The Yale Law Journal


How To Think About Law as Morality: A Comment on Greenberg and Hershovitz

20 Jan 2015

A Response To

The Moral Impact Theory of Law Mark Greenberg 17 Mar 2014
The End of Jurisprudence Scott Hershovitz 15 Jan 2015


In philosophy, we can sometimes hope to make progress just by looking at old issues in new ways. The hope is that we might see familiar facts and controversies differently and understand them better for it. In their recent Essays, Mark Greenberg and Scott Hershovitz make the case for such hope in jurisprudence: they argue that we can see the issues differently and understand them better for it. Greenberg and Hershovitz don’t see things in exactly the same way, of course, but they effectively agree that we should view law as morality.1 In other words, they effectively agree that we should understand legal rights and obligations (and other legal facts) as certain moral rights and obligations (and other moral facts) triggered by the actions of legal and political institutions.2

Greenberg and Hershovitz contend that the law-as-morality framework (as I’ll call it) is more natural and familiar than it first seems. To understand this framework, we can start with the familiar idea that our particular reasons, rights, and obligations depend on our circumstances. For example, the crises we happen to experience affect the details of our obligations to provide aid; the goals that our spouses set affect what we should do to support their projects; and the promises that people make to us affect the details of our rights to make demands of them. The actions of legal and political institutions—which include, for example, legislative enactments, executive branch decisions, and judicial proceedings—are no different in this respect. These institutions have an almost unmatched power to shape our circumstances, so it’s no surprise that their actions affect our reasons, rights, and obligations. Consider an example that even critics of the law-as-morality framework might accept. The actions of legal and political institutions affect traffic patterns, and traffic patterns affect our moral situation. For instance, given facts about our local traffic patterns, we have moral obligations not to accelerate through most red lights and not to tell children they can safely cross the street as soon as they’ve checked for cars on the right. So the idea that the actions of legal and political institutions affect our moral rights and obligations is familiar enough. Of course we also tend to think that the actions of legal and political institutions affect our legal rights and obligations. Greenberg and Hershovitz do too, but they propose a shift in perspective: they argue that what we ordinarily regard as legal rights and obligations are best understood as a certain subset of the moral rights and obligations affected by the actions of legal and political institutions.3

Greenberg and Hershovitz argue that there are advantages to adopting the law-as-morality framework. For example, both suggest that the theory explains what Dworkin and others call theoretical disagreement about the law.4 Sometimes we disagree about the existence and content of our legal rights and obligations, even when we agree about what’s been said and done by legal institutions in the past. One diagnosis is that we disagree in these cases because we disagree about the legal significance of the historical record.5 For example, think about American constitutional law, where two parties might agree about the way some constitutional clause would have been understood in the eighteenth century, but disagree about whether that fact plays an important role in determining our legal rights and obligations. The law-as-morality framework seems to offer a straightforward explanation: if legal facts are moral facts, then it should be no surprise that people can and do have theoretical disagreements. After all, we often disagree about the existence and content of our moral rights, even when we agree about what’s been said and done in the past.6

Greenberg and Hershovitz have much more to say about how and why to understand legal facts as moral facts, and I strongly recommend their Essays to anyone interested in jurisprudence. In this Response, however, I will focus on just one problem the law-as-morality framework must address. The problem, in short, is to explain the distinction that we ordinarily draw between legal facts and moral facts. We take ourselves to have moral rights and legal rights, moral obligations and legal obligations. But Greenberg and Hershovitz hold that legal rights and obligations simply are moral rights and obligations, and this seems to put them at odds with common sense. The challenge is to explain the common-sense distinction we draw between legal facts and moral facts without giving up the law-as-morality framework.

Consider an example that makes the problem more vivid. Imagine that the local legislature votes to implement a new traffic code and instructs other agencies to post and enforce lower speed limits. These actions will affect people’s moral rights and obligations, but the effects will be diverse: they might include an obligation to drive more slowly; but they might also include an obligation to leave earlier for appointments or perhaps, in the extreme, to find work closer to home (if, for example, the longer commute interferes with important familial obligations). The basic law-as-morality framework holds that legal rights and obligations are the moral rights and obligations affected by the actions of legal and political institutions.7So the basic law-as-morality framework seems to recommend that we classify all the obligations resulting from the legislature’s actions as legal obligations. That won’t do, because it recommends conclusions (for example, that we would have a legal obligation to leave earlier for appointments) that we confidently reject. The problem for Greenberg and Hershovitz is to explain what distinguishes moral obligations that are legal from other moral obligations so that the law-as-morality framework recommends conclusions that better fit and explain our confident judgments about the legal facts.

