a. A virtue-theoretic account of moral praiseworthiness and blameworthiness

In drawing out the sort of virtue theoretic account I have in mind, it will be helpful to look to Nomy Arpaly’s account of moral praiseworthiness and blameworthiness that she develops in Unprincipled Virtue. ^{Footnote 10} Arpaly defends two theses concerning moral praiseworthiness. First, for an agent to be praiseworthy is for her to have done the right thing for the relevant moral reasons.^{Footnote 11} This observation should be familiar. It is what animates Kant’s well-known grocer example.^{Footnote 12} The relevant moral reasons, in turn, should be understood as those considerations that are fitting given the circumstances. To get the basic idea, consider that there may be circumstances where worrying about whether one will dirty one’s clothes is fitting—for instance, if there are puddles on a street and one is walking to an important business meeting—and there can be circumstances where it is not—for instance, at the scene of a very bad car wreck. And a person who is preoccupied with cleanliness at the scene of a car wreck is to that degree less praiseworthy than he would be were he to have been focused solely on the needs of the victims of the crash.^{Footnote 13} The fastidious would-be rescuer is simply less heroic than the rescuer who is unconcerned about the state of his wardrobe. And this is so even if the fastidious rescuer soldiers on despite his misgivings about whether his drycleaner can save his pants.

Second, Arpaly argues that an agent is “more praiseworthy, other things being equal, the deeper the moral concern that has led to her action.”^{Footnote 14} Having deep concern is associated, Arpaly claims, with three things. First, a certain “diehard quality”—the more concern you have for a consideration, the more it takes to stop you from acting in accordance with it. Consider two agents, one who acts morally only when it is convenient, and the other who acts morally even in the face of substantial impediments. The latter is more morally praiseworthy than the first even when they do the same thing. Second, those who have moral concern have associated emotional dispositions. For instance, a person who cares deeply about his spouse’s well-being will take pride in her successes, experience joy when she is happy, and sadness when she suffers. Third, a praiseworthy person has a kind of awareness of those considerations that are morally relevant. Those who care about sexism are more aware of sexism when it occurs, for instance. The upshot of this sort of view is that being reason-responsive in the morally best sort of way is not just having a set of beliefs about what reasons there are, but involves an entire motivational complex, which constitutes an agent’s “read” on a moral situation.^{Footnote 15} One isn’t merely in an intellectual state of believing something to be a reason, but rather one is moved in certain ways and aware of certain things. Those considerations that are most morally relevant occupy the mind of the virtuous agent.

b. Why exclusionary reasons are at odds with the virtue theoretic account of moral motivation

However, if one accepts that moral praiseworthiness and blameworthiness should be understood in terms of responsiveness to objective moral reasons, and that responsiveness to moral reasons involves a set of motivational dispositions described above, then it will turn out that exclusionary reasons are at odds with a sound account of moral reasoning. On the exclusionary reasons model, when a person is confronted with an excluded moral reason, she should either not have the degree of concern for it that would typically accompany a virtuous outlook given its weight—e.g., a diehard quality, an awareness of its relevance, and the associated affective responses, or she should manifest these characteristics yet act contrary to those dispositions. Neither option is acceptable. The first option would entail that if moral reasons are excluded, then the traits that typify a virtuous disposition would be incompatible with the sorts of motivations an agent should have in the face of those excluded reasons. Such a result should be unacceptable to anyone sympathetic to the virtue theoretic account of moral motivation noted above.

