Hostname: page-component-74656f9884-2dmjx Total loading time: 0 Render date: 2026-11-08T17:18:52.705Z Has data issue: false hasContentIssue false
  • English
  • Français

NATURAL LAW JURISPRUDENCE

Published online by Cambridge University Press:  15 March 2004

Mark C. Murphy
Rights & Permissions [Opens in a new window]

Abstract

An abstract is not available for this content. As you have access to this content, full HTML content is provided on this page. A PDF of this content is also available in through the ‘Save PDF’ action button.

Information

Type
STATE OF THE ART ESSAY
Information
Legal Theory , Volume 9 , Issue 4 , December 2003 , pp. 241 - 267
Copyright
© 2004 Cambridge University Press

While it is not entirely misleading to describe the (very different) theories of law forwarded by Lon Fuller1

“Do the principles expounded in my second chapter [of THE MORALITY OF LAW] represent some variety of natural law? The answer is an emphatic, though qualified, yes.” See Fuller, THE MORALITY OF LAW 96 (1964). Fuller describes his natural law view as “procedural” rather than “substantive.” But if we read Fuller's account of the internal morality of law as an account of what law must be in order to provide rules to guide the conduct of rational beings like us, it seems that to appeal to the formal features such rules must possess (prospectivity, clarity, possibility, etc.) while eschewing entirely appeal to the substantive features such rules must possess (that is, adequately reflecting or extending the reasons for action that beings like us have) is an arbitrary truncation. Fuller's views have been making a bit of a comeback: see, e.g., David Luban, Natural Law as Professional Ethics: A Reading of Fuller, 18 SOC. PHIL. & POL'Y 176–205 (2001); and David Dyzenhaus, Positivism's Stagnant Research Programme, 20 OXFORD J. LEGAL STUD. 703–722, 720–722 (2000); see also REDISCOVERING FULLER: ESSAYS ON IMPLICIT LAW AND INSTITUTIONAL DESIGN (Willem J. Witteveen & Wibren van der Burg, eds., 1999); and also several of the essays in RECRAFTING THE RULE OF LAW: THE LIMITS OF LEGAL ORDER (David Dyzenhaus, ed., 1999).

and Ronald Dworkin2

Dworkin expresses a limited willingness to describe himself as an advocate of natural law theory, by which he means the view that “what the law is depends in some way on what the law should be”; Dworkin, “Natural” Law Revisited, 34 U. FLA. L. REV. 165–210, 165 (1982). But it is hard to square the notion of Dworkin as natural law theorist with Dworkin's limited theoretical ambitions, that is, to provide an account of our practice of law: “General theories of law, for us, are general interpretations of our own legal practice”; Dworkin, LAW'S EMPIRE 410 (1986). A parochial natural law theory is no natural law theory at all.

as natural law theories, it is pretty clear that the contemporary resuscitation of the full-blooded natural law view in analytical jurisprudence dates from John Finnis's 1980 Natural Law and Natural Rights.3

John Finnis, NATURAL LAW AND NATURAL RIGHTS (1980).

The theses defended by Finnis in that work cover the range of philosophical fields in which natural law doctrines have been affirmed: it includes natural law accounts of practical rationality,4

Finnis, supra, note 3, at 59–126.

of natural morality,5

Finnis, supra, note 3, at 126–127, 297–308.

of politics,6

Finnis, supra, note 3, at 134–259.

of law,7

Finnis, supra, note 3, at 3–29, 260–296, 308–368.

and of religious morality,8

Finnis, supra, note 3, at 371–413.

and these accounts are, like the natural law accounts of these matters offered by the paradigmatic natural law theorist, Thomas Aquinas,9

Aquinas' account of human law is located within an account of law in general, and hence Aquinas' understanding of human law is developed in relation to his views on eternal, natural, and divine law; further, Aquinas systematically connects his account of law to the notion of good, reason, virtue, and justice. Because Aquinas has, at the point at which the so-called Treatise on Law appears in the SUMMA THEOLOGIAE, developed substantial accounts of good, virtue, and rational action, he is able to draw on these accounts in offering his theory of human law.

tightly integrated. But it is possible to extract the natural law view in analytical jurisprudence from the specifics of the natural law views in ethics and politics, as long as one is willing to proceed at a certain level of abstraction.10

For a discussion of why both ethical and jurisprudential views go by the name “natural law theory” and why it is often difficult to disentangle them, see Philip Soper, Some Natural Confusions about Natural Law, 90 U. MICH. L. REV. 2393–2423 (1992).

