Against the Entitlement Model of Obligation

Abstract

The purpose of this paper is to reject what I call the entitlement model of directed obligation: the view that we can conclude from X is obligated to Y that therefore Y has an entitlement against X. I argue that rejecting the model clears up many otherwise puzzling aspects of ordinary moral interaction. The main goal is not to offer a new theory of obligation and entitlement. It is rather to show that, contrary to what most philosophers have assumed, directed obligation and entitlement are not the same normative concept seen from two different perspectives. They are two very different concepts, and much is gained by keeping them distinct.

1.

Suppose you have my bowling shoes. I lent them to you, or maybe I forgot them in the alley and you picked them up. It doesn’t matter. You have something of mine and I want it back. Now, the fact that the shoes are mine is quite important. For one, it makes the fact that you want them too irrelevant. Indeed, it makes all your wants vis-à-vis the shoes irrelevant. That they fit you better, that you had planned to give them to your aunt for her sixty-eighth birthday—these might otherwise be excellent reasons for keeping the shoes. But they cease to be so if the shoes are mine. If they are mine, then it’s not up to you what to do with them. That’s just what it means for the shoes to be mine. And that is why I don’t have to ask you to give them back. I can demand that you do. I can say: “Give me the shoes!” Such a demand underscores the crucial difference between us: I have full authority—and you have none—over the fate of my bowling shoes.

This simple case of ownership helps illustrate the more general concept of ‘entitlement.’ To be entitled to X is to have the kind of normative control over X that I have over my shoes. An entitlement gives one full authority to decide over X (hence to ignore or disregard others’ desires or intentions about it) and therefore the standing to demand X from others (or from particular others, if the entitlement is restricted to a given person).

A question arises whether the concept of entitlement can be used to give a general account of directed obligation; whether, that is, we can use the logic of entitlement to understand the logic of the obligations we have to other people. To see why one may be tempted—as most philosophers have been—to answer in the affirmative, consider what’s going on when we make an agreement or a promise. Suppose you promise me to show up at noon for the dance rehearsal. You “give me your word,” as we say. But in giving me your word you also give me something else. Your promise gives me an entitlement over whatever it is that you promised. If I am entitled to your showing up at noon, then whether or not you feel like going is irrelevant. It is no longer up to you to decide. That is why I don’t have to ask you to keep the promise. I can demand that you do. “The promisee,” H. L. A. Hart writes, “has a temporary authority or sovereignty in relation to some specific matter [showing up for the rehearsal] over the other’s will” (1955, 183–84). That is, I have the authority over your affairs at noon that I have over the fate of my shoes. I can let you keep my shoes or I can demand that you hand them back. Similarly, I can waive the promise and let you do what you will, or I can demand that you be there. “[The promisee] is, in other words, morally in a position to determine by his choice how [the promisor] shall act and in this way to limit the [the promisor’s] freedom of choice” (180).

There is, then, one important class of directed¹ obligation, that of agreements and promises, which does seem to conform to the basic model from which we began. In both cases, one person (the owner, the promisee) has an entitlement, a claim,² to something (a physical object,³ an action) against another (everyone else, the promisor) and thus the authority to demand compliance with his choice. The thought, a popular one as it turns out, is that this logic just is the logic of obligation in general.⁴ That to be obligated just is to be at the other side of an entitlement and under the authority that it grounds. Your obligation to give the shoes back, in other words, is not something that follows from the fact that I am entitled to the shoes. Your obligation just is that fact seen from your perspective.

But is obligation a matter of complying with an entitlement against us? Consider the following cases:⁵

SANDWICH: On your way home you run into a homeless man. He looks at you, waves a hand, says he’s hungry. “Can I have something to eat?” he asks. You bought a sandwich in the morning and didn’t eat it. You know you probably won’t eat it.

TUESDAY DINNER: You and your wife have the following agreement: you make dinner on Mondays, Wednesdays, and Fridays, while she does so on Tuesdays, Thursdays, and Sundays. Suppose it is Tuesday and your wife, a pediatrician, calls you from the hospital in the midst of an emergency. “I’ll have to stay late,” she says. “Can you make dinner tonight?”

There’s a lot to say here, but for now, I want to draw attention to one point. Though it is absolutely clear that the sandwich is yours—that you, and not the man, are entitled to it, and so that it’s up to you, and not to him, to decide what to do with the food—it is far from clear that you are not obligated to the man to give it to him. Similarly, it is absolutely clear that Tuesday is your day off—that you, and not your wife, are entitled to the meal, and so that it’s up to you, and not to her, to decide what to do—and yet there remains a question of whether you are obligated to her to cook.

The intuition that you may be obligated already suggests that there is a problem with extending the model of entitlement to obligation in general. If obligation were a matter of giving others what they are entitled to—of what we owe to each other—there would be no space to ask whether you might be obligated. For if there is obligation here, it is precisely an obligation to give to others something you yourself—and thus not the obligees—are entitled to. The question of obligation does not seem to go away when there is no entitlement. And this is, I argue, because entitlement and obligation follow a different logic and respond to different considerations.