Both Greenberg and Hershovitz attempt to meet the challenge. In this Response, I will explain and evaluate their attempts. In the next Part, I will explain Greenberg’s attempt to solve the problem. Greenberg’s proposal, in short, is to explain the conditions that set moral obligations that are legal apart from moral obligations that aren’t. In effect, Greenberg characterizes a legal domain of morality that is meant to fit and explain our confident judgments about which sorts of rights and obligations are legal and which are “merely moral,” to adopt Hershovitz’s phrase.8 Greenberg’s approach to the challenge seems promising to me, but the specific conditions he defends do not. I will explain why and briefly outline an alternative that seems more promising. In the final Part, I will explain Hershovitz’s response to the challenge. Hershovitz’s response is similar to Greenberg’s, but it comes with reservations. Like Greenberg, Hershovitz argues that we can explain our tendencies to regard certain moral rights and obligations as legal, and he employs similar strategies to explain particular cases. But Hershovitz has doubts about Greenberg’s attempt to answer the challenge by characterizing the legal domain of morality—that is, by identifying the conditions that make moral obligations legal. I will explain why I think Hershovitz’s doubts are premature.

I. greenberg’s moral impact theory

In this Part, I explain and evaluate Greenberg’s account of the difference between legal obligations and other moral obligations. First, I explain Greenberg’s attempt to distinguish legal obligations by their moral force and their source in our legal practices.9 Second, I explain why I don’t think Greenberg’s account will work. Third, I outline a different account of the difference between legal obligations and other moral obligations that seems more promising.

A. Greenberg’s Account of Legal Obligation

Greenberg’s account starts with the basic view I described above: “legal obligations,” he writes, “are those moral obligations created by the actions of legal institutions.”10 But he adjusts his initial characterization in two ways.

First, Greenberg attempts to distinguish legal facts by their moral force: he suggests that moral facts are legal facts only if they are all-things-considered moral facts. Call this Greenberg’s “force condition.” He writes: “The relevant obligations—the ones that, according to my theory, are legal obligations—are simply genuine, all-things-considered, practical obligations” created by the actions of legal institutions.11 On Greenberg’s view, “an all-things-considered obligation is one that, taking all relevant considerations into account, one should fulfill.”12 The upshot is that moral obligations are legal obligations, on Greenberg’s view, only if they are decisive.13

Second, Greenberg attempts to distinguish legal facts by their origin in the actions of legal and political institutions. Call this Greenberg’s “source condition.” The source condition has two parts, which he includes in response to distinct concerns about the basic view.

Greenberg’s first concern is that his account will include too few legal obligations. In particular, Greenberg’s concern is that the basic view can’t explain legal obligations with the same content as certain moral obligations (for example, not to harm or kill), since the relevant moral obligations pre-exist the actions of legal and political institutions, and hence aren’t created by them.14 The worry, in other words, is that Greenberg’s account would “have the consequence that some of what we take to be paradigmatic legal obligations, such as the obligation not to kill, are not legal obligations at all.”15 In response to this worry, Greenberg relaxes the required connection between the actions of legal and political institutions, on the one hand, and the resulting moral facts, on the other. It’s enough, Greenberg argues, if our moral obligations are affected by our community’s legal and political history.16 For instance, we have preexisting moral obligations not to drive impaired, but the content of these obligations is somewhat imprecise. Greenberg’s thought is that the actions of a community’s legal and political institutions might make the content of these preexisting obligations more precise. Suppose, for instance, that the legislature, after careful study, decides to prohibit anyone from driving with a blood alcohol level above 0.8%. The legislature’s decision might affect the content of our preexisting obligations, in the sense that its decision triggers a more precise obligation that is explained, in part, by the preexisting obligations.17 Because the legislature’s decision has a relevant moral effect, Greenberg suggests, we can classify the resulting obligation as legal.

