Navigation
Newsletters Index (06:2)
apaOnline Home Page

APA Newsletters

Spring 2007
Volume 06, Number 2

Newsletter on Philosophy and Law

Articles

Law and Political Morality*

Jules L. Coleman
Yale University and Rutgers University

I. **

The conventional story of contemporary jurisprudence begins with Hart’s formulation of legal positivism in The Concept of Law. Both the work and Hart himself have been hugely influential, especially so on three of his students—Ronald Dworkin, John Finnis, and Joseph Raz—each of whom has emerged as the leading figure within different jurisprudential traditions.

Dworkin is the most important and influential critic of Hart and of legal positivism more generally. In addition, he has developed an alternative jurisprudential outlook—interpretivism—that is as insightful, influential, and controversial as are his many objections to legal positivism. Finnis has replaced Fuller as the leading contemporary practitioner of traditional natural law theory, though in his version of it, Hart’s influence looms large; indeed, Finnis’ own views draw inspiration in roughly equal measures from Hart and Aquinas. Still, he is more faithful than Fuller to the natural law tradition, and far more so than is Dworkin who, at least in some circles, is likewise identified with the natural law tradition (which only goes to show that you can’t control what kind of mess others will make of your views). Though his first book contains many important criticisms of The Concept of Law, Raz has emerged as the uncontested heir both to Hart’s positvism and to the positivist mantle more broadly.1

Ironically, perhaps, Dworkin is more responsible than anyone else for the wide currency of Hart’s views as well as his continued influence, which is primarily a function of his sustained, if highly critical, engagement with it. For the most part, Hart has not responded to Dworkin’s many objections to his views. Other positivists have taken up Dworkin’s challenges both to Hart and to positivism more generally. Dworkin’s best known early objection to legal positivism was that it lacks the resources to explain the role of moral principles in legal argument; some moral principles can be binding on officials, that is, impose legal duties and confer legal rights, and legal positivism has no way of explaining how this could be so.

Raz correctly noted that not every standard that is binding on a judge in a particular jurisdiction is, for that reason alone, part of the law of that jurisdiction. So even if moral principles are sometimes legally binding on officials it does not follow that they are binding because they are law. All positivists accept this initial response to Dworkin but disagree as to how the response should proceed from there. Many positivists follow Raz in holding that moral principles cannot be authoritative legal texts. Others adopt a line of argument that I have championed: namely, that moral principles can be authoritative legal standards—provided their status as law depends not on their content but on their source.

Theorists of the first sort are often referred to as "exclusive" legal positivists, whereas those adopting the second line of response are typically called "inclusive" legal positivists. Both versions of positivism have a claim to Hart’s legacy. Some might think that the claim of inclusive legal positivism to Hart’s legacy is strengthened by the fact that Hart himself came to embrace it as the best understanding of his view; others remain steadfast in insisting that Hart did himself no service by keeping such unsavory company.2

This is an incomplete and insufficiently nuanced picture of the jurisprudential landscape. It ignores legal realism, which every now and again finds a serious advocate among philosophers of law while remaining a constant favorite among law professors—whose insight into matters jurisprudential is suspect at best. Brian Leiter is currently the most serious philosophical defender of the realist tradition.3 Nor can one escape the seemingly infinite variety of so-called critical theories, many of which claim a legal realist lineage. This group includes Critical Legal Studies, Critical Race Theory, and Feminist Jurisprudence, to name the most well-known critical theories. Incomplete as it may be, the basic picture remains intact. The central figures dotting the jurisprudential landscape include the two faces of positivism, various forms of natural law theory, and Dworkinian interpretivism, and this picture suffices for our purposes.

Historically, the most important divide among jurisprudential theories is that between legal positivism and natural law. So powerful is the pressure to characterize jurisprudential views in terms of how they relate to legal postivism and natural law that some clearly distinctive views—such as legal realism and Dworkinian interpretivism—are often (and wrongly) pigeonholed into one or the other. Legal realists and legal formalists as well are said to be essentially legal positivists, and much ink has been spilt arguing that Dworkin is either a natural lawyer or a legal positivist—when in fact he is neither.

All this would be harmless were it not terribly misleading, distracting, and, worse, altogether unhelpful. The root of the problem is that the distinction that allegedly marks the difference between legal positivism and natural law—the so-called "separability thesis"—cannot discharge this burden. The most narrow formulation of the separability thesis is given by Hart himself: "it is in no sense a necessary truth that laws reproduce or satisfy certain demands of morality, though in fact they have often done so" (CL 185-186). The problem is that there is likely no plausible interpretation of the separability thesis that positivists endorse and natural lawyers reject.

On one interpretation, the separability thesis asserts that there are no necessary moral constraints on the content of law. So understood, it allows that a rule or norm could be a legal rule even were it immoral: for example, even were it to require conduct prohibited by morality or prohibit conduct required by it. The problem is it is hard to imagine a plausible version of natural law theory that would reject this claim.

On the other hand, some have taken the claim that there are no necessary moral constraints on the content of law to imply that there are no necessary connections between the concepts of law and morality, or that there are no necessary moral properties of law. Natural lawyers are said to reject this claim because they believe that law has necessary moral properties, though they disagree among themselves as to what those properties are and whether the claim that law possesses them is a conceptual truth or an essential feature of law that is not part of our concept of it. In any case, the claim that natural lawyers apparently endorse is that nothing could be law—understood as a form of regulating human conduct—that did not possess morally desirable or valuable properties. In endorsing the separability thesis, legal positivists are presumed to be rejecting this claim. But do they, and, more importantly, must they?