Now, the goal of this paper is not to spell out these considerations. Indeed, I won’t even attempt to say what these considerations are. That project will have to wait for a future time. My goal is to show that entitlement and directed obligation are different concepts, and therefore that it is a mistake to expect that in theorizing one, the other will come into focus. The hope is to convince the reader that a fuller picture of both obligation and entitlement can only emerge if we abandon the view that they come down to the same thing.

My strategy throughout the paper is the following. I bring out what looks at first like a puzzling feature of ordinary moral interaction and show that it is only puzzling under the assumption that entitlement and obligation are the same concept seen from two different perspectives. In each case, I argue, resisting this assumption leads to a better understanding of the phenomenon in question. Not any one case in isolation, but the cumulative strength of the evidence as a whole, should lead us to reject the entitlement model of obligation.

2.

I start, in this section, by exploring a puzzle about the obligee’s standing in relation to the obligor. What kind of authority, if any, does the former have over the latter? If you are obligated to me to X and you haven’t delivered, what can I do or say to you in response? The answer is clear for those who follow the entitlement model of obligation: I can demand that you comply with the obligation—that you do X. Such authority is given, in effect, the moment one identifies your obligation to X with a corresponding entitlement on my part to X. And this is precisely what philosophers have done ever since Hohfeld’s (1913) famous analysis of rights as equivalent to directed duties, an equivalence that philosophers today have, for the most part, accepted (see, e.g., Darwall 2012; Feinberg 1970; Gilbert 2018; Hart 1955; May 2012; Sreenivasan 2010; Thompson 2004; Thomson 1990; Wenar 2013; Zylberman 2014).⁶ In following Hohfeld, that is, philosophers have taken on board, either consciously or without much notice, the claim that obligees have the standing to demand that the obligor comply.

Let me pause here to clarify the claim I am ascribing to most philosophers and that I take to be standard fare in discussions of directed obligation. The claim is simply that directed obligation implies the obligee’s standing to demand. There are, of course, many and important differences among philosophers in how they understand this standing and its place in a theory of directed obligation. My point is that despite their many differences, they all agree at least with the following conditional: if you are obligated to me to X, then I have the standing to demand that you X.

To illustrate the point I’m trying to make, consider the following two examples of substantive disagreement among theorists of obligation. Julian Jonker (2020) has argued that the obligee’s standing to demand cannot be what makes directed obligation (in contrast to nondirected obligation) normatively significant.⁷ This is because the standing to demand, he argues, is not restricted to the obligee: third parties may also demand compliance with a directed obligation.⁸ This view stands in contrast to what he calls a Demand Theory of directed duties, which holds that “If J owes it to S to φ, then S has special standing to demand that J φ” (Jonker 2020, 5). Jonker’s point, contra those who support the Demand Theory (perhaps Darwall 2006 and Gilbert 2018), is that no such special standing exists. But notice that in rejecting the Demand Theory, Jonker is not rejecting the conditional I stated above—namely, that if you are obligated to me to X, then I have the standing to demand that you X. He simply holds that in addition to the obligee, other people may also demand compliance. Both Jonker and supporters of the Demand Theory, then, agree that directed obligation implies (or comes along with) the standing to demand.

Let’s turn to the second example. In The Moral Nexus, R. Jay Wallace presents an extensive attack on what he calls “voluntarist” theories of obligation. According to these theories, moral obligation is grounded in the actual or hypothetical demands that people make or could make on each other. The details of Wallace’s argument need not concern us (see Wallace 2019, chap 2, esp. 41–43). What’s important for my purposes is that while Wallace rejects the voluntarist picture, in which obligation itself is understood in terms of the standing to demand, he does not reject the claim that directed obligation involves such a standing. In fact, the opposite is true. Wallace comes back time and again to the idea that obligation is closely tied to accountability, which for him involves not only expecting others to comply with certain actions and standards, but also, crucially, the standing to demand that they do (see Wallace 2019, chap. 3). The lesson is the same: both Wallace and supporters of voluntarist theories, despite their dispute about the nature of obligation, are in complete agreement about the less ambitious claim that directed obligation involves the standing to demand compliance.

It is only this less ambitious claim that I am describing as a near platitude among philosophers. I am not suggesting that it is part of the entitlement model to maintain that the standing to demand is the very ground of obligation, or that we should understand the latter in terms of the former, or that such a standing is what makes directed obligation normatively significant. I am only noting that most philosophers hold that directed obligation corresponds with entitlement (or ‘claims,’ as the terminology may be), and thus that if you are obligated to me to X, I have, therefore, the standing to demand that you X. ⁹

With these clarifying remarks out of the way, let’s come back to the question with which I started this section. What kind of authority does the obligee have with respect to the obligor? On the entitlement model, as we have just seen, it is the authority to demand compliance.