The second option—that an agent faced with excluded moral reasons should manifest the kinds of dispositions characteristic of virtuous agency, but act contrary to those dispositions, however, creates a kind of cognitive dissonance that cannot plausibly figure into any account of virtuous agency.^{Footnote 16} To see why, let’s return to Jeremy. If Jeremy is virtuous and the excluded reason—the harm to the tradesman—is a compelling one, then Jeremy would feel the pull of acting contrary to his orders. But, because any competing reasons are excluded, he must muscle through, despite his better sentiments, and do what he has been ordered to do. In the normal case, when a morally decent person is confronted with what she takes to be a pressing moral reason, she acts. So, what we have in the case where there is an exclusionary reason—an order, say—and a pressing first order reason to act contrary to the order, is a conflict between the recognition of something as morally salient in the case, and the putative order or command. Exclusionary reasons put the virtuous agent in a position of recognizing something a reason, and hence being motivated by it, but simultaneously judging that it would be inappropriate to actually be so motivated.^{Footnote 17}

Raz himself is aware of the apparent motivational paradox that arises if we accept his account of exclusionary reasons, though it is not clear that he grasps its full implications. He allows that there is something a little paradoxical about “reasons for not being guided by reasons whose very nature is that they should guide.”^{Footnote 18} But he tries to lessen this worry by arguing that reasons are mere candidate guides. That is, reasons are only potential guides; they need not actually motivate a person in acting as she does. As Raz puts it, reasons require conformity, not compliance. One conforms with a reason if one does what the reason favors. One complies if one also acts for that reason. For instance, if Samantha ought to give money to a charity because doing so will alleviate animal suffering, she conforms if she gives, even if merely to impress her friends, but she complies if she gives out of concern for the animals. And Raz thinks that if reasons don’t require compliance this will partially undermine one source of resistance to exclusionary reasons—the assumption that reasons ought to motivate an agent. But, as should be apparent, one upshot of the virtue theoretic account is that, contra Raz, at least moral reasons do indeed require compliance.

Raz cites three considerations that favor the view that reasons require only conformity. First, Raz notes that there may be multiple reasons for an act, each of which is sufficient to warrant acting in that way, and there is no fault in being motivated by only some of these.^{Footnote 19} For instance, going to the party may be enjoyable for me, it may please my friend, and it may help raise money for a good cause. If I go to the party because it will please my friend, without even considering the other reasons, then on Raz’s account, I have done nothing wrong, assuming that going to the party is what I ought to do.

Second, Raz argues that if we accept that reasons can be outweighed, there is nothing wrong with not being motivated by the outweighed reasons. As Raz observes, if a person could spend the afternoon visiting his mother, or alternately weeding the garden, and the weight of reasons favors visiting his mother, there is no sense in which he does something wrong if he visits his mother, and pays no mind to his garden.^{Footnote 20} Indeed, there is no sense in which he does something wrong if he doesn’t give his garden even a passing thought. And so, Raz argues, it can’t be the case that reasons require compliance, because there is no fault in failing to attend to an outweighed reason.

Third, Raz notes that in the case of acts that we ought not do—kill another, say—it is often most admirable to not act for a reason at all. That is, the person who never considers the possibility of killing is more admirable than the person who entertains the thought and refrains because killing is wrong. As Raz puts the point in discussing a motive to kill, “the moment we are morally motivated in such cases we are behaving in a less admirable way than those to whom the wrongful act simply does not occur.”^{Footnote 21}

None of these three arguments is compelling. The first—the fact that one does nothing wrong in being motivated by only some of the considerations that favor what one ought to do—seems to rely on the thought that the moral worth of an act should be assessed along a single axis—i.e., whether the act was right or wrong. There may be a point to thinking that there is a fact of the matter about what we ought to do all things considered. But thinking of moral worth as uni-dimensional, and assessed primarily in terms of whether one did what one ought to have done all-things-considered, is an impoverished picture of the moral landscape, and it encourages an impoverished picture of the contours of sound practical rationality. Rather, we can assess an agent and her acts along many different dimensions. If one attends a fundraising event with no concern whatsoever for the cause, but only to please a friend, this might show a person to be a good friend, but also shows her not to be particularly generous. If one does it solely because she will have a good time, the act involves a kind of self-absorption. There are many senses in which an act can be assessed as praiseworthy or blameworthy. The fact that one can be morally criticisable in cases like this—those where the reasons over-determine what one ought to do, and one acts for some subset of those reasons—suggests that mere conformity is not enough, at least in many cases. We may not expect an agent to act with an eye towards all the reasons that might bear on a practical situation, but this fact doesn’t license the further conclusion that we are unconcerned with the motivational makeup of an agent when she acts, so long as she does what she ought.