And proceeding at this level of abstraction is worthwhile, for it is not clear to what extent these other natural law positions are needed to provide support for the natural law position in jurisprudence, and it is not clear whether the natural law account of human law is more or less compelling than the natural law moral and political theories. Their fates should not be tied together.

With respect to natural law theory in analytical jurisprudence, Finnis set himself three tasks in Natural Law and Natural Rights. The first was to deny that a number of absurd theses that have been associated with the natural law doctrine are in fact asserted by that view;11

See the chapter in Finnis, supra, note 3, entitled Images and Objections, at 23–49, esp. 25–29.

the second was to provide a more adequate formulation of the central natural law jurisprudential thesis;12

For Finnis's final definition of law, which incorporates the natural law elements that he defends at various points in the book, see Finnis, supra, note 3, at 276–277.

and the third was to offer an account of the sort of argument that is needed to establish the truth of central natural law thesis.13

See esp. Finnis, supra, note 3, at 3–18.

Since the publication of that book, he has also devoted himself to a fourth task, that of explaining to what extent the natural law thesis is incompatible with, and to that extent superior to, families of claims affirmed by prominent versions of legal positivism and other competing jurisprudential views.14

On legal positivism, see Finnis, On the Incoherence of Legal Positivism, 75 U. NOTRE DAME L. REV. 1597–1611 (2000); on critical legal studies, see Finnis, On “The Critical Legal Studies Movement,” in OXFORD ESSAYS IN JURISPRUDENCE, 3rd series 145–165 (John Eekelaar & John Bell, eds., 1987); on Dworkin's view, see Finnis, On Reason and Authority in LAW'S EMPIRE, 6 LAW & PHIL. 357–380 (1987), and Natural Law and Legal Reasoning, in NATURAL LAW THEORY 134–157, 143–148 (Robert P. George, ed., 1992).

Since Finnis's 1980 book, a number of writers—some self-consciously following Finnis, others not; some adherents of the natural law position, others not—have taken up these questions as well.

I am going to proceed as if it were no longer an important task for natural law theorists to dissociate their view from the stupid positions that are sometimes associated with it, for example, that a norm's status as a correct moral norm is sufficient for its being a legal norm, or that even if a norm's status as a correct moral norm is not sufficient for its being a legal norm, the law can do no more than transcribe the content of moral norms as legal norms. That this task is no longer an important one for natural law theorists is almost certainly false, but there is only so much clarification that natural law theorists or sympathetic exponents of that view can do before one must wonder if anyone is listening.15

It is simply obvious that the natural law view is not interested in claiming that Φ-ing's being morally required is sufficient for Φ-ing's being legally required, or that Φ-ing's being independently morally required is necessary for Φ-ing's being legally required. No thesis that Aquinas, the paradigmatic natural law theorist, takes pains to deny could be essential to the natural law position. But Aquinas explicitly denies that a norm's status as part of the natural law is sufficient to make it part of human law (SUMMA THEOLOGIAE IaIIae 96, 2) and explicitly affirms that legal norms can properly go beyond what is dictated by the natural law alone (SUMMA THEOLOGIAE IaIIae 95, 2). It also seems to be, sadly, pretty obvious that this cartoon understanding of natural law theory is not sufficiently eradicated from the space of jurisprudential argument. No conception of natural law theory that is employed both by one of the leading legal positivists of our time and by one of the leading legal realists of our time can be one that is eradicated from the space of jurisprudential argument. But note the following argument from a recent piece coauthored by Jules Coleman and Brian Leiter:

Now we can see the problem with the natural lawyer's account of authority. For in order to be law, a norm must be required by morality. Morality has authority, in the sense that the fact that a norm is a requirement of morality gives agents a (perhaps overriding) reason to comply with it. If morality has authority, and legal norms are necessarily moral, then law has authority too.

This argument for the authority of law, however, is actually fatal to it, because it makes law's authority redundant on morality's….Natural law theory, then, fails to account for the authority of law.