Joel Feinberg (1970) offers the cleanest example of this dialectic. He begins his discussion of the value of rights by identifying rights with directed obligation. To say that a debtor is obligated to a creditor, he tells us, is to say that the latter has a right against the former. The obligation of the one and the right of the other “are precisely the same relation seen from two different vantage points, as inextricably linked as the two sides of the same coin” (249–50). The fact that the debtor is obligated to the creditor—the fact of directionality—is what distinguishes a relation of right from other normative relations. This is why in Nowheresville, Feinberg’s imaginary “world without rights,” all obligations are directed to an external authority, a “sovereign right-monopoly” (247). The point being that a world without rights just is a world without directed obligations (a world, that is, in which people are not obligated to each other).

Once Feinberg has made this move—once he has equated Y is obligated to Z to Z has a right (or entitlement) against Y—the question of the standing of the obligee has been virtually answered. For being entitled to X (having a right to X) implies the authority to decide upon X,¹⁰ which in turn implies the standing to demand X of others.¹¹ And this is exactly what Feinberg goes on to say. He asks: what’s wrong with a world without rights? What are the citizens of Nowheresville deprived of? His answer is that they lack the standing to make demands on each other:

Nowheresvillians, even when they are discriminated against invidiously, or left without the things they need, or otherwise badly treated, do not think to leap to their feet and make righteous demands against one another, though they may not hesitate to resort to force and trickery to get what they want (249).

For Feinberg, obligees not only have the standing to demand obligors to comply but, moreover, that is the value of occupying the normative place of obligee. It is a loss that the citizens of Nowheresville cannot claim anything from anybody (at least without the mediation of the sovereign)—that no one has the authority to demand of another to comply with his or her obligations. And so it is precisely this authority that gives rights, and therefore, given the previous identification, directed obligation, its point. Feinberg concludes, “there is no doubt that their characteristic use and that for which they are distinctively well suited, is to be claimed, demanded, affirmed, insisted upon” (252).¹²

On the entitlement model of obligation, then, obligees have the standing to demand, and such a standing is, at least for Feinberg, what distinguishes them as obligees. ¹³ In the rest of the section, I argue that this is false. Obligees, qua obligees, do not have the standing that the model grants them. Their standing, I believe, is more modest, more limited. It is the standing to blame, to complain, perhaps even to demand a justification, but it is not the standing to demand compliance.

Let me come back to SANDWICH and TUESDAY DINNER, our initial examples. I want to bracket for a moment the question of obligation to think about the question of standing. How should we characterize the authority of the homeless man and of your wife in relation to what they ask of you?

The question, put this way, contains the seed of its answer, for notice that both the man and your wife ask for, but do not demand, the things in question. And notice that this is not an artificial feature of the vignettes, but a fundamental feature in the structure of the cases. To see this, try replacing their requests for demands. Instead of the rather unassuming “Can I have something to eat?” of the man, imagine something like, “Give me that sandwich!” Or instead of your wife’s considerate “I’ll be late. Can you make dinner tonight?” try, “I’ll be late. Cook me some dinner!” The demands, I take it, sound altogether off, even strange. Why? Because demanding presupposes an authority that here the speakers clearly do not possess. If the sandwich is mine, then it is me, and not the man, who has the normative power—the authority—to decide what to do with it. But a demand denies me that power. As David Owens (2012) puts it, explicating Raz (1986):

The recipient of a valid order is not meant to judge the case on its merits. In particular he is not meant to treat the order as one factor in his deliberations, to be weighed against other considerations … Rather the order is there to preempt deliberation, to take the decision out of his hands. (86)

The man not only takes (or presumes to take) the decision out of your hands, but also in demanding, puts his hands all over your decision. He makes the decision for you and expects you to comply. It is this presumption of authority that strikes us a deeply misguided. We might even say… entitled.

The man (even if hungry) and your wife (even if in a serious emergency) do not have the standing to demand. This follows directly from the fact that they are not entitled to the things they want (and in fact you are). So far so good, says the entitlement model of obligation. No entitlement, no authority to demand. What’s the problem? The problem starts to show its tail once we turn our attention to the response we might expect from the man and your wife if you decline to do as asked.

Suppose you simply refuse to cook dinner on Tuesday. It’s not that you have something urgent or better to do. You think about it for a moment and decide that it is after all your day and, therefore, emergency or not, your wife should make dinner. Now, assuming a minimally healthy and loving relationship, I take it that your wife would be justified in resenting your decision. She might get angry with you (and not only at the situation), complain about your lack of regard, and expect some kind of apology later on. That is, your wife would be justified in blaming you (and the same is true, mutatis mutandis, for the homeless man).

The blaming response shows that your refusal runs afoul of a normative expectation—that your decision comes with the need to account for it. The man, in other words, might look at you for a while waiting to hear why you won’t give him the food you don’t plan to eat. And if there is no good reason, no justification to offer, resentment can, and may justifiably, ensue. One might deny that the man can really resent you. He may complain, of course, but does he have the grounds to blame you? A complaint might register disagreement with your decision, perhaps even anger, but resentment goes further. Resentment signals a normative expectation; it represents your action as wronging him. So one might properly ask: Would he be warranted in taking it further by resenting you? I think the answer is in the affirmative. It is one thing to refuse to give your food (or anything else) to a needy stranger if you are going to use it yourself. But to explicitly prefer to throw your food away than to give it to someone who needs it, and who is addressing you directly, is to express a cruel indifference to the plight of others,¹⁴ which the man is justified in resenting. We can summarize this discussion by saying that you are accountable to the man, or from the other direction, that he can hold you accountable.