The second consideration—that there is nothing wrong with not being motivated by an outweighed reason—mischaracterizes the resistance on the part of the person who objects to exclusionary reasons.^{Footnote 22} Raz characterizes the resistance to exclusionary reasons as arising out of the thought that it is in the nature of reasons that they guide, and “[s]urely there cannot be reasons for not being guided by reasons whose very nature is that they should guide.”^{Footnote 23} But, the resistance to exclusionary reasons arises not from the fact that there can be reasons that ought not to be motivationally effective—that is, that ought not lead to action—but from the oddness of being in a situation where one correctly judges that there are stronger reasons to prefer one course of action to another, but that one ought to not be motivated by what one correctly judges to be a stronger reason to act. That is, the resistance is not to the idea that there could be reasons that ought not guide—that will be true of any outweighed reason. Rather, it is to the idea that one ought not be guided by what is otherwise a sound exercise of practical reason—where one is getting the first order reasons right. Exclusionary reasons require that one shouldn’t be guided by what, at the first-order level, is an exemplary exercise of practical rationality. And it is this result, not the mere fact that there could be some reasons that ought not to guide, that grounds the resistance.

And Raz’s assertion that there is “nothing wrong in not being aware of, and not being motivated to act for, reasons which are overridden”^{Footnote 24} is far too strong. Of course, an agent confronted with competing first-order reasons will also be in a position of judging that something is a reason, but also judging that she ought to act contrary to that reason.^{Footnote 25} Those partial to virtue ethics often describe such situations as involving a remainder of some kind. And, indeed, it is often more fitting for an agent to act with regret, say, since that is what it means to give the reasons their due in cases where there is a bona fide conflict.^{Footnote 26} But that virtue-theoretic insight runs contrary to the assertion that there is nothing wrong with not being motivated by a reason that is outweighed. If I am forced to choose between saving my nephew and saving my dog, presumably I ought to save my nephew, but a complete lack of concern for my dog seems callous at the very least. Indeed, it is noteworthy that in Raz’s example the reason that is outweighed is that his garden needs to be weeded. One might have thought that this was a paradigmatic example of a mere enticing reason—a reason that doesn’t get us to an ought, even if there are no competing concerns.^{Footnote 27} If there are enticing reasons, and the weedy quality of one’s garden is one of them, then it should not be surprising that there is nothing wrong in being oblivious to this reason. Indeed, the strength of Raz’s example may depend on him having selected a reason that isn’t of the right kind to yield an ought.

The third observation simply doesn’t support the claim that reasons only require conformity, and not compliance. Raz’s supposition seems to be that compliance with the reasons not to murder would require something like the thought “I could kill now, and I won’t because there are powerful reasons not to kill.” But, it is unclear why this should be so. We don’t consider killing, because, for the most part, there are no reasons to kill. And there is little point in attending to the reasons not to do something when there is no reason to do it. Where killing becomes a salient possibility, we expect compliance, and not mere conformity. That is, if the possibility of killing is put on the table by another, we expect revulsion at the very idea, and not mere accidental conformity. It is unclear why we can’t see this entire disposition—where one doesn’t consider murder at all unless it is presented as a salient possibility—as a way of acting in compliance with reasons. Killing just isn’t a salient possibility for the most part, and hence doesn’t figure in the psychology of most people most of the time. But, where it becomes salient, one does display a kind of moral defect if one merely conforms to the reasons not to kill.