Greenberg’s second concern is that his account will include far too many legal obligations. His basic account regards all the moral obligations affected by the actions of legal and political institutions as legal obligations.18 But this sweeps in too much. Greenberg focuses his attention on what he calls “paradoxical” moral obligations generated by the actions of legal and political institutions, which might include obligations to vote the bums out, to protest, to resist, or perhaps to simply flee.19 But Greenberg must also consider what I will call “incidental” moral obligations generated by the actions of legal and political institutions, though I do not mean to suggest these moral obligations will be unimportant. For example, the class of incidental moral obligations will include the obligation to leave earlier for appointments that we identified when we imagined the local legislature acting to implement a new traffic code.20 But the category includes much else. Greenberg considers what happens when the legislature merely suggests a solution to a coordination problem that, by making the solution salient, leads to a moral obligation to adopt it.21 And Hershovitz asks what we should think when the legislature enacts whistleblower protection statutes that, by providing a new option (to blow the whistle with legal protection), make coming forward with information obligatory, not supererogatory.22 In response to these concerns, Greenberg argues that legal obligations aren’t just any moral obligations created by the actions of legal and political institutions: moral obligations are legal obligations only if they are created by the actions of legal and political institutions “in the legally proper way.”23 Greenberg’s thought is that paradoxical and incidental moral obligation don’t trace to the actions of legal institutions in the proper way, and therefore aren’t legal obligations.24 For example, he argues that, by their nature, legal systems are supposed to improve the moral situation—given this, when the actions of legal institutions make the moral situation worse, the resulting obligations to fix the moral situation won’t be legal obligations.25 Greenberg acknowledges that this condition won’t cover all the cases, and he regards the account as a work in progress. 26 But Greenberg seems to put its finger on the kind of account we need. The intuitively correct response to cases of paradoxical and incidental obligations does seem to be: sure, those are moral consequences of legal action, but not the right sort.

In the end, then, Greenberg’s considered view is as follows: a community’s legal facts are just those all-things-considered moral facts affected by the actions of its legal and political institutions in the legally proper way.27 These conditions characterize a set of moral facts, which we can think of as a legal domain of morality—if Greenberg is right, it’s the domain of morality that fits our confident judgments of the legal facts and explains what they are judgments of.

B. Some Doubts About Greenberg’s Account

In this section, I will offer reasons to think Greenberg’s account of legal obligation won’t work. I will focus on Greenberg’s force condition, but I will voice some concerns about the source condition too.

The problem with Greenberg’s force condition is straightforward: this condition seems to imply that we have many fewer legal obligations than we ordinarily think, or that the legal obligations we have are much stronger than we ordinarily think.28 We ordinarily think that legal obligations can conflict with other obligations. Greenberg’s view doesn’t rule out such conflicts, at least not entirely. He allows that we can face conflicts between legal obligations and other pro tanto obligations. But part of what we ordinarily think, I take it, is that we can face conflicts between legal obligations and other obligations where the right thing to do is to fulfill our non-legal obligations. (Or more broadly, we can face situations where the right thing to do is something other than what we’re legally obligated to do.) Greenberg’s view rules this out. He holds that legal obligations are all-things-considered moral obligations, and that all-things-considered moral obligations settle the question of what to do. It can only appear that the thing to do is to violate a legal obligation: in reality, either we have no legal obligation, or the thing to do is to fulfill it.

Does this conflict with common sense? It certainly seems to. Suppose that you are the only coach of your child’s soccer team, and you have a game this afternoon. You left for your team’s game at a responsible time, but ran into terrible traffic. You arrive with just minutes to spare and frantically look for a parking space. You find what seems like an open space but a city sign clearly says: “No parking.” You draw the conclusion we all draw: that you have a legal obligation not to park there. You take everything into account: is parking there the best thing to do, despite your legal obligation not to? I’m inclined to think it is, though of course you’re on the hook for a ticket, perhaps worse.29 And this is a case where the moral stakes are relatively low. We can imagine cases involving more serious moral concerns. Suppose, for example, that you must trespass to reach someone drowning in a private pond; or that your cancer-stricken mother asks you to buy her cannabis in order to deal with the nausea caused by chemotherapy; or even that you must violate one legal obligation (for example, to pay the credit card bill) in order to fulfill another (for example, to pay the gas bill so your children have a warm place to sleep). It seems clear that in these cases—or in some cases like them—we face situations where we should violate at least one of our legal obligations in the name of competing moral obligations or concerns. Greenberg’s view doesn’t let him describe the cases this way: he must say that our moral responses are misguided, or that these aren’t cases where we have the relevant legal obligations at all.30