Their standing turns out to be more complex than it might have at first appeared. Though you don’t owe them the thing in question, they can hold you accountable if you refuse to give it to them. They lack the standing to demand, but they have the standing to blame. Put differently, they can’t take the decision out of your hands, but they can complain if you decide badly. What to make of this?

The suggestion is that these are cases of obligation without entitlement. Their standing to blame, together with the need to explain yourself, to justify yourself to them, suggests the existence of a directed obligation. While their lack of authority to impose their will on you through a demand speaks of a lack of entitlement. The proposal is that things fall into place once we abandon the entitlement model of obligation. On my view, Feinberg et al. are absolutely correct in virtually everything they say about rights and their value. But they obscure important features of the obligee-obligor relation by failing to distinguish rights or entitlements from directed obligation. We need to keep them distinct to see the complete picture.

3.

In this section, I continue to build the case against the entitlement model by looking at a different contrast. The focus here is between the blaming response of obligees when a violation has been flouted and their response when it is fulfilled.

If you have something of mine, like my bowling shoes, and you do give them back, I may say “thanks,” by way of courtesy. I may appreciate your respecting my rights and even acknowledge that fact to you. But nothing more than that would be needed or expected from my part. This is true in general when we receive something we were entitled to receive. The fitting attitude towards your Xing, if you owed me X, is some sort of acknowledgment that the debt has been paid off.

Now consider the attitude with which the man, or your wife, would receive the things in question. How ought the man receive your sandwich if you do decide to give it to him? What response should we expect from your wife if you make dinner on Tuesday? I surmise that we would see a very different attitude from (and in a sense the mirror image of) the “acknowledgment of payment” attitude that typically marks cases of entitlement. The appropriate attitude here—we might call it ‘gratitude’ in the absence of a more specific word—is one that expresses, and is meant specifically to express, precisely a lack of entitlement. The man and your wife would be thankful, not for receiving something owed to them but rather for the opposite.¹⁵

A good way to get a feel for the contrast is to compare your wife’s attitude to Tuesday dinner with her attitude to the regular Monday dinners. On Mondays, she is simply getting what you owe her, and her demeanor would likely reflect this. If she thanks you after dinner, as she probably does, her thanks are almost a formality, a display of good manners. They simply recognize that the agreement has been respected, that you held your end of the bargain. But things are very different on Tuesday. Her overall stance here should instead be one of gratitude. The difference is not one of degree but of kind. The point is not that in one case her reaction is understated while in the other is overflowing with excitement. The point is that no matter how she happens to express it—regardless of the idiosyncrasies of her character and the particularities of the situation—her attitude would reflect the normative fact that she received something she was not entitled to receive.

This response is hardly worth remarking in itself. Considered in isolation, that is, it raises no issues to the entitlement model of obligation. But the response becomes interesting, and becomes problematic for the model, when we see it in light of the response we outlined in the previous section—in light of what happens when you do not accede to their requests. For notice where things stand: when you do as asked, they receive your actions with the kind of gratitude that signals a lack of entitlement. But when you don’t, their response is a blaming response. They seem justified in expecting a justification and, in its absence, in resenting your actions.¹⁶

In short, we arrive (through a different route) at a now familiar place. And that familiar place is hard to make sense of under the entitlement model of obligation. For, again, the suggestion is that both sides of the response become intelligible only after rejecting the model; that is, once we see the blaming response as picking out an obligation which, as shown by how its fulfillment is received (that is, with gratitude), is not accompanied by a corresponding entitlement.

4.

How might the entitlement model accommodate the results of the last two sections? The trouble for the model, in a nutshell, is that it seems incapable of explaining both sides of the relevant contrasts. Take the contrast of the last section. If TUESDAY DINNER is a case of obligation, and hence of entitlement, how to account for your wife’s gratitude? And if it isn’t, if you are not obligated to cook, how to account for her standing to blame you? Either way the model has some explaining to do. In this section, I consider, on behalf of the model, the first horn: the view that your wife is indeed entitled to the meal. I consider the second horn in the next section. I argue that both horns are dead ends: the model lacks the resources to make sense of the facts.

The model’s problem in treating our cases as cases of obligation is to account for the way your actions are received when you do as asked.¹⁷ The problem, I claimed, is that the response seems to signal a lack of entitlement. The model, then, needs to deny this claim. In what follows, I consider two different strategies. The first strategy proceeds by way of counterexamples. The second strategy holds that I have mistaken the object of gratitude in the cases at hand. The goal of the two is the same: both attempt to show that feeling gratitude for X is not incompatible with being entitled to X.

Let me start with the first strategy. Supporters of the model can point out that gratitude is a common response even in uncontroversial cases of entitlement, and so that regardless of how we describe it, its presence need not express a lack of entitlement. I have in mind, particularly, cases of promises in which we appear to respond with gratitude at seeing them fulfilled. If you promise to do something that goes beyond what I could reasonably expect from you, then, given the promise, it seems that I am entitled to your compliance. And yet gratitude would be a very natural response to your carrying out the promise.