c. Why a parallel problem doesn’t arise with outweighed reasons

A defender of exclusionary reasons may protest that competing reasons, especially important moral ones, also put a virtuous agent in the position of recognizing something as a reason, and hence manifesting the affective responses associated with virtue, yet acting contrary to those dispositions. Hence, this tension can provide no basis for objecting to exclusionary reasons, unless one is also willing to object to the existence of competing reasons. I’m not so sure. There are at least two reasons for thinking that the tension between the virtuous agent’s motivational dispositions and the act that she ought to perform isn’t parallel as between competing reasons, and exclusionary reasons. First, competing first-order reasons compete over our actions, but not over our motivations. And there is a perfectly natural way of accommodating the sense in which one can motivationally give both reasons their due, while acting only as the strongest reason requires. That is just the point about regret. But it is hard to understand how one can give the first order reasons their due, while acting contrary to weight of such reasons. And the mystery is deepened when we remember that second order reasons don’t favor or disfavor action at all. That is, they do nothing to show a course of action in a positive or negative light.

Second, paradigmatic examples of second-order reasons play their putative normative role independently of their content. That is, the putative exclusionary force of an order, or an authoritative directive, obtains independently of what was ordered or directed. So, if there are exclusionary reasons, there is something that must figure in the motivation of an agent so as to overcome first order reasons, but, since the exclusionary reasons are content neutral, the exclusionary reasons, qua exclusionary, do nothing to favor contrary action. Being content neutral, they cannot be said to characterize anything that could figure into a conception of the good. In contrast, competing first order reasons do figure into some conception of the good. So, if we accept the virtue theoretic account of moral motivation according to which the possession of virtue is a matter of having a certain motivational complex which gives reasons their due, we can account for any tension arising out of competing reasons in terms of how those considerations figure into a conception of the good. We can tell a story about how competing goods are arranged in different contexts, and what an appropriate response might be in the face of competing considerations—when to feel regret, say. Not so with exclusionary reasons, as there is no content to them, qua exclusionary, that might ground such a story.^{Footnote 28}

d. Why we don’t need exclusionary reasons to account for the range of commonplace judgements about agents who defy orders or act contrary to the law

I’ve argued that a conception of virtue as involving responsiveness to reasons is at odds with the role Raz attributes to exclusionary reasons. And I’ve argued, contra Raz, that reasons are not merely candidate guides; salient moral reasons have a grip on the mind of a virtuous person and motivate action. All pressing moral reasons, including excluded ones, would motivate a virtuous agent in accordance with their weight. But if an agent ought not to act on excluded reasons, then a virtuous agent would be motivated to do something she ought not to do. This is at odds with the virtue theoretic idea that virtue sets the standard for right action.

In closing out this section, it is worth considering one final line of argument Raz offers for exclusionary reasons. With regard to a person who acts contrary to the weight of first order reasons because she takes herself to be confronted with a valid exclusionary reason, Raz claims that we are conflicted.^{Footnote 29} And he suggests that the existence of exclusionary reasons best explains this sense of conflict. As Raz puts it, we experience a “peculiar feeling of unease, which will show itself when we wish to censure a person who acted on the balance of reasons for disregarding the exclusionary reason and when we have to justify someone’s acting on an exclusionary reason against claims that the person should have acted on the balance of reasons.”^{Footnote 30} And, further, he claims that as compared to a person who does wrong because she acts on a reason that is outweighed, we are more ambivalent about our condemnation of a person who acts on excluded reasons.^{Footnote 31} This unease suggests that exclusionary reasons play an explanatory role in capturing certain features of common judgments about praiseworthiness and blameworthiness.