And particular cases aren’t the only problem for Greenberg’s force condition. This condition would also seem to require us to revise the way we think about legal obligation in general. Consider two quick examples. First, the force condition seems to imply that we are morally required to obey the law, since it holds that our legal obligations simply are all-things-considered moral obligations, and we are morally required (I assume) to obey our all-things-considered moral obligations. If so, then Greenberg’s force condition either puts him at odds with the widely held view that there is no general moral obligation to obey the law, or it requires him to understand the question in a different way (for example, to interpret the question as whether there is a general moral obligation to obey what the legislature puts forward as law).31 Second, and in a similar way, the force condition seems incompatible with the most straightforward account of civil disobedience—after all, if civil disobedients are right about their reasons, then they might have no genuine legal obligations to violate.32

My point is not that Greenberg has nothing to say in response, or that his responses could not be plausible. My point is only that Greenberg’s force condition recommends judgments that we intuitively reject, and this comes at some cost to his account of legal obligations, especially since the charge is that the law-as-morality framework fails to fit and explain what we ordinarily think.33 Suppose for now that Greenberg should drop the force condition. Where would that leave his account of legal obligation? The most straightforward response would be to drop the force condition and keep the source condition. The resulting account would be that legal obligations are the moral obligations (whether pro tanto or all-things-considered) affected by the actions of legal and political institutions in the legally proper way. In fact, Greenberg already indicates that this is his fallback position.34

I find this view more promising, but I have two related concerns. First, I worry that the source condition doesn’t really explain what legal obligations have in common or what distinguishes them from other moral obligations. (At worst, the source condition comes closer to naming the problem than solving it.) This charge isn’t completely fair, of course, because Greenberg stresses that the account is a work in progress, and he might yet show that our questions and disputes about legal obligations are best construed as questions and disputes about moral obligations that have arisen in the proper way, whatever that turns out to be. So I won’t press the point about the source condition as it now stands. Second, and more important, however, I worry that the filled-in account will be lacking in a similar way. Consider some cases we’ve seen that an adequate source condition will need to sort and explain: an obligation to vote the bums out or protest government action; an obligation, given the new traffic patterns, to leave earlier for appointments or to live closer to work or to reinstruct your children on how to safely navigate the streets; an obligation, given that you have legal protection, to blow the whistle on your employer; and an obligation, given the state’s recommendation and the way others have responded to it, to adopt a solution to some coordination problem. These cases hint at the diverse ways that the actions of legal institutions can result in obligations that clearly aren’t legal. And it’s not obvious to me that the reason they aren’t legal obligations is that they didn’t arise in the proper way, unless that just means that they didn’t arise in a way that makes them legal obligations. But suppose I waive that concern, and grant that, in each case, the relevant obligation didn’t trace to the actions of legal institutions in the right way. Even then, I worry that it will be hard to develop criteria of legal properness that hang together in a theoretically appealing way.35

I don’t want to lean too hard on these criticisms because I’m not yet certain they’ll bear the weight. But I turn now to putting a different sort of pressure on Greenberg to defend his account. My strategy is to outline a different way to think about legal obligation within the law-as-morality framework, one that seems more promising.

C. Another Way To Think About Legal Obligations

The view that I will outline in this section holds that what sets legal obligations apart from other moral obligations are the distinctive liabilities we incur for violating them. To a first approximation, the view is that legal obligations are moral obligations that legal institutions have the moral standing to hold us to. I will call this the “legal liability condition,” but keep in mind that it refers to a moral (that is, genuine normative) liability to certain responses from legal institutions. To see the condition’s appeal, think again about a local legislature’s decision to adopt and enforce a new traffic code. The legal liability condition promises to capture one key difference between the resulting obligation to observe the posted speed and the resulting obligation to leave earlier for appointments. The local legal institutions—through the police—have the moral standing, or authority, to hold us to the posted speed, but not to our lunch plans. That, anyway, is the basic idea. In the rest of the section, I will explain this approach to the law-as-morality framework in somewhat more detail. If the view is as promising as it seems, then the challenge to Greenberg is sharper: he must not only show that the force and source conditions are defensible, but also that they are superior the legal liability condition.

