kinds of actions represent violations of these rights by invoking a type of explanation that has the paradigmatic form of wishful thinking, then so much the worse for value-based explanations! I think this is in fact a reasonable line to take.Footnote 25 It is important to keep in view the overall dialectic, however. The problems that made it necessary to supplement NPP with a value-based grounding story in the first place presupposed a close explanatory connection between value and reasons along the lines of VRN. Should the argument from autonomy-rights just presented lead us to abandon the idea that showing a connection to value is necessary for justifying a given moral principle, then it is no longer clear why we should not have stuck with the simple proposal of NPP in the first place.Footnote 26

The results of the discussion can therefore perhaps be restated in conditional form: if one generally is on board with the project of giving value-based explanations, then one must countenance grounding of the type [it would be good if p, therefore p], at least as long as one wants to capture some very important moral intuitions. If one rejects the project of value-based grounding altogether, it is not clear how the worries against non-reductive views of normative powers are supposed to get off the ground to begin with.

6 A possible explanation of the results

So far, I have argued that the type of explanations that the wishful thinking objection takes issue with are crucial for making plausible value-based grounding stories for important other types of (moral) duty that a convincing normative theory would want to capture. Most importantly, these are the aforementioned duties to respect autonomy-rights, though there might indeed be others.Footnote 27

Ultimately, I believe that this appeal to “companions in guilt” may be enough to undermine the objection from wishful thinking, showing that we should be happy to countenance explanations of the type [it would be good if p, therefore p] in the domain of the normative, at least if we are on board with the project of grounding the entirety of this domain in the facts about value. However, it is one thing to show that we have reason to accept a certain claim, on pain of leading to counterintuitive results, and quite another to show why this is so. We may thus rightfully wonder what kind of consideration can be brought forward to give a deeper explanation of the validity of the arguments that I have offered above. At the maximum, such an explanation might provide for a defence of what we may call a Principle of Normative Optimism.Footnote 28

(Normative Optimism) Where P and Q denote two incompatible moral theories, whenever it is true that, if P were true, the world would be a better place than if Q were true, we have a reason to believe that P, rather than Q, is the correct moral theory.

In trying to provide such an explanation, we would not be alone. Recently, several philosophers, who were likewise struck with the prevalence of grounding explanations of the form [it would be good if p, therefore p] across various debates in normative ethics, have tried to offer defences of claims in the ballpark of Normative Optimism.Footnote 29

It is important to first note that nothing said so far entails that Normative Optimism is true. I have argued that for a number of high-level normative propositions, simply showing that it would be good if they were true can support the claim they are in fact true. This does not entail that this is true for all high-level normative propositions that constitute the true moral theory. Nonetheless, a sustained case for the truth of Normative Optimism would not only lend support the individual arguments above but could also provide a more profound understanding of them, weaving them together into a cohesive whole. While an attempt to make a full case for a claim such as Normative Optimism would take us too far afield at this point, it is nonetheless worth at least sketching the outlines such a case could take, as well as briefly dwelling on its prospects.

At first sight, Normative Optimism might strike some as a frankly outrageous claim. Is the claim that we live in the best of all worlds, normatively speaking, not simply almost as outrageous as Leibniz’ famous dictum that we live in the best of all worlds, period? While it may seem that way at first blush, I think the claim at least warrants some more careful consideration. It might very well just turn out to be amongst that select number of principles that appear more and more plausible the longer one thinks about it.

One important first step in rendering the principle more plausible is clarifying its contents. Most importantly, this concerns the notion of a moral theory. In the context of Normative Optimism, I understand a moral theory as simply and exclusively being composed of all true moral principles, where a moral principle is a general moral statement setting out the conditions under which, and because of which, an action is right or wrong, or favoured or disfavoured by moral reasons.Footnote 30

This restriction to the deontic side of things is important, for if Normative Optimism were to apply equally to questions of moral worth and/or value, it would not only be as implausible as the Leibnizian view, but in fact straightforwardly entail it (van Someren Greve, 2011: 448–9). If we had reason to always prefer the theory of value the truth of which would make the world better a better place, then we would ultimately be forced to settle on a theory that confers value to every state of affairs contained in the actual world. For surely, if that theory were true, the world would be as good a place as it could be. This, however, is clearly absurd.

For similar reasons, we must be careful to exclude deontic value from the value at issue in Normative Optimism. If the mere fact that a right action was performed were to make the world a better place in the sense at issue in the principle, we would again be forced into implausible conclusions. Under this assumption, a world in which all the actions which agents actually perform were morally required, down to the most menial and pointless, would be a world in which more deontic value is instantiated than one in which these acts are merely permissible. More right things are done, more deontic value is instantiated. Allowing for such value to make the relevant differences, Normative Optimism would then entail that we have reason to believe in this kind of ridiculous proliferation of requirements, which is again clearly absurd.

If, however, correctly constrained to just the non-deontic value of the kind of high-level normative principles which (PP) features amongst, Normative Optimism will appear much less outrageous than one might at first sight think it to be. Of course, it is not enough to show certain misconceptions about Normative Optimism to be false to render it plausible. At least some positive reasons to believe it are equally required.