In this vein, let us consider the moral worth of a person’s act when, because of a putative exclusionary reason, she acts contrary to the weight of first-order reasons. I submit that in such cases, our responses are not univocal. That is, we don’t always feel a sense of conflict or unease about such cases. Indeed, at least some of the time, the person who disobeys an order or violates the law is more praiseworthy for doing so precisely because her having done so comes at great potential cost. Such actions are often held up as paradigmatic of courage. These persons exhibit a deeper concern for morality, because the potential personal repercussions for disobedience are so great. But it should be admitted that it isn’t always the case that one who violates an order or the law is more praiseworthy. Whether the person is more praiseworthy will depend on the facts of the case. In some cases we might say that a person who violates an order or a law exhibits a kind of hubris, even when she does so for moral reasons. But, by the same token, we might also say of a person who simply follows the order in the face of competing moral concerns that he exhibits a kind of lack of judgment, or moral blindness, or something similar. So, the question is how to account for this diverse range of judgments regarding the moral praiseworthiness or blameworthiness of an agent, and whether any such account requires exclusionary reasons. I submit that in we can capture the relevant intuitions in such cases without positing exclusionary reasons.^{Footnote 32}

One possible way to capture these intuitions is to understand putative exclusionary reasons—orders or authoritative directives—in terms of what Don Regan calls indicator rules.^{Footnote 33} Roughly an indicator rule is a rule that indicates the layout of first order reasons, but is not itself an intrinsic reason for action, or for acting or not acting for other first order reasons.^{Footnote 34} Indicator rules need not pick out first order reasons with exact precision. But, nevertheless, such rules usually should be followed, at least under normal circumstances. As Regan explains, the rule “professors should not have sexual relations with students” is a good rule to follow. It doesn’t capture anything of intrinsic moral import, but it does generally track considerations of intrinsic moral import—e.g., power differentials, and the need to be a good mentor to students, say. And we would most often do best to follow the rule rather than to try to assess the intrinsic merits of our acts themselves, since there are good reasons to believe that we won’t be especially good at doing so. Regan’s indicator rules fit well with some of the observations from the preceding paragraph. Roughly, we tend to blame those who violate the law, or who fail to follow orders precisely when we think they overestimate their ability to assess the intrinsic moral merits of their acts, but not otherwise. In cases where a person stands up to an obvious injustice, contrary law or orders don’t undermine our moral esteem for such acts. This is so because in such cases we take the agent to be justified in her decision to act contrary to the order. The idea, then, is that whether we experience unease in such cases will depend on whether a person is epistemically justified in disregarding the order or the law. Where they are, we hold them in higher regard, but where they aren’t we hold them in lower regard because their act indicates a kind of hubris. And, importantly, all of this can be captured in terms of the normativity of first-order moral reasons. Indicator rules are decision procedures that aim at getting first order reasons correct, and as Regan points out, one can revise one’s indicator rules in the face of evidence that a better decision procedure is available.^{Footnote 35} And one is free to ignore an indicator rule when it is clear that it leads to the wrong result in some case. All of this requires an agent to continue to assess and act on first order reasons. Indicator rules do not exclude such reasons. Indeed, when one adopts an indicator rule—a general rule against sexual relations with students, say—one does so precisely because of the relevant first order reasons. An indicator rule, if legitimately adopted, should represent the agent’s best effort to give first order reasons their due, in the face of recognized epistemic challenges. Consider an agent who follows an indicator rule because of limited epistemic access to the facts that constitute what she already acknowledges as a reason for action, for instance, that power differentials should be avoided in sexual relationships. In such a case, the agent isn’t excluding acting based on her assessment of first order-reasons. Indeed, it is precisely the assessment that power differentials are reason-giving that leads the agent to follow the indicator rule. Following the rule is merely a means to achieving what, at the first order level, the agent properly judges she has reason to do.

a. Putative second order reasons can be explained solely in terms of first-order normative force