The problem here is twofold. First, it may well be a necessary truth about law that it possesses moral properties, and why would positivism want to find itself in the unenviable position of rejecting an important truth about law? The answer may well be that positivism’s internal logic commits it to rejecting this important truth. But there is, in fact, nothing in legal positivism that would entail its denying that law necessarily has morally desirable features or properties.

Here is a simple argument whose premises are perfectly compatible with legal positivism and whose conclusion is that law has necessary moral properties. It is a necessary property of law that it claims a power to enforce its demands through the collective use of force. It is also a necessary truth about law that it addresses those to whom it issues its demands as agents, that is, as individuals having the capacity to act for reasons. It is by reason, not force, that law regulates our affairs with one another. On the other hand, it claims a power to employ force to compel agents to act on the reasons it provides. Arguably, given that the law is coercive and claims that the reasons it provides are legitimate ones, the fact that it addresses those over whom it claims coercive authority as agents capable of acting for reasons is a morally attractive feature of law, and a necessary one at that.

Nothing really hangs on whether this is a compelling argument. Many positivists, including Raz and myself, take the law to have necessary moral properties. And there is no reason at all why a positivist would want to insist that this could not be so. Positivists insist that there are no necessary moral constraints on the content of law, not that the law has no necessary moral properties. These are two very different kinds of claims. The problem is that no positivist need reject the latter and no natural law need reject the former. Not to put too fine a point on it, but the separability thesis is simply not up to the task that most jurisprudents have assigned to it.

This is not to say that there are no interesting distinctions among jurisprudential views regarding the relationship between law and morality. The first pertains to the semantics of law; the second to the normative methodology of jurisprudence.

In criticizing the Austinian view of law as commands backed by threats, Hart notes that the language of law is the language of "rights," "duties," "obligations," and so on. How are we to interpret these terms and the sentences in which they appear? The law asserts that no matter what else we have reason to do, we are required to do what the law directs us to do. The law tells us what we morally ought to do. Of course, the law can be wrong about what we ought to do, but it "sees itself" as telling us what we ought to do: what we have compelling moral reasons to do. This suggests that the law calls for a "moral semantics." The language of the law is the language of morality. When the law directs or commands us to act, we should interpret it to be imposing moral duties and conferring moral rights.4

There may be a problem in claiming that legal statements call for a moral semantics. If a legislature passes a law to the effect that we have a duty to do X, then (on the moral semantics view) we are to understand that to mean that the law claims that we have a moral duty to do X. If, however, the legislature were to have passed a law to the effect that we have a duty not to do X, then we should interpret that to mean that we have a moral duty not to do X. Apparently, the duty the law imposes (if it succeeds in imposing the duty it seeks to impose) is both moral and independent of its content. And that seems at odds with our normal understanding of the grounds of moral duties and rights, which generally depend on their content, not their source.5

Hart himself had a rather complex account of the semantics of legal judgments. First, he was a non-cognitivist. In his view, normative judgments were not apt for truth. The semantics of law is the semantics of morality but both are non-cognitivist. Legal statements are endorsements about what everyone ought to do. This would make them exactly like moral statements, but Hart held that the notion of "obligation" and "right" in law were nevertheless different from those at work in ordinary moral discourse. He drew that distinction in terms of the distinctive kinds of pressures of compliance that were brought to bear in order to encourage compliance with legal duties on the one hand and moral duties on the other. Thus, the semantics of law and morality were the same—non-cogntivist—but the notion of an obligation at work in the law was unique to it and not reducible to the ordinary moral one.

These two features of Hart’s account can be detached from one another. One can follow Hart and be a noncogntiivist about both law and morality while rejecting the view that legal obligations are distinctive (beyond the obvious sense that they are obligations under the law); or one might agree with Hart that there is something distinctive about legal obligations while advancing a cognitivist interpretation of legal propositions. And so on. Raz and I (and many others) reject both features of Hart’s account.

Arguably, the problem for positivists is to make intelligible the idea that the language of law calls for a moral semantics consistent with the claim that law itself is determined entirely by social—that is, non-moral—facts. grounded ultimately in morality. By my lights, one of Raz’s most important and least fully appreciated contributions to jurisprudence is the simple but brilliant explanation of how legal sentences (whose identity and content is fixed entirely by social facts) can be understood as making moral claims, by which I mean to say that they make truth-apt claims about what we are morally required or permitted to do.6

II.

Many jurisprudential views take the law to be a kind of "code." I characterize this jurisprudential outlook as "law-through-laws." On this account, a legal system is a set of rules that stand in a certain kind of distinctive relationship to one another: that relationship constitutes a code (of conduct). Understood in this way, the important project of jurisprudence is to identify the membership conditions that must be satisfied for a norm or a rule to be part of a legal code. Certainly, this is one of the tasks Hart sets for the concept of law and for the Concept of Law.

Many of Hart’s positivist predecessors focused on the same issue though they divided it into two parts. First, they sought to identify a non-relational property that is unique to legal rules. For many of these positivists, legal rules are distinguishable from other rules by the fact that sanctions or threats of sanctions are part of their content. With this account (or an alternative) of the uniqueness of legal rules in hand, these positivists then asked how legal rules come together to form a legal system.