One quite straightforward first reason to take seriously the possibility of Normative Optimism is simply the fact that, once properly restricted, counterexamples to the claim seem very hard to come by. It is difficult to envision a situation in which it is sensible to say something along the lines of “If only X were not a true moral principle! The world would be a better place for it. Alas, we have no choice but to follow it, if we want to act rightly.”Footnote 31

Consider what it would mean to defy one of the abovementioned arguments from the goodness of p (or badness of non-p) to p by simply denying Normative Optimism. Take a consequentialist who is faced with an overdemandingness-objection á la Slote (1985): her theory cannot be true because if it were, nobody would be allowed to pursue the kind of personal projects that confer meaning and value to our lives. Imagine the consequentialist answering in the following way: “Well, you are right, it is pretty bad for everyone that we cannot pursue our personal projects without violating our moral duties. But unfortunately for us, consequentialism just happens to be true. I wish a deontological theory were true instead, but the truth of consequentialism just is something we have to deal with, the world’s a tough place.” To me, this seems straightforwardly bizarre. A much more natural response would be to either deny that consequentialism does not permit us to pursue projects autonomously, or to argue that it is not all-things-considered bad that it does so, since other, more important values would otherwise not be served. Either way, simply conceding the badness of the truth of one’s favoured moral theory seems not be a live option. Normative Optimism is one straightforward way to explain this datum.

I therefore believe that at least those committed to something along the lines of VRN have reason to not to dismiss Normative Optimism offhand, and instead carefully consider arguments in its favour. To conclude the discussion, I will briefly summarise the core of one, as I take it, quite promising defence of the principle, recently put forward by Geoff Sayre-McCord (ms). Sayre-McCord’s argument builds on what he takes to be an important feature of normative concepts. As he puts it, normative concepts in general.

“mak[e] a claim on our allegiance, purporting to be such that their satisfaction ipso facto means someone or other has reason to act or react in certain ways to what satisfies them.” (Sayre-McCord ms., p. 16)

Sayre-McCord’s idea is that, since competence with our normative concepts commits us to application of their very same standards to themselves, arguments to the effect that there are better criteria for their application support the claim that the currently used criteria need to be rejected or revised (Sayre-McCord ms., p. 17). To put it bluntly, if our currently endorsed conception of what reasons are by itself yielded the verdict that we have reason to revise the concept, instead embracing reason*, this shows that we made a mistake in embracing the first conception to begin with.

As such, Sayre-McCord holds, to find that a set of moral principles is morally better than the set we have been thinking constituted morality is to find reason to think we have been wrong. This explains why the arguments sketched in Sect. 4 above do not strike us as absurd, while traditional cases of wishful thinking, like the desperate belief in one’s own popularity, do. There is a specific feature of normative concepts which establishes the kind of link between something being the moral goodness of p and the truth of p, rationally justifying the move from the former to the latter.

It is important to note, however, that if Sayre-McCord’s thoughts on the nature of normative concepts are to actually support Normative Optimism, and thus provide an explanation to the types of arguments listed above, they have to be read to implicitly presuppose an intimate link between value and reasons such as proposed by VRN. Unless we already assume that our normative concepts all eventually bottom out in value, it is not clear why application of their own standards to themselves should lead us to the set of principles that is best, rather than just least immoral or perhaps most rational for us to hold.

Sayre-McCord’s considerations, though by his own admittance themselves not yet fully developed, seem to me to point towards a simple but powerful way to explain the validity of the arguments in defence of the two-level version of NPP laid out above. Again, the arguments fielded in defence of this account against the wishful thinking objection do not presuppose anything like Normative Optimism, nor a specific explanation of the phenomena. The core of my argument in defence of the structure of the explanation at issue drew only on specific intuitions about the nature and applicability of autonomy rights, as well as the (potential) badness of their violation. This, I believe, is enough to vindicate the form of grounding explanation involved in the two-level version of NPP, at least when holding fixed the background assumption of a value-based conception of the normative (which I have argued we should, at least for dialectical reasons).

Nonetheless, arguments such as McCord’s can play an important role in contributing to a fuller understanding of the issues, showing us not only that, but also why types of reasoning that would strike us as objectionable wishful thinking elsewhere actually have their place in the normative. Further inquiry into, and elaboration of them is thus a task that is highly worthwhile.

7 Conclusion

In this paper, I have sketched a view of promissory obligation that combines a non-reductive view of the bindingness of promises with a value-based grounding explanation of why this bindingness obtains. I have argued that this view, as other views of normative powers like it, enjoys considerable initial appeal. As we have seen, however, accounts with this structure are subject to a powerful worry in that they appear to involve a type of explanation that normally amounts to objectionable wishful thinking. Here, I have argued that this wishful thinking worry ultimately need not dishearten defenders of these views. As I have shown, value-based accounts of normative powers should, when well-understood, only aspire to provide a value-based grounding story for the normative component of any given normative power. And with regards to just this, things are plausibly less problematic. I have argued that explanations of the form [It would be good if p, therefore p] must be countenanced, at least if we retain the framework of value-based grounding in the normative. For if we do not countenance them, we are forced to forced to give up extremely plausible claims in other areas of the normative, most importantly with regards to rights of autonomy.