Raz takes it to be obvious that there can be reasons not to act for reasons. On this point, I agree. However, Raz also suggests that all reasons against acting for a reason function as exclusionary reasons.^{Footnote 37} This latter claim is false. In this subsection I argue, first, that not all reasons to not act for a reason—call them putative second order reasons—are exclusionary.^{Footnote 38} What separates exclusionary reasons from non-exclusionary reasons to not act for reasons is the trumping role that Raz ascribes to exclusionary reasons. As Raz puts it, conflicts between first order reasons and exclusionary reasons “are resolved not by the strength of the competing reasons but by a general principle of practical reasoning which determines that exclusionary reasons always prevail, when in conflict with first-order reasons.”^{Footnote 39} However, many reasons not to act for reasons are simply workaday first order reasons to be weighed against competing first-order reasons. And hence, not all reasons to not act for reasons are exclusionary in the relevant sense. Second, I argue that once we see our way to accounting for some reasons not to act for reasons in terms of first order concepts like weight, we have the tools to account for all reasons of that kind in first order terms. To put the point another way, I grant that there are reasons not to act for reasons. But I argue that where there are such reasons, we need not account for them in terms of second-order exclusionary reasons.^{Footnote 40} Rather, we can better understand them as just first order reasons for doing a particular kind of action—acting with a particular motive. This will require some unpacking.

Consider first, that putative second-order reasons are just reasons to act with certain motivations. That is, they stand in relation to an agent’s motives—to what an agent takes as reason giving—not to the underlying first-order normative reason. This is what distinguishes exclusionary reasons from disabling conditions. Disabling conditions make a consideration that would otherwise be a first order normative reason not a first order normative reason. Exclusionary reasons, in contrast, leave the first order normative force of a reason untouched. They simply require that an agent not act for that reason. To see this point, consider the following: Suppose there is a hate crime statute that holds that if a person assaults another for reasons of racial animosity, that person is subject to a penalty of an additional five years in prison. It follows from Raz’s view that the penalty is an exclusionary reason insofar as it a reason not to act for a reason. But there is no first order normative reason to act on the basis of racial animosity. Hence, in this case, it would be a mistake to say that what is excluded is a first-order normative reason. What is excluded is a motivational reason—acting for reasons of racial animosity. So far, I think Raz accepts all of this; putative second-order reasons bear on an agent’s motives and they leave the underlying first-order normativity untouched. So, properly understood, the putative second order reason is just a reason to do a particular kind of thing—avoid acting with a certain motive.

This brings me to the next point. Having a reason to act with a certain mental state is not different from any other reason to perform an action in a particular way, with regards the normative force of the reason provided. One can have reasons to do things thoughtfully, and reasons to avoid doing things angrily. And it is unclear why a move to mental states requires positing a distinct second order normativity. If doing something with a positive disposition will make you enjoy it more, that is a reason to do it with a positive disposition. But if doing the same thing with a negative disposition will help make your warranted contempt more evident, perhaps that is a reason to not have a positive disposition. But again, all of this can be captured in terms of first-order reasons and the related idea of weight. If that is right, then it is hard to grasp why we need to posit a distinctive kind of second-order normative status when we are dealing with acting with a particular kind of mental state, namely, acting for a reason—that is, with a certain motivation. So, at the very least it is a mistake to assume that reasons for acting for a reason are always exclusionary.

So far I’ve argued that at least some reasons for (or against) acting with certain motivations can be construed solely in terms of first-order normative force. Nothing about the possibility of reasons for having certain motivations requires two kinds of normativity—the normativity associated with weight, and the normativity associated with exclusion. To drive the point home, consider the following example: John is told by God that if he will perform act X out of love for his wife, God will eliminate childhood cancer, and if he does it for any other reason, She won’t. Surely John has a powerful reason to perform X out of love for his wife, and not for any other reason. But, what is that reason? It’s just the first order reason that doing so will cure childhood cancer. And, importantly, whether John ought to act out of love for his wife will depend on how that reason relates to other competing reasons. If God* will destroy the heavens and the Earth if John does act X, then there are more powerful reasons not to X, and hence not to X out of love for his wife. But this is all captured in terms of the first-order weighing relationship. John has stronger reasons to not X at all, than to X out of love for his wife. The fact that John has a reason to not act with a certain motivation excludes nothing. In deciding what to do, John must compare the reason to act with a certain mental state against all reasons to do anything else.^{Footnote 41}