Famously, Hart argued that: (1) legal rules are of at least two irreducible kinds, (2) sanctions are not essential to either of them, (3) there is no property intrinsic to legal rules that distinguishes them from other kinds of rules, and (4) the relevant law-making property is a relational one—validity under a rule of recognition. Membership in a legal code is determined by a rule of recognition.7
Though the Dworkin of the "Model of Rules"8 approaches the law from the point of view of identifying criteria for membership in a legal code, he ultimately abandons this familiar understanding of the project of jurisprudence. Recall that in the MOR his principal objection to legal positivism is that it has a bad theory of membership conditions. On his accounting, positivists—and, of course, he has Hart primarily in mind—hold that to be a law is to be a rule, and to be a legal rule is to be picked out as such by a rule of recognition. This criterion precludes moral principles from having the status of law since such principles are not rules and are not part of the law in virtue of the relevant relational property. They are law in virtue of their content as moral principles. Thus, on the natural reading of the MOR, Dworkin is following the standard approach to jurisprudence adopted by his predecessors. He has the same view about what question to ask but a different answer to it.

In "Hard Cases,"9 Dworkin characterizes his position as the "rights thesis," by which I take him to mean that the law is best understood as a set of coercively enforceable political rights—rights that particular litigants have to a decision in their favor—and not as a set of rules comprising a code. Within the "rights thesis," the project of jurisprudence is to determine the sources of law—those materials that ground the relevant political rights. In Law’s Empire, his view is that the law is best understood in terms of a set of principles that illuminates a practice in its best light.10 So-called legal rules are relevant insofar as they guide one to the right set of underlying principles. Neither the rights nor the principles theses collapse into versions of the law-as-laws approach. It is a good question to ask what the relationship between the principles and the rights theses is, an answer to which is best left for another occasion.11

Though he rejects the rule of recognition as setting out the test for membership within a legal community, Raz is himself to committed to the Sources Thesis, according to which a rule’s status as law depends on its social source and not, for example, its content. In this regard at least, Raz, like Hart before him, can be understood as approaching the law as a kind of code consisting largely of authoritative directives. In part, the project of jurisprudence is to determine membership condition, and the Sources Thesis is part of Raz’s theory of legal membership.12

III.

In addition to distinguishing between understanding law as a code and as a set of political rights or principles, we can also distinguish between so-called "normative" and "descriptive" (approaches to) jurisprudence. If we allow some of the proponents of normative jurisprudence the privilege of characterizing the descriptivists (an unsavory group that apparently includes Hart and me, but not Raz) we might wonder whether descripitivists satisfy even the most minimal standards of philosophical competence.

To see the distinction, we might begin, as some normativists would have us, with a concept like liberty or justice, rather than law. According to the normativists, so-called descriptivists believe that in order to be acceptable, an account of either liberty or justice must be neutral with regard to the normative question of whether liberty or justice is valuable, and if so why? The normativists smirk at this level of philosophical naivety. On their view, liberty and justice are valuable and any account of what they are will be in part an account of what it is that makes them valuable, or explains why they have the value that they do.

The normativist then invites us to transpose these characterizations to the analysis of law. As the normativists see it, the descriptivist apparently believes that an analysis of law should be neutral with regard to its value, whereas the normativist believes that an account of law must be an explanation of how it is that law has the value that it does.

There are several problems with this formulation of the difference between the descriptivists and the normativists. The first mistake is to treat all descriptivists alike. Suppose we begin instead by characterizing normative jurisprudence, and then see if we can distinguish descriptive jurisprudence by contrast. The key features of normative jurisprudence are these: (1) law is valuable; (2) to explain what law is is to explain in part what is valuable about it; and (3) in order to explain what is valuable about law, one has to provide an account of what law is that is itself embedded in a political theory—one that begins by assigning a value to law.

Arguably, this line of argument draws on an analogy with justice or liberty, and so we might formulate it even more generally as an account of normative or evaluative concepts or, for convenience, as a way of thinking about morality itself. In that case, the argument might proceed as follows: (1) morality is valuable; (2) to explain what morality is requires that we explain what is valuable about it; and (3) to explain what is valuable about morality we have to embed it in an appropriate, substantive, and thus controversial theory of value.
We can now distinguish among various forms of descriptivism according to where they disagree with the above line of argument. One kind of descriptivist gets off the train before it leaves the station, for he rejects the first premise of the argument. This descriptivist thinks that it is an open question whether law is valuable. He may believe that there are circumstances under which governance by law is desirable, valuable, or important, but he does not think that it is part of our understanding of what law is that it have a distinctive value of any sort. He certainly rejects the idea that we will return to below that law has a distinctive value intrinsic to it, what we might think of as the value of legality.
This kind of descriptivist may or may not believe that liberty or equality are valuable or that they have a distinctive value intrinsic to them. We can assume that he believes that justice is valuable. So this kind of descriptivist might be characterized as resisting the analogy between justice on the one hand and law on the other.

Let’s consider another sort of descriptivist, and I put myself in this category. I am not convinced that law is intrinsically valuable, but my descriptivism does not depend on whether or not law is valuable. I am prepared to allow that law is intrinsically valuable. The form of descriptivism I associate myself with takes the inference from the second premise to the conclusion to be invalid. For even if law is something of value it hardly follows that in order to analyze the concept of law one must appeal to substantive political morality. Indeed, part of the problem is trying to understand how an account of the concept of law or the nature of law is supposed to figure in an explanation of its value. Let’s take up both of these points, beginning with the invalidity of the inference from the value of law to a commitment to political theory of the concept.