Let’s take one step back. Our legal practice—the practice of making, interpreting, disputing, and implementing the law—involves questions and competing claims about the exact distribution of legal obligations, rights, powers, and immunities. In other words, a central part of the practice involves moral notions like “right” and “obligation.” In a way, Greenberg’s insight—shared by other law-as-morality proponents—is to take this part of legal practice at face value, or near enough. Greenberg argues, in effect, that legal rights and obligations could in fact be what they seem: moral rights and obligations with a special connection to and relevance for legal practice. To make the case, Greenberg stresses the source and force of some of our moral obligations. These don’t seem to be the right points to stress, for the reasons I’ve given. But Greenberg might have tried to make the case by stressing a point about the moral notions involved. One standard view (though not the only possible one) about notions like “right” and “obligation,” but especially about “power” and “immunity,” is that they are fundamentally relational: they refer to constituent parts of the moral relationships individuals stand in to one another.36 To fully characterize the moral facts, on this sort of view, we would have to characterize the complex web of moral relationships that individuals stand in to one another.37 So if legal facts simply were certain moral facts, as we’re supposing, then to fully characterize the legal facts, we would have to characterize part of the complex web of moral relationships that individuals stand in to one another.

This suggests an alternative approach to characterizing a legal domain of morality. Legal facts are moral facts, we might say, but they are primarily distinguished from other moral facts by the particular sort of moral relationships in which they figure, not by their practical force or their source in our legal practice. Here’s a working hypothesis: legal facts are moral facts that figure in the moral relationship between a community’s legal and political institutions and the community’s individual members. 38 Let’s think about how this hypothesis might apply to legal obligation. One way to understand obligation is by its connection to accountability or answerability.39 Obligatory actions are those we are accountable to others (including ourselves and perhaps everyone) for performing, and this sets them apart from actions that are only recommended (no matter how strongly). We might distinguish our legal obligations, then, as those moral obligations that legal institutions have the distinctive standing to hold us to, a standing that characteristically involves the standing to impose sanctions.40

So far, this puts the view in terms of the characteristic powers (that is, the standing to hold to account) that figure in the moral relationship between a community’s legal and political institutions and its members. We can also put it in terms of the characteristic liabilities involved. That is, we can say that legal obligations are distinguished from other obligations by the liabilities we incur for failing to fulfill them: specifically, they make us liable to certain responses from the legal institutions themselves. That is the crux of what I called the legal liability condition. Roughly put, it holds that the legal domain of morality is defined by the standards that we are answerable to our community’s legal institutions for meeting. I don’t mean to suggest that the legal liability condition will be easy to spell out—I don’t think it will be.41 But I think the general approach has some promising features.

First, the legal liability condition doesn’t imply Greenberg’s thesis that legal obligations must be decisive. On the legal liability approach, the question of my having an obligation to you is a moral question, to be sure, but a question about the structure of our moral relationship. Roughly put: do you have the standing to hold me accountable for acting a certain way? The question of what moral force this fact has is a further moral question. So this approach to distinguishing law from morality would enable us to separate the question of what the legal facts are from the question of their force.42 Given this, the legal liability condition promises to explain cases involving moral emergencies in a way that better fits with what we ordinarily think about them. Consider, for example, the case where I’m late for the youth soccer game I’ve promised to coach and decide to park illegally. I have a legal obligation not to park there, in the sense that the police have the standing to demand that I don’t and to ticket me if I do. But this doesn’t automatically exclude my having enough reason to park there.43

Second, the legal liability condition promises to explain the phenomena Greenberg puts forward to motivate the claim that legal obligations must arise “in the legally proper way.”44 The first of these problems, recall, involves legal obligations that seem redundant because their content coincides with the content of preexisting moral obligations. The legal liability condition can accommodate our intuition that the actions of legal and political institutions aren’t idle: they succeed in generating non-redundant content, because they succeed in generating non-redundant moral liabilities. This might sound strange. But it’s a familiar enough idea that we can make ourselves newly accountable for acting in ways that we are already accountable for acting.45 I might promise my partner that I’ll exercise more often and then later promise my mom the same thing.