So, there is at least some reason to suspect that if there are reasons to act for reasons, they just have first-order normative force. The question I want to pursue in what remains is whether we can capture everything we want solely in terms of first-order normative force. In closing, I consider two possible grounds for thinking that we need the exclusionary relationship, and argue that we can capture the relevant intuitions with first-order normative concepts. Obviously, there may be other grounds for saying that exclusionary reasons are necessary to capture the relevant normative phenomenon. But I hope at the very least to have provided a model for assessing putative examples where exclusionary reasons might be thought to figure in an adequate explanation of the normative landscape.

b. Can First-Order Reasons Accommodate the Phenomenology?

One might insist that exclusionary reasons capture something important that can’t be captured in terms of first order reasons and the outweighing and disabling relationships—namely, that there are certain social roles where we want to say that a person making a decision should not act for certain reasons. For instance, a judge rendering a decision should not favor one party over another because of a familial relationship with that party, or because he has a friendship with that person. And, the thought goes, this can’t be captured in terms of outweighing, because the competing reason may be more powerful—e.g., familial duty might be more important than the outcome of a relatively minor contract dispute, say. So, a person who rejects exclusionary reasons is forced to say that these reasons are disabled if they want to preserve the idea that judges ought not to act for those sorts of reasons. That is, they will have to say that while familial duty might be a reason for many things, it just isn’t a reason for a judge to favor one party over another in a lawsuit. But, the objection continues, this result is unsatisfying because exclusionary reasons can explain why it is intelligible to us when a person feels the pull of a consideration that runs counter to what the exclusionary reason would require (or, indeed, acts contrary to what the exclusionary reason would require). Namely, there still is a first order moral reason to favor one’s daughter, say. And, the thought goes, this is something that can’t be captured if the reason is disabled.^{Footnote 42} After all, a disabled reason is a non-reason.

However, we should not confuse acting intelligibly and acting for a normative reason. One may act intelligibly even when one fails to act for a normative reason. For an act to be intelligible, we just have to see it as responding to something that is normatively salient. It would be unintelligible, say, if someone were to give money to Oxfam because the sky was blue. But often a person responds to something that is morally salient, and thus makes it the case that the act is intelligible, even if the salient consideration is not a reason. To illustrate, consider a judge who (unbeknownst to others) is deciding a case where his daughter is the defendant. She is clearly liable, let’s suppose. However, the judge grants summary judgment in her favor out of a feeling of familial loyalty. My suspicion is that most will think that the act in this case is even worse in virtue of being a kind of nepotism. Nepotism in judging is a wrong-making feature, not a right-making feature. But, nevertheless, the father’s behavior is intelligible to us. We get it, but not because we think that the status of the plaintiff as his daughter is a reason to render a verdict in her favor. Indeed, we think it counts strongly against rendering a verdict at all. We get it because we think the motivation to want to aid one’s daughter is very often commendable (and, frankly, to be expected), it just isn’t commendable here: It isn’t a reason to decide for his daughter. What does this example show? It shows an act can be normatively intelligible even if it is motivated by a reason that is normatively disabled. Remember, the fact that the defendant is the judge’s daughter counts against his granting summary judgment for her. It is a wrong-making feature of his act, and hence not a reason to do it. The point, here, is that to render an action intelligible, all that is required is that the act be related to something that has a kind of general moral salience. It need not be the case that the salience takes the form of being a reason in favor of the act performed. Indeed, disabled reasons will often have this character. We get their general normative relevance. That is why it is worth pointing out that they have been disabled; they are candidates for our consideration. We would not say, for instance, that the blueness of the sky had been disabled as a reason. It just isn’t even a candidate reason (for giving money to OxFam, anyway). A person who never bluffed when playing poker out of a sense of honesty is misguided, to be sure. She treats as a reason something that isn’t a reason in the context. But her behavior is intelligible to us. We can imagine, indeed, her having a preference for avoiding games that involve misleading others. But this is intelligible because of the general normative salience of honesty, not because honesty is a reason when deciding whether to bluff.^{Footnote 43} If this sort of analysis iterates to other cases where we might be tempted to say that a person—because of their social role—should not act for certain reasons, there is a way to capture the relevant phenomenon without having to posit exclusionary reasons.

c. Who Decides?