See Coleman & Leiter, Legal Positivism, in A COMPANION TO PhILOSOPHY OF LAW AND LEGAL THEORY 241–260, 244 (Dennis Patterson, ed., 1996). This argument against natural law theory entirely depends on natural law theory's affirming the “copy” view of the relationship between law and practical reasonableness, where human law can do no more than transcribe these natural requirements of reason. Coleman and Leiter are undoubtedly right that this is a stupid view. But it is no part of the natural law position. Hence those who affirm a natural law position in jurisprudence, if they are interested in the question of political authority, must go on to explain how and when those legal norms that go beyond the natural law are authoritative. See, e.g., Finnis, supra, note 3, at 231–252; Finnis, The Authority of Law in the Predicament of Contemporary Social Theory, 1 NOTRE DAME J.L. ETHICS & PUB. POL'Y 114–137 (1984); see also my Natural Law, Consent, and Political Obligation, 18 SOC. PHIL. & POL'Y 70–92 (2001).

It is, however, worthwhile to spend some time on the other three tasks: asking about what progress has been made in providing an acceptable formulation of the natural law thesis, determining what sort or sorts of argument can support that thesis, and seeing whether there remain any significant theoretical differences with putative jurisprudential rivals, in particular, with positivist views. I will conclude with some brief remarks on natural law theories of adjudication.

FORMULATING THE NATURAL LAW THESIS

The dominant contemporary understanding of natural law theory is, strangely enough, not drawn from any reading of natural law theorists themselves, but from Hart. Natural-law theory is the view that Hart rejects in his 1957 Holmes Lecture on “Positivism and the Separation of Law and Morals.”16

H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 HARV. L. REV. 593–629 (1958); cited to the reprint in Hart, ESSAYS IN JURISPRUDENCE AND PHILOSOPHY 49–87, 55 (1983).

Positivism believes in the separation, or at least the separability, of law and morals; natural law theory must be the denial of that claim. To be a natural law theorist is, then, to reject the view that law and morality are separable.

This way of putting the natural law thesis is unsatisfactory in a number of ways. It is first of all unclear as to whether the natural law view should be framed in terms of morality. The notion of morality is arguably a peculiarly modern notion,17

See, e.g., Alasdair MacIntyre, AFTER VIRTUE 1–5 (2nd ed., 1984); and Bernard Williams, ETHICS AND THE LIMITS OF PHILOSOPHY 6 (1985).

and it would be courting anachronism to refer to the views of Aquinas, the paradigmatic natural law theorist, in such terms. Better to adopt Finnis's suggestion and simply appeal to “practical reasonableness,” a semitechnical term that is meant to cover the entire range of good reasons for action rather than suggesting a subset of them.18

“The term ‘moral’ is of somewhat uncertain connotation. So it is preferable to frame our conclusion in terms of practical reasonableness. If there is a viewpoint in which the institution of the Rule of Law, and compliance with rules and principles of law according to their tenor, are regarded as at least presumptive requirements of practical reasonableness itself, such a viewpoint is the viewpoint which should be used as the standard of reference by the theorist describing the features of the legal order”; Finnis, supra, note 3, at 15. Michael Moore essentially follows this line as well; though formulating the natural law thesis in terms of a necessary constraint on law by moral reasons, he explicitly counts all normative reasons as moral reasons. See Moore, Law as a Functional Kind, in NATURAL LAW THEORY 188–242, 189, 196–197 (Robert P. George, ed., 1992).

Nor is inseparability a sufficiently precise concept to get at what the natural law theorist has in mind:19

Leslie Green, Legal Positivism, in THE STANFORD ENCYCLOPEDIA OF PHILOSOPHY (Edward N. Zalta, ed., Spring 2003) available at http://plato.stanford.edu/archives/spr2003/entries/legal-positivism/.

The natural law theorist is concerned with a necessary continuity between law and the requirements of practical reasonableness. Positively put, the natural law thesis is that, necessarily, law is a rational standard for conduct. It is of the nature of law to provide a set of standards that rational agents should take as a guide to their conduct.

This first cut at the natural law thesis is rough, but it has the benefits of closely following Aquinas' formulation of the position20

SUMMA THEOLOGIAE IaIIae Q. 90, A. 4; IaIIae Q. 91, A. 3.

and would be affirmed by contemporary writers self-consciously working in this tradition. What has separated writers in this tradition is how the natural law thesis ought to be understood and developed. The central difficulty has been to provide a clearer understanding of the natural law thesis that is both interesting—that is, an understanding that is or would be denied by some otherwise sensible legal theorists—and not obviously false.