The second of these problems involves “paradoxical” and “incidental” moral obligations. The problem, recall, is that the actions of legal and political institutions bring about many moral obligations—including, perhaps, obligations to resist or flee, and obligations to leave earlier for appointments—that couldn’t plausibly be legal obligations. The legal liability condition seems to identify why these sorts of obligations won’t count as legal obligations. The legal obligations, roughly put, are the obligations that legal institutions have the distinctive standing to hold us accountable for fulfilling. Any obligations we have to resist or flee wicked legal institutions won’t have this feature. (I can change your moral obligations by attacking you, but it seems obvious that I won’t have any standing to hold you to the resulting obligations.) And the legal liability also seems to identify why incidental obligations aren’t legal obligations: legislative actions might affect the way I’m obligated to drive and when I’m obligated to leave for appointments, but violating the resulting obligations involves different moral consequences, and that’s how we can distinguish my legal obligation to observe the posted speed from my merely moral obligation to leave early enough to keep my appointments.

My aim with these brief remarks is to offer a glimpse into yet another way to see law as morality, one with evident appeal.46 But one effect of these remarks is to sharpen a potential challenge to Greenberg’s view. A defense of Greenberg needn’t only show that his view scores better than I’ve given it credit for; it also needs to show that it scores better than the alternatives.

II. hershovitz’s alternative

Greenberg and I agree that law is best seen as part of morality, but we disagree about which part. Greenberg proposes that we see law as the binding part of morality triggered by the actions of legal and political institutions in the legally proper way. In response, I have argued that Greenberg’s proposal doesn’t fit what we tend to think about the law. And I have proposed that we might instead see law as the part of morality that courts and other legal institutions have the moral standing to hold us to. My proposal, once spelled out, might prove to be more successful than Greenberg’s, or it might not.

But let’s ask another question: why have this dispute with Greenberg at all? At the outset, I suggested that what justified the effort was the need to make sense of certain familiar ideas within the law-as-morality framework. We ordinarily take there to be something distinct from morality, whether the law or the content of the law, and we ask what it requires or allows. In effect, Greenberg attempts to show that our familiar ideas are correct, if not quite in the way we might have thought. That is, Greenberg attempts to show that the law-as-morality framework can and should vindicate the idea that there is something that answers to our ordinary thoughts about the content of the law. The law-as-morality framework simply holds that the something—the content of the law—is a characteristic subset of morality.

Scott Hershovitz has concerns about developing the law-as-morality framework this way.47Like Greenberg, Hershovitz endorses the central claim of the law-as-morality framework, that questions about what the law requires of us or entitles us to are best understood as questions about the moral effects of the actions of legal and political institutions. But Hershovitz doesn’t think this insight should lead us to refurbish and vindicate the idea that these is something—the law’s content—that’s distinct from morality. Instead, he argues, it should lead us to leave the idea behind entirely.48

Let’s grant that we could leave it behind. After all, Greenberg should agree that we could ask and answer all the underlying moral questions without asking which are best regarded as legal—that is, we could figure out morality’s content without asking which parts are properly included in the content of the law.49 So the question is whether we should leave the idea behind. Hershovitz says yes. He argues that there are reasons to think law-as-morality proponents can’t vindicate the notion of the law’s content, and reasons to doubt they should even if they could.

A. Is Greenberg’s Project Based on a Mistake?

First, Hershovitz argues that the attempt to vindicate the idea of the content of the law by characterizing a legal domain of morality rests on a mistake, because it misunderstands the nature and significance of the distinction we draw between law and morality.50 This attempt assumes that the distinction we draw between legal facts and moral facts is systematic, in the sense that it tracks the same conditions every time it’s drawn correctly.51 But Hershovitz thinks we should reject this assumption, or at least that we have strong reasons to doubt it. He argues that, if you look closely, we often draw different distinctions between law and morality, and for different purposes. So it’s a mistake to try to characterize “the” distinction between law and morality.52

I will return to Hershovitz’s arguments below, but first consider a case where his position seems right. We regard some of our obligations as “work obligations,” but it seems like a mistake to read too much into the distinction between work obligations and other obligations, since we likely draw that distinction in different ways to make different points. In some cases, as when our concern is to keep our jobs, we might distinguish our work obligations from other obligations by the fact that our bosses have the standing to hold us to them. But other concerns might lead us to draw different distinctions. For example, suppose you promise a work friend that you’ll attend the optional holiday party. Your boss won’t regard the resulting obligation as one of your work obligations. But you might regard it as such, especially if you only need to explain the general source of the obligation (for example, because you need to explain to non-work friends why you have to decline their invitation to be elsewhere that night). Then again, you might not (for example, if your non-work friends press you, you might clarify that it’s not something that might get you fired, but something you’re committed to all the same). But the underlying moral facts are the same, and it seems unmotivated to insist that there’s a privileged way to describe them. At the very least, it seems like we need good reason to think there is.