I don’t think these examples do much for the model. This is because even here gratitude is responding to a lack of entitlement. Consider the following mundane example: “Thanks so much for showing up today!” “It’s nothing, I promised I’d do that.” “Yes, you promised and therefore I was entitled to your showing up. The entitlement gave me the assurance to rely on your word, and hence the standing to hold you accountable if you let me down. But you didn’t have to promise. You were under no obligation to do the things you are now obligated to do. So yes, you did give me something I was not entitled to get, namely, the promise itself and thus the very entitlement that I now have. I am grateful for that.” That is, the voluntary undertaking, the issuing of the promise (with its ensuing, nonvoluntary fulfillment), is often a fitting object of gratitude. But this is because the promisee was not entitled to the promise.¹⁸ The promisor did not owe the promisee the special kind of authority that the promise gives him. The entitlement is in an important sense a gift, and this explains why gratitude is an appropriate response to it. But this feature is peculiar to promises and as a result it is of no use to the entitlement model in dealing with our cases.

Let’s now turn to the second strategy. The idea here is to give a different account of how gratitude functions in the cases at hand. Particularly, the goal is to explain how gratitude can in general be a fitting response to receiving something one was entitled to receive. Here’s the proposal. Gratitude is fitting when someone fulfills an obligation toward you (and thus, according to the model, when someone gives you what you were entitled to get from them), provided you benefit from their action and the benefit is not part of what makes the action obligatory (and thus not part of what you were entitled to). Take the following dialogue from TUESDAY DINNER. “Thanks for making dinner.” “No problem. That’s what a spouse does in this situation.” “True, but it was also delicious!” Your wife, in other words, is grateful for a benefit (a delicious meal) that resulted from your fulfilling an obligation (a decent meal). But you didn’t have an obligation to make dinner because the act would provide her with a delicious meal.¹⁹ The conferral of the benefit is thus a fitting object of gratitude, but not the obligatory act. That is, contrary to my description, the presence of gratitude here does not signal a lack of entitlement. Or rather, it signals a lack of entitlement though not for the action under dispute (dinner on Tuesday), but only for one of its benefits (a delicious meal).²⁰

I think this proposal is mistaken because the principle behind it is false: we are often grateful precisely for the benefit that grounds an obligation. The benefit the homeless man accrues in your fulfilling your obligation, the alleviation of hunger, plays a central role in grounding the obligation to give him your food. And the same, I believe, is true of your wife. To see this, suppose you don’t make her a delicious meal. Due to your poor cooking skills, you only make her a decent meal. I take it that gratitude would still be in order. And I take it that she would be grateful, at least in part, for not having to make dinner after a late night at work. But this benefit is indeed part of what makes the action obligatory. Gratitude, therefore, can be a direct response to the benefit that grounds an obligation.

More importantly, gratitude might remain fitting even in the absence of any benefit—even if, for instance, the meal turns out so bad as to be uneatable (supposing you tried in good faith to make a decent meal). This suggests, I think, that your wife’s response is not really about her benefiting from your action, but rather aims to express something deeper about the action itself. What can that be? The current proposal has no answer at this point. And yet the answer, if we reject the entitlement model of obligation, is clear. Her gratitude expresses that she received something she was not entitled to receive. And this fact is no less important—and no less worth expressing—if what she receives brings her no benefit.

The entitlement model of obligation does not fare well under the first horn. I have considered two attempts to render gratitude compatible with entitlement. None of them worked. There may be better attempts I have failed to consider. I believe, however, that they are bound to fail. The present discussion gives us strong reasons to believe that gratitude, at least as expressed in our cases, responds to a lack of entitlement and that it functions to acknowledge, both to oneself and to others, this very fact.

The better approach, it now appears, is to take the second horn—to deny that these are cases of obligation.

5.

The second horn may seem, at first glance, less threatening to the entitlement model of obligation. By denying obligation to TUESDAY DINNER and SANDWICH the model can simply grant the analysis of gratitude I have given. Your wife should be grateful because she is not entitled to your cooking for her. In fact, the model adds, gratitude is what we should expect given the supererogatory nature of the dinner. The model can similarly embrace our conclusions on the standing to demand: if there is no obligation, there is no entitlement. And if there is no entitlement, there is no standing to demand. What’s left, of course, is blame. The model needs to explain why your wife is justified in blaming you if you had no obligation to cook.

This might not seem that big of a problem. Many philosophers have come up with cases that attempt to show that one can be blameworthy without violating an obligation. The upshot, according to these authors, is that it is a mistake to infer from the fact that one is worthy of blame that one has violated an obligation. Consider:

• • A doctor tries to poison her patient by giving her a pill that turns out to be the patient’s cure. (Haji 1998)

• • In wartime, a prime minister approves a military operation that will cause only “proportional” harm while minimizing harm to noncombatants, but does so with the express desire to “undermine morale by killing the civilians.” (Scanlon 2008, 31)

• • A woman kills a man solely out of malice. Unbeknownst to her, the man was about to kill her daughter and killing him was the only way to prevent him from so doing. (Capes 2012)

Now, I am not for a second convinced that these are genuine instances of blameworthiness without wrongdoing. Still, to evaluate the claim would force us into some muddy waters in the theory of action and metaethics. Since I have no desire to swim on those waters, I will simply avoid them. So let me grant, for the sake of argument, that these cases show what they intend to show. The question becomes: Are they of any help to the model? Do they show that it is a mistake to treat your wife’s blame as evidence that you had an obligation to make dinner? I think not.