The best way to see the problem is to have a look at the concept of morality, in part because all but the most skeptical among us allow that morality has value. As Kenneth Himma has pointed out to me, William Frankena analyzed the concept of morality as picking out a set of norms that restrict our behavior as it pertains to other human beings. (He might have been wrong to restrict the scope of morality to humans, but that is not our concern here.) The norms are categorical in the sense that they apply to a person regardless of how she feels or what she believes. They are universalizable in that they apply in the same way to everyone in relevantly similar circumstances; and they are supreme in the sense that the obligations or duties they impose win out when in conflict with other kinds of obligations or normative claims. We can understand Frankena to be offering an account of the application conditions of the term "moral,"—the conditions that must be satisfied for a norm to count as a moral rule. Nothing in this account of the concept of morality commits one to endorsing the content of any of the rules that satisfy the application-conditions of the concept. Nevertheless, no one seriously denies that morality is valuable. It is just that an account of what morality is does not require substantive claims about its value.
But it is exactly at this point that our normativist enters his protest: If there are no normative propositions anywhere to be found in the analysis of our concept of morality, then how are we to establish that morality is something of value? How can we draw the normative inference without normative premises? And that shows that the Frankena style analysis is inadequate. Rather than undermining normativism in jurisprudence, this line of argument establishes that it is inescapable.

Not really. The key is to understand exactly how an account of the concept of law or an account of what law is figures in an explanation of the value of law. Suppose that we take law to be valuable, by which we mean that governance by law is desirable. What is the connection between this fact about law and the concept of law or the nature of law? The right answer to this question—according to the descriptivist—is this: an explanation of why law is desirable must make reference to at least some of law’s essential features. If it is true that law is desirable but we cannot make sense of this by reference to some list of its essential features, then that list is incomplete or we have an inadequate account of what law is. The key point for a descriptivist is that to understand why law is desirable under certain circumstances, we are going to have to know much more than what we can read off the concept or infer from law’s nature. We are going to have to know more about human beings—their interests, the constitutive elements of their welfare, what they want to accomplish jointly and severally, and so on.

The problem with the normativist, then, is that she gives us the impression that we should be able to read the value or desirability of law off the concept itself or infer it from the nature of law itself. Once we realize that this is very implausible, we see that properly understood descriptivism provides all the normative insight into law that is appropriate. Moreover, it is a feature of descriptivism, so understood, that it can provide an explanation of the value or desirability of law that is compatible with a broad range of political moralities, and not just with the one that is presumed to be embedded within the concept or nature of law on any account that says that in order to analyze the concept we must begin by attributing a value to law.

The claim that normative jurisprudence is distinguishable from descriptive jurisprudence because only the former can explain the value of law is simply mistaken. Law may not be intrinsically valuable, but if it is, descriptivism cannot only explain how that might be, it does so in a way that is methodologically sound: something that cannot be said for normative jurisprudence so understood.

How else might we make out the alleged distinction between descriptive and normative jurisprudence? Some have argued that a theory of law presupposes an account of the proper function of law. A theory of law is in part a theory of its proper function. The idea of a proper function is at least with regard to law a moral one; and so an account of law is an account of the morality of law.
This line of argument is not likely to succeed. First, though a legal system is a social artifact designed by agents to secure certain ends, some more valuable than others, it does not follow that law has a function in the philosophically interesting sense. That is, law may possess no functional property that would explain its existence, persistence, and the shape it takes in its mature form. Secondly, even if law had a function, it need not be a proper function in the moral sense.13

My view is that there is an important sense in which Raz adopts a normative jurisprudence, but we haven’t hit upon that sense yet. For when it comes to the value of law, I am pretty confident that Raz does not believe that law has an intrinsic political value. I am also confident that he believes that the philosopher of law takes a detached point of view, and so in providing an account of what law is he is not taking a posture of endorsement. Instead, he is looking to uncover "application-conditions" of the concept or looking to uncover necessary truths about law. In what sense, then, is Raz a normative jurisprudent? In what sense is he closer to Dworkin than to Hart?

IV.

If we want to understand the ways in which Raz approaches law similarly to the way Dworkin does, perhaps we should begin by having a look at the way Dworkin does. On my reading of him, Dworkin’s view is that law is an answer to a particular question in political morality, and that question is: Under what conditions is the collective use of force (of all against some) justified?
The problem of justifying the collective use of force is clearly a problem in political philosophy. We can imagine a range of possible answers to it. One answer might be that the collective use of force is justified if and only if it is required by justice. Another answer is that collective force is justified if it is authorized by law. So one might argue that essentially law is an answer to the question of when the collective use of force is justified. This means, of course, that not everything that purports to be an answer to the question—when is collective force justifiably imposed—law. Law is a distinctive kind of answer to that question and so one cannot derive a full account of what law is from this feature of it. I will not have much else to say in this essay about what law is on the Dworkinian picture beyond emphasizing this feature of it: namely, that it is an answer to a problem in political morality.

The particular answer that law gives is that collective force is justified when it is warranted by past political acts. (Not every political act is relevant, of course, but determining which are and which are not are parts of the theory that I am going to set aside in this essay.) Imagine that a case comes before a court and the question is which litigant has a right to a decision in her favor. That right is coercively enforceable and is based on past political acts (that is, authoritative decisions of one sort or another). Now we are to understand law as an answer to the question: When is collective force justifiably imposed? This implies that in determining which litigant has the relevant legal right in her favor, the judge is committed to making that determination in a way that would make her decision as close to a justification for using the coercive authority of that state as she could possibly make it. This means that she must interpret the relevant political texts in their best light—as providing the best justification that they can for the exercise of that force. And it is easy to see that what does the work for Dworkin is, at the basic level, the set of principles that provide the best possible interpretation of the relevant political acts, and thus which provide the best available justification for the use of force.