Next, an exclusionary reasons theorist might point to the common thought that the questions “what should I do” and “who decides what I should do” can come apart. And they might contend that we need exclusionary reasons to capture some aspects of the “who decides” question. For instance, a judge may say of a statute that it is morally awful, but unfortunately it is the law, and proceed to apply the morally awful statute. And, the thought goes, there is a good explanation of this—namely that it is for the legislature, and not for the judge, to decide what the rules governing our society should be. In the face of examples like this the exclusionary reasons theorist might say that we need the category of exclusionary reasons to explain this phenomenon. After all, we need to capture both the sense in which the statute is judged to be morally awful—that’s the first order assessment—and the claim that it is not for the judge to decide—that’s the exclusionary bit. But, this too can be captured in terms of first order reasons. The question of who decides arises not just in the context of law, but also in other contexts as well. Indeed, anyone who is concerned to respect the autonomy of others, in some sense, thinks that there are decisions that should be left to another person. This is what animates our concern about paternalistic action. We recognize that people will sometimes make bad decisions, but we think that those decisions are for them to make. And the reason why is that we respect their autonomy. However, all of this can be captured in first order terms. When a friend makes a bad choice we may refuse to stop them in many cases, but in many others we may intervene. While there may be much disagreement about when, whether, or how to intervene, I submit that when we do it is a very familiar first order moral decision. It is not as though we have a well-defined zone of reasons for intervening that are excluded, but that outside that zone we are free to intervene.^{Footnote 44} Rather, we ask whether respect for the other person’s autonomy is overcome by the harm she is doing to herself. And it is important to note that this isn’t an all or nothing decision. We may, in the face of a friend who is starting to make bad decisions voice our concerns. And we may refuse to help him financially, and at the limit, if things are bad enough, we might intervene more forcefully, or even have the person committed. But all of this makes sense in light of the first order balancing of two competing concerns—the individuals’ autonomy on the one hand, and their well-being on the other.

Similarly, a court’s behavior can also be best assessed as balancing a number of competing concerns. I will focus on just three. There is a concern for a kind of social stability and predictability characterized by the rule of law. There is a concern, at least in democracies, for respecting the autonomy of the people expressed through their legislatures. But, there is also a concern for substantive justice. The first two concerns can look like exclusionary reasons. A judge might say in the face of a prior precedent that it is not for her to decide, and in the face of a statute that it is not for her to contradict the will of the legislature. But what this comes to is just the judgment that these sorts of things matter, morally speaking—indeed, they may matter a great deal. However, in the face of a law that is morally noxious, the Fugitive Slave Act, say, we often find reasoning that looks very similar to the paternalism case mentioned earlier—or at least we should hope to so find. At some point, a judge may narrow the reading of the statute or the precedent in order to try to give justice more of its due. But, if it is deemed noxious enough a judge might simply decide to ignore it, or find a way to deem it unconstitutional. And this is all just perfectly ordinary first order moral reasoning. As with paternalism, there is not a well-defined zone of reasons that are excluded within which the court is constrained. Rather, there are just reasons for or against certain decisions, some of which might outweigh the others.^{Footnote 45}

Of course there are cases and there are cases, and it may be that there are some cases where we do need exclusionary reasons to explain the relevant normative phenomenon. I suspect not, however, and in this section I have hoped to show how cases that might invite an analysis in terms of exclusionary reasons can be captured in terms of first-order normativity. Indeed, given the puzzling features of exclusionary reasons I submit that before assuming that there are exclusionary reasons, we would do well to first ask whether we would be missing anything by giving an analysis in terms of first order reasons. If we can, then we should prefer that account.