Hershovitz suspects the situation is the same in law: just as we characterize different “work domains” of morality, so too do we characterize different legal domains. In other words, Hershovitz suspects that we use “legal” to stress different sets of obligations for different purposes: sometimes we use the label to stress the source of our obligations, other times to stress the liabilities we face, and still other times to stress the point that these obligations don’t exhaust our moral responsibilities.53 And Hershovitz argues that individuals who occupy different roles regard—and should regard—different sets of rights and obligations as legal.54 For example, we might regard judges as having legal obligations even when they can’t be made to answer for their conduct. When we focus on ordinary citizens, however, we might draw the boundary more strictly and include only those obligations that involve certain liabilities. If Hershovitz is right, then perhaps Greenberg and I simply draw different distinctions to stress different points. But then we don’t (or need not) disagree, since we can agree about how to distinguish legal obligations from moral obligations once we have a concern in mind. And it seems unmotivated to insist that there’s a privileged concern to have. At the very least, it seems like we need a good reason to think there is.

I think the best response to Hershovitz’s doubts, given the early stage of the debate, is to make the case for cautious optimism. Hershovitz might be right that we distinguish legal obligations from other obligations in a pragmatic and flexible way, and hence he might be right about how to develop the law-as-morality framework, but I don’t think we should give up on the boundary-drawing project too quickly. The basic reason for optimism is that our practice of making and disputing claims of legal obligation is more robust and (looks to be) more systematic than our practice of making and disputing claims of work obligation—in other words, we seem to distinguish legal obligations from other obligations in a more disciplined and principled way.55 Given this, careful attention to the way we ordinarily draw the distinction might yet support Greenberg’s thesis that we’re concerned with a particular class of moral obligations, whose general features we might explain. Then again, careful attention to the data might support Hershovitz’s view. But I don’t think the examples we have seen settle the question either way.56 And it’s worth stressing that Greenberg’s attempt to vindicate the notion of the content of the law might survive a counterexample or two. It could turn out that disputes about legal obligation are best understood as disputes about a particular kind of moral obligation in all but a few cases. In that case, we might succeed in characterizing the central distinction we draw between law and morality, and we might offer partial vindication of our ordinary ways of thinking in the process.57 So I think we should regard the jury as out on the question of whether we can vindicate the notion of the content of the law by characterizing a legal domain of morality.

B. Is Greenberg’s Project Worth Doing?

But let’s turn to Hershovitz’s second argument, for thinking we should move on even if we could vindicate notions like the content of the law. Hershovitz argues, in short, that there aren’t good reasons to pursue this vindicatory law-as-morality framework, and several good reasons not to. Hershovitz doubts there are good reasons to pursue Greenberg’s vindicatory approach because he doubts that concepts like the content of the law play a sufficiently important role in our thinking. Hershovitz writes: “The thought that there is an existing body of law that comprises all the legal rights, obligations, privileges, and powers in force in a legal system plays no role in legal practice.”58 Lawyers, that is, take no special interest in the content of the law: “Rather, they read records of their community’s legal history . . . and then they construct arguments about what obligations people have as a result.”59 Hershovitz’s suggestion, I think, is that there is pressure to vindicate ways of thinking only when they play some valuable role in our practice. The challenge, then, is to explain why notions like the content of the law are sufficiently valuable to incorporate into the law-as-morality framework.