Other features of Dworkin’s jurisprudence are intertwined with this most basic feature of it. First, for Dworkin, the important relationship in political morality is coercion: the power collectively to enforce directives by force. Secondly, political morality is a distinctive domain of normative discourse that concerns itself with a distinctive set of questions and which is distinguished from other normative domains by certain institutions. Thirdly (and relatedly), political institutions are regulated by norms specific to them, and they embody or exhibit values distinctive of them. As I read him, Dworkin’s view is that the political virtue of law is integrity.14

As I noted, most commentators (including Dworkin) treat Raz as the leading contemporary legal positivist and as the heir to Hart’s positivism (as well as his throne). The main reason for labeling Raz a positivist is his commitment to the Sources Thesis. The Sources Thesis is the claim that the identity and content of law depends on social facts alone. Social facts are facts about human behavior and attitude. If we focus only on the Sources Thesis then it is natural to place Raz within the tradition that includes Austin and Bentham then runs through Kelsen and Hart. And this seems just right. With regard to his "substantive jurisprudence" Raz is a positivist—if anyone is.

On the other hand, one would miss a lot about Raz’s views were one focused only on the Sources Thesis in isolation from its relationship to the problem of determining the conditions of legitimate authority.

Though Raz does not approach the law from the point of view of it providing an answer to the question when is the collective use of force justified, he does approach the law from the vantage point of a problem in political morality. Raz believes that the law is a special instance of practical authority and that one can only understand law by understanding the nature of practical authority. More importantly, Raz draws an important distinction between de facto and de jure practical authority. A de facto authority is someone who issues directives that others treat as binding on them. A legitimate authority is one that in fact imposes obligations, one whose directives actually bind others.

The relationship between authority and legitimate authority is complex. The notion of an authority is itself explained in terms of the concept of legitimate authority. One who claims to be an authority claims to be a legitimate authority and not merely a de facto authority. In that sense, the notion of legitimate authority is more basic in the order of explanation. On the other hand, the notion of a legitimate authority is conceptually "constrained" by the possibility conditions for being an authority. By that I mean that nothing can be a legitimate authority unless it is the sort of thing that can stand in an authority relationship to others. An authority must be capable of contriving, formulating, and issuing directives. These are features of persons. The failure to notice these conceptual constraints on the idea of a legitimate authority has led to a good deal of confusion and to some implausible objections to Raz’s views on legitimate authority.15

According to Raz, law necessarily claims to be a legitimate authority. The claim may be false, but it cannot be necessarily false. It is a claim that could be true of law. Raz’s view is that the claim can only be true if the identity and content of law can be fully determined by social facts alone. If the identity or content of law is determined in part by moral facts, then law’s claim to authority is vitiated: thus, the Sources Thesis. Whatever else law is, it must be (that is, it is essential to law) that it be capable of being a legitimate authority. So the capacity to be a legitimate authority imposes a substantive constraint on what law is. And thus the nature of law—or essential features of it—derives from a constraint of political morality: that is, the conditions of legitimate authority. This is the sense in which Raz is a normativist, and it is also one of the important and interesting ways in which his approach to jurisprudence is much closer to Dworkin’s than it is to Hart’s.

V.

Claims of authority take the following form: X has (claims, exercises) authority over Y with regard to domain of activity Z. We can distinguish a Razian from a Hobbesian conception of authority. Hobbesian authority is the rightful power to issue directives. Those directives in turn can be justifiably enforced. Hobbesian authorities do not necessarily claim a power to alter the normative status of agents with regard to the actions that are available to them, only a power to issue enforceable directives. This is not to say that the Sovereign does not change what moral reasons for acting those he governs have. If they have contracted with one another to create a Sovereign, then they may have under the contract moral duties to obey the Sovereign’s directives; and in that sense, by issuing directives the Sovereign alters what they have moral reason to do. It is just not a part of the Hobbesian conception of authority that the authority claims the power to alter the moral reasons for acting his subjects have.
In contrast, according to the Razian account, an authority claims a power to create moral reasons for acting. And it is because the reasons for acting he creates are moral ones that the exercise of coercion to enforce them is justified.

Legal rules are authoritative directives. Authoritative directives have a certain structure. Authoritative directives are second order exclusionary reasons and first order reasons for acting. In saying that they are exclusionary reasons, Raz means that they preclude the agent from considering some set of first order reasons that would otherwise apply to him. In saying that they are first order reasons for acting, Raz means that they are to figure in an agent’s deliberations about what the balance of first order reasons requires him to do—that is, the balance to be struck between the authoritative directive and the remaining first order reasons that apply to him that are not excluded by the authoritative directive.

This way of understanding the concept of an authoritative directive does not rest on empirical claims about how such directives actually figure in this or that person’s deliberations. Instead, it purports to express something important about the nature of rational deliberation. What is the connection between authority, rational deliberation, and reason? Why might the very idea of rational deliberation entail an authoritative directive excluding at least some first-order reasons for acting that would otherwise apply?

One answer is that an authoritative directive is itself justified by those very reasons that it excludes. Someone issuing such a directive claims, in effect, to have balanced those reasons and determined what the balance of reasons requires. When someone claiming to be a legitimate authority issues such a directive, we must interpret her as asserting that she has balanced the reasons properly and determined accurately what reason demands. Thus, if it is rational to accept a directive as authoritative, then it is rational to exclude the reasons on which the directive is based. And those reasons—the dependent reasons—are at least some of the first order reasons that the agent would normally balance herself. To rationally accept an authority is for it to be rational to displace aspects of one’s own practical reasoning with that of the authority.
Raz calls the reasons that justify an authoritative directive "dependent reasons," and he dubs the theory of legitimate authority the "service conception." Authoritative directives exclude the dependent reasons that justify them, and the claim of legitimate authority is true only if an agent will do better acting on the authoritative balancing of the reasons that apply to him than he would by acting on his own assessment of them.