The cases, though different in subtle ways, all share the following structure: there is a principle that renders a course of action permissible (e.g., a principle permitting the killing of unjust aggressors) and there is an agent who undertakes that very course of action (kills the unjust aggressor). But what moves the agent to act (malice) is not what makes the action permissible (the protection of an innocent victim against an unjust aggressor). The considerations that make the action permissible thus play no role whatsoever in the agent’s motivation to act. (She would have killed him even if he were not an unjust aggressor; indeed this is what she took herself to be doing.) This is what enables the divergence between blameworthiness and wrongdoing to take place. We can test this explanation by observing that had they acted from the right considerations (to defend her daughter, to cure her patient, to respect the laws of war) their actions would cease to be blameworthy. There would be nothing to blame them for.

But once we see this, once we understand what’s behind the alleged divergence between blame and obligation, it becomes clear that these cases are of no help to the model. For it is clear that our cases do not have this structure. Start by positing a principle rendering your actions permissible. A natural candidate, it seems, would appeal to the fact that the man is not entitled to your sandwich or your wife to the meal. In TUESDAY DINNER, for instance, the principle might be something like this: given the agreement, your wife is not entitled to your time, and therefore you are permitted to do whatever you please on Tuesday evenings. Now suppose you act from (and not only according with) this principle; in other words, your motives pick out exactly the considerations that make the action permissible. Suppose you reason thus: “I understand that my wife had an emergency, but I feel like taking a long nap. And since it is Tuesday, and she is not entitled to a meal, I am permitted to do so.” Finally, ask whether your action ceases to be blameworthy. I take it that the answer is in the negative. Here, in contrast to the cases from Haji and company, acting from the considerations that (allegedly) make the action permissible, if anything, makes the judgment that you are blameworthy even clearer.

The lesson: even if (and the “if” is no small one) the cases show that wrongdoing is not a necessary condition for blameworthiness, they offer no explanation whatsoever for why, in TUESDAY DINNER, you are blameworthy if there is no obligation.

So these cases by themselves cannot rescue the model. But perhaps they offer a clue that can. Let’s start with this question: Why is the woman in Capes’s case blameworthy if she acted permissibly? Because, the story goes, her reasons for action—her “quality of will,” in Capes’s preferred terms—are morally objectionable. Killing the man out of malice is morally objectionable and hence blameworthy, even if killing the man, given the facts of the situation, is permissible. This shows that blame responds directly to the agent’s quality of will, and only indirectly to wrongdoing per se.

Now, the two often go together. Acting in ways that express ill will tends to result in wrongdoing. But not always. In one kind of case—the kind presented by Haji, Scanlon, and Capes—this is because the agent’s action is made permissible for reasons other than those on which the agent acted. But there could be other kinds of cases, TUESDAY DINNER and SANDWICH among them, in which acting for the very reasons that make the action permissible itself expresses ill will. Or to put it differently, a principle of permissibility might function precisely to permit certain actions even if they express ill will. For example, the principle that given the agreement, you are permitted to do whatever you want on Tuesday evenings permits you to disregard your wife’s needs. The point is not that napping is not objectionable, but rather that the principle makes such objectionable behavior morally permissible. If this is true, napping on Tuesday would be suberogatory—that is, “bad to do, but not forbidden” (Driver 1992, 286).

But why think there is such a principle? If the action is bad—if it expresses ill will and is therefore blameworthy—why think that it is permissible? Well, supporters of the suberogatory might reason, because your wife is not entitled to the dinner. Because she has no claim to your time. Because it is your day off. Needless to say, this response would not do in this context. For this is simply to assume what’s under dispute, namely, the central claim of the entitlement model of obligation.²¹ Appealing to the fact that your wife is not entitled to the meal to account for the permissibility of not making dinner is to assert rather than to argue for the model.

Still, the real question at this point is whether there is something substantive under dispute. If napping is morally objectionable, if feelings of resentment and hurt are warranted, if your wife can hold you accountable through blame, if guilt is justified, if it is intelligible for you to apologize and for her to forgive you, if we agree about all of this … what does it matter how we call it? I, for one, do not care. To me, the force of calling an action obligatory is precisely to justify this mosaic of attitudes and responses. What’s the point of calling an action permissible and then go on to say that it justifies the full array of the accountability practices? If there is no motivation other than to salvage the entitlement model of obligation, I suggest that we give up the model.²² But the suggestion, at this point in the argument, is terminological. I’m not fighting for a label.

6.