I think this challenge might be met. Here’s one possibility I find tempting, though I won’t pursue it at length here. The thought is that it might be useful to talk about the law’s content because doing so enables us to better track an entire dimension of our moral lives. On Greenberg’s picture, we discover our legal obligations by looking to the facts about what our community’s legal institutions have said and done and figuring out their moral consequences. Greenberg then adds that the content of the law includes all the relevant obligations, rights, powers, and immunities that we discover. Hershovitz wonders what the point of the addition is.60 I think Hershovitz’s complaint would have more bite if we only had to worry about the moral consequences of what legal institutions have done. But of course that’s only part of what we worry about when we worry about our moral situation. We also have to ask how our promises, our families and friendships, our jobs, and our religions contribute to our moral situation. But this still leaves us with questions about what to do in the end, and the outputs of our earlier moral inquiries (for example into our promissory obligations) might serve as inputs to this one. And at this stage, moral categories (for example, the law, the house rules, and so on) serve a useful purpose, both as a way to keep track of the relevant considerations, and as a way to pose the question of whether, say, legal obligations as a category should have any special moral force in deliberating about what to do. A superhuman deliberator might proceed differently: he or she might be able to deliberate about what to do without intermediate steps and without the help of categories like “the law’s content” or “the house rules.” But none of us is superhuman: we deliberate in stages, and general moral categories like the content of the law play an important role in our ordinary moral thinking.

Perhaps these comments point to good enough reasons to maintain and vindicate the notion of the law’s content within the law-as-morality framework—provided, of course, that there aren’t downsides to doing so. But Hershovitz thinks there are downsides to consider. He worries—like Dworkin did before him—that notions like the content of the law are prone to cause confusion in jurisprudence.61And Hershovitz isn’t only concerned about misunderstandings. He’s also concerned about misallocated attention. He worries—and, given the content of this Response, presciently—that jurisprudence will focus on the classificatory questions at the expense of the moral questions (for example, the question about how and why our community’s legal history affects our rights and obligations and powers) that, in his view, belong at the center of jurisprudence.62

Hershovitz might be right about some of these costs, but I think it’s too soon to tell. He and Greenberg have laid out these issues much more clearly than before, and confusion might well be less common and less distracting going forward.63 At the very least, then, it’s premature to give up Greenberg’s vindicatory version of the law-as-morality framework. Even so, I share Hershovitz’s concerns about the potential misallocation of attention in jurisprudence. The moral questions are more important—certainly as a practical matter, but even as philosophical puzzles that are worth solving—and they deserve the lion’s share of the attention. But I’m not sure that the proper response to this point is to give up Greenberg’s project, provided—and this is a big proviso—it can be done, and that we have reasons to do it.


Let’s quickly take stock. This Response has focused on one question: if law is best understood as a part of morality, then what explains the ordinary distinction we draw between law and morality? Greenberg argues that certain criteria sort moral rights and obligations that are legal from those that aren’t. Hershovitz, in contrast, thinks it’s most important to start by understanding how and why we classify some moral obligations as legal. Once we understand this, he argues, we’ll see why Greenberg’s attempt misfires. We don’t classify moral obligations as legal simply because they satisfy certain criteria that set them apart. Instead we classify moral obligations as legal to accomplish different goals. And because our goals vary, so do the moral obligations we classify as legal. In this Response, I have defended Greenberg’s view against Hershovitz’s alternative. I don’t think Hershovitz has offered sufficient reason to reject the principled boundary-drawing project, and I have offered several reasons to think it can and should be done. But I argued that Greenberg’s attempt to draw the boundary fails, and outlined what seems like a more promising approach to the challenge.

In conclusion, it’s worth noting how things look if we take a step back. I have granted much of the law-as-morality framework for the sake of discussion, both because I think it’s on the right track and because doing so enables me to reach the important and interesting choice between Greenberg’s approach and Hershovitz’s. But the law-as-morality framework is quite radical in some respects, and other readers will be less inclined to grant Greenberg and Hershovitz all that I have. So the most important debate—at least in the sociological sense, but perhaps philosophically too—is between those who accept the law-as-morality framework, and those who reject it. And in this debate, Greenberg and Hershovitz are allies with a tall task ahead. But their Essays, taken together, are an impressive start.

Steven Schaus is a Ph.D. Candidate at the University of Michigan; Brown University, A.B., 2006. He is grateful to Johanna Dennehy, Daniel Drucker, Ben Eidelson, Allan Gibbard, Scott Hershovitz, Alex Sarch, Scott Shapiro, and Rohan Sud for helpful comments and conversations. He also benefited from discussing these (and related) ideas with audiences at Harvard University and the University of Michigan. Finally, he would like to thank the editors of the Yale Law Journal—and especially Rachel Bayefsky—for their patient and thoughtful guidance. 

Preferred Citation: Steven Schaus, How To Think About Law as Morality: A Comment on Greenberg and Hershovitz, 124 Yale L.J. F. 224 (2015),