Those who accept this view of legitimate authority and of the role authority plays in our practical reasoning believe that it has important implications for the law. As a general matter, if an individual must appeal to the dependant reasons that apply in order to determine either the identity of an authoritative directive or its content, then in doing so she vitiates the claim to authority that the directive makes. That is because the claim to authority is a claim to exclude such reasons in one’s deliberations. And this means that the content and identity of law must be determined by factors other than the moral or dependent reasons that would justify it.

The law necessarily claims to be a legitimate authority. The Razian view is that this claim may be false but it is not incoherent and could not therefore be false necessarily. Thus, law must be the sort of thing the claim could be true of (here is where the nature of law answers to an issue in political morality). Given the above argument, this means that the identity and content of legal directives has to be ascertainable without appealing to the moral principles that justify the directives. And it is this argument that gives rise to the Sources Thesis—the claim that the identity and content of law must be determined by social facts alone.

The emphasis on the authority relationship means that law implicates reasons for acting in two distinct ways: internal reason and external reason. Quite apart from authority, law will invariably implicate internal reasons for acting. For example, if we think of law as a coordinated activity or as involving planned or rationally structured activity, then it must conform to the norms governing planning, coordination, and rational structuring of collective action and decision. These are norms internal to legal activity—norms of consistency and coherent rational collective action.
The more important claim is that the authority relationship means that law engages reasons that are otherwise external to it. Whether or not a community has law, individuals are bound by the reasons that apply to them. The law’s claim to authority just is a claim that individuals will do better acting on the balance of those reasons by acting on the reason the law provides than they will do by acting on their own assessment of what those reasons require. Legal reasons in this sense are in the mix with moral reasons—indeed, legal reasons claim often to displace such reasons. We might think of the law as being governed by or engaging external moral reasons and internal reasons of "integrity."16
The claim that law requires a moral semantics is likewise connected to this train of thought. The law engages our moral reasons for acting that it claims often to displace. So the best way to understand what the law is saying is to understand it as giving moral reasons for acting. So if we are to understand the claims the law is making about what we ought to do, then this is how we ought to understand the law.
Relatedly, it is easy to see a connection between the centrality of the authority relationship for Raz and his liberal perfectionism. According to perfectionism the domain of the state’s legitimate authority is determined by what matters to persons—what is valuable to them. What is valuable to them gives them reasons for acting. The law justifiably enters this domain and its claim to legitimate authority is in a way just the claim that you will better serve the reasons that apply to you and which you ought as a person to value. This is of a piece with the perfectionist outlook and again stands in contrast to the familiar liberal picture that both Hart and Dworkin, among others, embrace.
We may put the point in a rough and crude but nevertheless illuminating way, I think. The perfectionist is drawn to the authority relationship as fundamental to law in part because it engages individuals and the reasons that apply to them; and the law is properly concerned with what matters to agents. The traditional liberal (and in this case I have in mind Dworkin and the Hart of Law, Liberty and Morality) is drawn to the coercion relationship because the domain of law is restricted to the justified interferences with individual liberty.
I am not sure whether it is the perfectionist/liberal divide or the authority/coercion divide that is of fundamental importance, but it is also important to note that Raz and Dworkin take somewhat different postures towards the political domain and to the set of institutions that are distinctive of the political sphere. As I read him, Raz does not deny that the political domain is in some sense distinctive. Still, the political domain engages human agents through the authority relationship and so it is best understood in terms of the role it plays in the lives of agents—that is, in affecting their ability to act on the reasons that apply to them. This seems to me the force of understanding authority in terms of "service." It is all about persons and reasons. The state simply (but not uncomplicatedly) mediates between the two.
I take it that Raz is not committed to understanding political institutions as exhibiting distinctive political virtues that are in a sense intrinsic to them. And so he does not see it as a burden of his account to identify a distinctive political value (beyond service) intrinsic to law. After all, the value of serving the interests of agents in acting on the reasons that apply to them is not distinctive of law; indeed, it is not distinctive of authoritative institutions more generally.
In contrast, one of the distinctive and important features of law for Dworkin is that in answering the question what is the use of collective force justified it does so by expressing or embodying a political value that is distinctive of it as an institution—and that is the value of integrity. If we understand jurisprudence as an account of law and of cognate concepts or properties like legality, then we can understand Dworkin as holding that an account of law is in the end an account of the value of legality. And that concept of legality is to be spelled out in terms of integrity.
VI.
To this point I have argued that from an important methodological point of view Raz has more in common with Dworkin than he has with Hart. To be sure, Raz does share with Hart the "law-as-laws" approach to jurisprudence with its emphasis on membership conditions and authoritative acts. In contrast, as I read him, Dworkin does not identify law with the idea of a code.
On the other hand, both Raz and Dworkin approach law from the point of view of a problem in political morality: the nature of legitimate authority in Raz’s case, and the justification of collective force in Dworkin’s. Raz turns out to be a positivist in my view because the answer he gives to the problem of political morality is the Sources Thesis. His positivism has a normative foundation. And this again distinguishes him from Hart. And it places him squarely with Dworkin in the normativist camp—given the common parlance—which, left to my own devices, I would eschew entirely.
Unlike Hart, both Raz and Dworkin take the law to have a moral semantics: the claims the law makes are best interpreted as moral assertions—as imposing moral duties, conferring moral rights, privileges, and powers. The problem is to explain how this is possible. There is no problem for Dworkin. After all, law offers an answer to what justifies the use of collective force, and so in each case that the law speaks we must understand what it is saying as providing reasons that it takes to be adequate to answering that question. Any answer to that question will take moral claims (or claims of political morality more broadly) as among its premises. The content of the law, for Dworkin (and others) is fixed by some complex function of interpretation in which a set of moral principles is applied to social facts.
The problem of explaining how it is that the law can make moral claims and purport to impose moral duties arises for the positivist like Raz, not for the interpretivist like Dworkin. The theory of authority entails the Sources Thesis. The Sources Thesis entails that the identity and the content of the law must be determined by social facts alone. The question is, how do we square the claim that the identity and content of law must be fixed by social facts with law’s requiring a moral semantics?
The answer is that we are to understand the law as "a point of view." That is, law is a perspective on the demands of morality. Legal judgments are statements about what morality requires from a particular point of view—namely, the point of view of the law. One aim of jurisprudence, then, is to characterize the notion of a point of view and, in particular, how to understand the idea of the law’s point of view—that is, what it is to understand the law as a point of view. And the value of jurisprudential theory, then, is the light it sheds on the importance of such a point of view in our practical and social life, and, in the contribution it makes to our self-understanding.