In making the case against the entitlement model of obligation, I have drawn from an influential view of obligation according to which it gives the obligee the standing to hold the obligor accountable through blame and the rest of the accountability practices. When our actions affect other people, particularly when it affects them in adverse ways, the question arises as to whether we can give those affected a proper account—a justification for treating them in this or that way. Directed obligation, on this view, serves the function of marking out the normative fact that a given course of action would not be justifiable to a particular person, and therefore that the latter can rightly hold the agent accountable.

In this last section, I want to consider the implication of this view for the role of obligation in deliberation. My claim is going to be that it is a mistake to think, with most philosophers, that obligation serves the function of constraining deliberation. If obligation is about justifiability—about justifying your actions to other people—then we need to deliberate to find out whether our actions are justifiable to others. Obligation, I argue, is a conclusion of deliberation. For it is not given to the agent, in advance, whether a course of action is justifiable or not. Nevertheless, there is indeed a normative concept whose function is precisely to constrain deliberation. The reader will not be surprised to learn that this is the concept of entitlement. The lesson of this section is that we need to distinguish obligation from entitlement in order to accommodate this important aspect of practical deliberation.

It is a platitude among philosophers that obligation has a distinct effect on practical deliberation. The central idea is that obligation shapes deliberation by excluding otherwise valid reasons from the agent’s consideration. The fact that I feel grumpy is a reason for me to skip the high school reunion later today. But if I promise to meet you there, and I know that you are relying on my presence, then my mood ceases to be relevant in deciding whether to attend. Or at least it ceases to have the force it had before I made the promise. Before the promise, I can weigh the reasons for not going (I feel grumpy, Doug may be there, etc.) against the reasons for going (the beer will be free, Doug may not be there, etc.), and reach a decision. But the promise, in effect, functions to block this kind of weighing—it works, to use Wallace’s words, “as a constraint on our agency” (2019, 30). The promise is not one more thing to weigh in deciding what to do, but rather influences what gets weighed, the weight of those things, and even whether there should be a weighing at all. Wallace again:

We might summarize these distinctive features by saying that practical requirements²³ enter the deliberative field in the guise of presumptive constraints on the agent’s behavior. They function in this way, insofar as their deliberative role is to determine certain options for action to be either on or off the table from the start, fixing assumptions within and around which the rest of our planning agency will operate, as it proceeds. (27)

There is of course a significant amount of disagreement about how precisely to understand the nature of this constraint.²⁴ But that obligation functions to constrain (or limit or preempt) deliberation no one debates, and indeed is taken as one of the central features a theory of obligation must explain.

And yet the claim is wrong. And wrong in the exact same way that Feinberg’s claim about the standing to demand is wrong, which is to say that it is entirely correct—but it is correct as a claim about entitlement. The mistake, just like Feinberg’s mistake, is to think that what is true of one must be true of the other (or rather, to think that one and the other are the same thing, and thus that theorizing about one is theorizing about the other). On the contrary, obligations and entitlement play different roles in practical deliberation. What functions as a constraint on deliberation is entitlement. It is the fact that the obligee has an entitlement against the obligor that reconfigures the “deliberative field” in the way Wallace describes. Obligation, on the other hand, serves a different function. It signals that a particular course of action would not be justifiable, and thus that the obligee can hold the obligor accountable. But the obligor needs to deliberate to this conclusion. Obligation is thus a result of deliberation—not something that constrains or preempts it, but rather a conclusion one arrives through it.

To start, consider again TUESDAY DINNER. Suppose you have been quite busy at work and at last have free time on Tuesday evening. You’ve been looking forward to lying down on the couch, drinking some cold Arizona tea, and finishing The Charterhouse of Parma. You are pouring the tea when your wife calls:

“Can you cook dinner tonight? I’m running late.”

“Well …”

“I know it’s Tuesday, but one of my patients had …”

“… an emergency. Right, well I kind of had plans for the evening.”

“Oh, really. What plans?”

“I was hoping to finish Stendhal.”

“Stendhal? You mean you were hoping to lie down on the couch. I’m at the hospital dealing with an emergency!”

“I’ve had two crazy weeks at work myself.”

“I know. But I’m not asking for a five-course meal. You can still read Stendhal if you cook something easy.”

“I don’t know about that.”

“You mean you can’t interrupt your reading for fifteen minutes … to feed your wife?”

* And so on ad nauseam.

You and your wife disagree about the weight of the relevant considerations. That’s what makes the argument infuriating and potentially infinite. But the disagreement reveals a deeper kind of agreement—one that underlies and structures the whole exchange—namely, that the way to resolve the present dispute is by weighing; by considering the force of one consideration against another. Your wife thinks, and wants to convince you, that her emergency at the hospital has more weight than tea and Stendhal. But that’s the point; she wants you to weigh the two and see that there is no valid justification not to. Compare this with the following exchange, now on Monday.

“I’ll be late tonight. Start dinner without me.”

“About that … I wasn’t planning to cook tonight.”

“What do you mean?”

“Well, I kind of had other plans.”

“We have an agreement, remember?”

“I know, but I was hoping to finish The Charterhouse of Parma.”

“But we have an agreement.”

“Yes, but …”

*She hangs up the phone.

This dispute might at first seem just like the previous one. But I want to suggest that it is quite different. In the first case, your wife objects to your plans by presenting you with reasons to consider against other reasons. She objects to a particular weighing of these reasons, and her purpose in the argument is to make a case for her way of evaluating the different considerations. In the second exchange, however, she is objecting to the weighing itself. Her purpose in insisting on the agreement is precisely to insist that, given the agreement, there should be no weighing of reasons.²⁵ She doesn’t need to hear your plans (whether you’re reading Stendhal or Stephenie Meyer is beside the point). The agreement blocks any plan from mattering. In both cases, then, she complains about your unwillingness to cook, but the complaints are of a different kind. One is about the weighing, about your failure to take the proper force of each fact into account. The second complaint is that of weighing—not how you weighed, but that you weighed.

My claim is that these back-and-forths with your wife have their exact analogues in first person deliberation. They reflect, in other words, the structural difference between obligation and entitlement in relation to practical deliberation. Your wife asks you to cook on Tuesday, letting you know about the emergency. What should you do? It depends, of course. But there is a question and, crucially, the way to answer it is by deliberating. “On the one hand, I am tired and I really want to finish the book,” you might think, “On the other, my wife’s (or Antonia’s²⁶) in an emergency.” You deliberate and conclude that The Charterhouse of Parma is not enough to justify—to justify to your wife—a refusal to make dinner. Making dinner, you come to think, is the only justifiable decision. It is at this point that you have arrived at the conclusion that you are obligated to her to cook. But the conclusion is something you arrive at through weighing—through a process of deliberation (even if a short one). Notice here that the conclusion is different from a judgment about what you have more reasons to do. For what you learn when you conclude that your action would not be justifiable to your wife is not that it would be imprudent or irrational to do it, but rather that she can hold you accountable for it. Still, the central point is that this important conclusion is a conclusion that you draw in deliberation by weighing your interests against hers and deciding where the balance of reasons lies.

Things are different on Monday, when she’s entitled to the dinner. To deliberate here is already a mistake. To think, “On the one hand, I made an agreement; on the other, I really want to finish The Charterhouse of Parma,” is to misunderstand, or to disregard, the normative force of the agreement and the entitlement that it grounds.²⁷ Here Wallace’s description seems absolutely right. Her entitlement, qua entitlement, enters your deliberative field from the start, fixing the assumption that you ought to make dinner, and thus constraining any further deliberation about your evening plans. In short: on Tuesday, “I am obligated to cook” is the result of deliberation—the conclusion of the right weighing of reasons. On Monday, it is a constraint, preventing further deliberation about it.²⁸

This difference leads us back to the beginning of the essay, where we considered the standing of the obligee to demand compliance from the obligor.

If I am right, and entitlement, but not obligation, serves to constrain deliberation, it makes sense that obligees can demand compliance only when they have an entitlement against the obligor. For demands, as we saw, function to block deliberation. Let me quote again the bit from Owens (2012):

The recipient of a valid order is not meant to judge the case on its merits. In particular he is not meant to treat the order as one factor in his deliberations, to be weighed against other considerations … Rather the order is there to preempt deliberation, to take the decision out of his hands. (86)

In issuing a demand, then, the obligee is insisting that the obligor respect the normative facts. The speech act of demanding reiterates what is already the case,²⁹ namely, that the entitlement ought to be treated as a presumptive constraint on deliberation. On the other hand, your wife’s request puts forward her emergency as a reason for you to take into account in deliberation. In asking, she expresses her understanding that her emergency does not settle whether you should cook; that there might be other valid reasons to consider.

This also explains, I think, why she doesn’t point directly to the obligation in the initial exchange. If you are obligated to cook, that is, why doesn’t she say so—why doesn’t she say, “Forget about Stendhal, you are obligated to cook”? The answer is that saying this, though true, would be to beg the question, to assume from the start that the balance of reasons is such that no other course of action is justifiable. And this would be to deny that the dispute ought to be resolved by actually weighing the reasons for and against. As we have seen, this is precisely what goes on in the second exchange, when she has an entitlement against you. Insisting on the obligation from the start, that is, would get perilously close to presuming that she is entitled to the dinner. Instead, she argues for the conclusion. She makes the case—pointing out, for example, that you can both cook and read—that there is no justification for refusing her request. She has to make the case that you are obligated to her to cook. With entitlement, the case is closed from the start.

7.

My aim has been a modest one. I have not attempted, even in preliminary form, to give a theory of directed obligation and entitlement. Indeed, it is one of the upshots of this paper that an adequate theory of these things can only emerge if we realize that what we need is not one theory but two. I have restricted my focus to arguing for this realization.

The separation of obligation from entitlement, however, is not simply a prelude to a future theory, but carries important implications on its own. It opens up the scope of moral obligation and accountability, revealing that our obligations to others go beyond what they are entitled to get from us. To put it paradoxically, we owe others more than what we owe them. This enlargement of the moral terrain, though, brings with it more, rather than less, freedom. For when it comes to our obligations, no one, not even those to whom we are obligated, has the authority to demand that we fulfill them. It is fully on us.