Two final points: the first is one of my favorites. Given Raz’s views, the claim to legitimate authority entails that one cannot look behind authoritative directives or particular legal rules in order to determine the identity or content of those directives. Doing so would vitiate the claim to authority. Thus, the authority relationship creates a nearly impenetrable barrier between law and the dependent reasons on which the law is grounded. To accept the law as authoritative is in effect to shut oneself off from assessing or confronting the balance of those reasons.
In contrast, for Dworkin, particular legal acts or authoritative directives must be transparent (or at least translucent) to the set of principles that provide the best justification of them. That is because the law is, in effect, a brief on behalf of the use of collective force; and to know the law is to know what principles figure prominently in that brief. So from the same methodological point of departure we get two completely different accounts of the relationship of law to its justifying grounds.
The point with which I want to close is much more speculative and I mention it primarily to encourage discussion of it. It is also an ironic point that requires a little bit of table-setting.
In another paper, Ori Simchen and I drew a distinction among three different kinds of jurisprudential questions. These are:

I. What is law (understood as a system of governance)?
II. What is it to be a law?
III. What is the law around here (indexed to a particular legal system)?
And we made much of the importance of these distinctions and the relationships among the different questions.

If we understand (as Dworkin does) law to be an answer to the first question, we should be struck by the fact that it almost never seems to give a persuasive answer to it. Here’s what I mean. Take a different possible answer to the first question: collective force is justified if and only if it is required by justice. I don’t mean to suggest that this answer will always be correct, but at least it seems to be in the right ballpark. If it weren’t a good answer, we would have to explain why that would be so for, intuitively, the fact that justice requires that collective force be employed provides strong support for doing so. No one would be surprised were this one’s opening gambit in providing an answer to the question.
In contrast, some might find the answer "the collective use of force is justified whenever the law says so," more puzzling for a number of reasons.
It strikes me as odd that it is in the nature of law to offer an answer that is almost always going to be inadequate to the task at hand. The law is an answer to the question, what justifies collective force? But the answer it gives, constrained by history, accident, and political pressures, is almost always going to fall short. In fact, it is hard to imagine that the answer law given under any but the most ideal circumstances is going to be unpersuasive. All of these factors contribute to the answer the law gives and, at most, the best we can expect law to do is make a pretty good case for the use of collective force every now and again. I am having difficulty imagining being drawn to a theory of the nature of law that sees it as failing as a matter of course at its essential task. This is not a criticism, just an observation.
To be sure, Raz is not asking Dworkin’s question, but the irony is that he has a more persuasive answer to a slightly different version of that question than Dworkin does—at least by my lights. To see this, let’s shift attention from what law is to what the law around here is. Now ask Dworkin’s question: When is collective force justified? Here is Raz’s answer. The law around here is best understood as an answer to that question, and invariably a plausible one—even when it is mistaken. After all, the law is a perspective on what morality requires in various domains of activity. Because the law just is such a perspective on the demands of morality, it is plausible to understand the law (around here) as an answer to Dworkin’s question. And this just is the difference between Raz and Dworkin on this point. In Dworkin’s view, law is not a perspective on demands of morality. Rather, it is a distinctive kind of putative justification of the coercive authority of the state—one that relies especially on past political acts. The fact that on his view law is a perspective on the demands of morality provides Raz with a better—or more coherent—answer to Dworkin’s question than Dworkin himself has on offer.
VII.
I have argued for several distinct but related points. First, I have tried to shed some doubt on the value of the traditional distinctions in among jurisprudential views, while taking a few shots at the conventional understandings of where everyone fits on the jurisprudential landscape. In particular, I have raised some doubts about thinking of Raz as the natural heir to the Hartian legacy.
I have introduced several distinctions that are more illuminating than the ones philosophers of law have typically focused on. In the course of the argument, I have suggested that the most important distinctions are methodological ones, and from that perspective Raz has more in common with Dworkin than with Hart. Most of the essay was devoted to making out this claim and then distinguishing Raz from Dworkin given their very similar methodological starting points.
I hope that in doing so I have provided a fresh look at Raz’s work while highlighting some of the most important and compelling contributions he has made to jurisprudence. Despite our substantive disagreements—the least important of which is the one that other commentators have most closely focused on, namely, the dispute between inclusive and exclusive legal positivism—no contemporary legal philosopher has had greater influence on my own thinking. Along with Dworkin, Raz is the true giant of our field. I hope I have gone part of the way to explaining why that is so.

Endnotes

* This essay is drawn from several sections of the 2006 H.L.A. Hart Lecture, "Rethinking Jurisprudence," to appear in the Oxford Journal of Legal Studies (forthcoming).

** I want to thank Scott Shapiro, Ben Zipursky, and especially Kenneth Himma for extremely helpful comments on this essay in earlier stages of its development.

1. Joseph Raz. The Concept of a Legal System: An Introduction to the Theory of Legal System, 2nd ed. (Oxford: Clarendon Press, 1980).

2. In the Postscript to the Concept of Law, Hart appears conflicted. On page 254, he suggests that inclusive legal positivism is the correct view—provided that moral objectivism is true. Hart himself was a non-cognitivist. This might lead one to think that he did not really embrace inclusive legal positivism; yet on page 250, he unambiguously and unconditionally embraces inclusive legal positivism.

3. Brian Leiter. "Legal Realism and Legal Positivism Reconsidered," Ethics 111 (2001): 278-301.

4. The claim here is not that every legal proposition asserts a moral proposition, only that legal reasons are best understood as expressing moral claims.

5. Some might think that this consequence is reason enough to reconsider whether we might have underappreciated the plausibility of the command view. According to the command theory, legal sentences are typically commands and do not assert facts about our duties or rights—moral or otherwise. "Shut the door!" makes no claims.

6. In other words, Raz holds both that the identity and content of law must be determined without regard to the moral principles or reasons that would justify the law and that the semantic content of the law is to be understood in terms of its making moral claims about what we have rights, powers, and privileges to do on the one hand, and duties and obligations to do on the other. We shall return to Raz’s account of how this is possible below.

7. Arguably, the emphasis that Raz places on the so-called Sources Thesis suggests that he, like Hart before him, approaches law-through-laws and thus takes providing a theory of law’s membership conditions to be an important task of jurisprudence. Many non-positivists, including natural lawyers, also understand the law as a kind of code, and so they too take one of the central tasks of jurisprudence to be determining membership conditions for legal codes; and so it would be a mistake to identify the law-through-laws approach to legal positivism.

8. Ronald Dworkin. "The Model of Rules," The University of Chicago Law Review 35 (1967): 14-46.

9. Ronald Dworkin. "Hard Cases," Harvard Law Review 88 (1975): 1057-1109.

10.Ronald Dworkin. Law’s Empire (Cambridge: Harvard University Press, 1986).

11.Some have argued that because Dworkin’s method requires a test for identifying the relevant set of "authoritative acts" it too is committed to the centrality of membership conditions, and thus (implicitly at least) to law as a code: that, in other words, the distinction between "law-through-laws" and "law-as-principles" is artificial. Tempting yes, persuasive no. Dworkin’s view is that the law is the product of applying an interpretive theory to pre-interpretive data. That data is not determined by applying a test of membership but is constituted by convergence of paradigm instances that are themselves revisable in the light of the outcome of applying the interpretive method to them. Only once we interpret the relevant materials can one identify the underlying political principles. And once the principles are in hand, one can derive what counts as criteria of membership, and given those criteria, what the actual authoritative legal acts are. The actual authoritative acts are determined post-interpretively, a consequence of applying the interpretive method, not a product of a test for membership that precedes applying that method. On the other hand, the set of authoritative acts that are derived by applying the interpretive method are at the same time the pre-interpretive materials to which the interpretive method will be applied next time around. It is a consequence of Dworkin’s approach that the set of binding legal acts is always subject to revision insofar as any case that comes before a court in principle provides it with the opportunity to revisit the question as to what the underlying principles are.

Everything is in the mix (so to speak) all the time, always in principle subject to revision. What counts as a central or a peripheral issue—a core or penumbra case—is itself a theoretical, not a factual, question, as is a function of applying the interpretive method. In that sense, the set of authoritative legal acts is in a sense epiphenomenal. This gives "rules" a very different status than they have within the "law-through-laws" approach.

12.So while Hart, Raz, myself, and most other legal positivists as well as many natural lawyers fall squarely within the law-through-laws camp, Dworkin, Greenberg, and Stavropoulos clearly do not. Dworkin himself has pursued at various times at least two different approaches to law: "law-as-enforceable-political rights" and "law-as-principles."

13.It does not help those pursuing this line of argument that neither Hart nor Raz, Dworkin nor I believe that law has an essential function (proper or otherwise). Many readers mistakenly think that Dworkin believes that the essential function of law is to justify the coercive authority of the state. Dworkin, in fact, analyzes law in terms having to do with the exercise of coercion, but he does not claim that justifying coercion is the function of law. I take this up in detail below.

14. Thus, his view: law-as-integrity.

15. So some have argued that Delphic Oracles and inanimate objects can satisfy the normal justification thesis—that is, one might do better acting on their directives than one might do acting on one’s own assessment of the balance of reasons. In that case, inanimate objects can be legitimate authorities, which would be a counter-intuitive result. But these aren’t plausible objections for the reasons given in the text. Inanimate objects lack the capacities that are necessary for being an authority. And in this (conceptual) sense, the notion of an authority is more basic in the order of explanation than is that of a legitimate authority.

16. This difference between internal and external is analytically sound of course, but one way of understanding some jurisprudential theories, say, Dworkin’s, is that he denies that the internal reasons—especially those of the kind of consistency that constitutes integrity—are "purely" internal to law. Their status as reasons in law depends on considerations of political morality more generally.

Previous Article | Index | Next Article

APA Newsletters

Spring 2007 Volume 06, Number 2

Spring 2007
Volume 06, Number 2