One way to understand the basic natural law thesis is what I will call the strong reading of the natural law thesis. According to the strong reading, the fact that it is of the nature of law to provide a set of standards that rational agents should take as a guide to their conduct entails that any standard that rational agents could not take as a guide to their conduct is not law but is simply invalid. Lex iniusta—or, better lex sine rationem—non est lex. While there have been some doubts expressed as to whether anyone actually held this view,21

Norman Kretzmann notes that no occurrence of the sentence lex iniusta non est lex appears either in Aquinas or in Augustine, whom Aquinas cites in introducing the idea into his discussion of law; Kretzmann, Lex Iniusta Non Est Lex: Laws on Trial in Aquinas' Court of Conscience, AM. J. JURIS. 99–122, 100–101 (1988). Of course, this does not at all settle the question of whether anyone in that tradition affirmed the rather stark proposition about the connection between law and practical reasonableness that the slogan suggests. Finnis expresses doubts about whether anyone in the natural law tradition affirmed it, and Finnis, George, and Soper are particularly confident that Aquinas did not mean to assert so stark a claim. Their main reasons for doubting that Aquinas held this view was that they find the view incoherent; on this alleged incoherence, see below in this paper.

it does seem to have been pretty clearly affirmed by Blackstone, and Michael Moore's recent sketch of the core natural law position includes the strong reading as the target at which the natural law theorist is aiming.22

Moore, supra note 18, at 194–195; see also Moore, Law as Justice, 18 SOC. PHIL. & POL'Y 115–145, 115–117 (2001).

Law that cannot serve as an adequate guide to conduct for a rational being is no law at all. Critics—even those otherwise sympathetic to some reading of the natural law thesis—have been very exuberant in their zeal to show that this sort of view is on its face paradoxical or otherwise deeply implausible. Finnis, George, and Soper have charged that the lex iniusta non est lex slogan expresses an absurd view—literally, “unjust law is not law”—that carries its self-contradiction out in the open and hence should not be considered an accurate statement of the natural law position. Finnis has argued that the natural law motto that unjust law is not law is, construed literally, “pure nonsense, flatly self-contradictory”;23

Finnis, supra, note 3, at 364.

Soper has written that “the very obviousness of this contradiction” shows that no one could ever have meant to affirm the strong natural law thesis;24

Philip Soper, Legal Theory and the Problem of Definition, 50 U. CHI. L. REV 1170–1200, 1181 (1983).

and George has remarked that the fact that Aquinas was perfectly willing to talk about unjust laws shows that the paradigmatic natural law position does not affirm the lex iniusta thesis.25

Robert P. George, Kelsen and Aquinas on “the Natural Law Doctrine,” 75 NOTRE DAME L. REV. 1625–1646, 1641 (2000).

But none of this is at all persuasive. For, first of all, the core of the strong view can remain without the paradoxical formulation. This is obvious. All that one needs to do is to restate the position as the claim that no norm or social rule (etc.) that is unreasonable can be law. This lacks even the appearance of paradox to which the critics object.

But for all that there may be good reasons to stick with the formulation and to reject the view that it is at all paradoxical. Norman Kretzmann has made one case for this view in his exposition of Aquinas' position.26

Kretzmann, supra note 21, at 102–107.

Kretzmann has defended the lex iniusta slogan by noting that it is a common phenomenon for one term to have two sets of conditions of application, one of which is nonevaluative, the other of which is evaluative. So one might claim that this doctor is no doctor at all; or that one's son is no son at all. In each of these cases, the correct application of the former term depends entirely on nonevaluative conditions (having the socially recognized credentials of physicians for the former, and perhaps being one's male biological offspring or being legally recognized as a male dependent of a certain sort), while the correct application of the latter term depends at least in part on evaluative conditions (having a proper care and competence with respect to furthering health or showing the proper sort of care for and deference to one's parents). The crucial point to be made here is that this is not merely a matter of equivocation: the former and latter conditions of application are nonarbitrarily related to one another. Kretzmann does not elaborate sufficiently on the nature of this nonarbitrary relationship, leaving his view open to the charge that the sense in which the lex iniusta claim is true may be a sense that everyone, including the most hard-boiled positivists, can accept.27

J.S. Russell, Trial by Slogan: Natural Law and Lex Iniusta Non Est Lex, 19 LAW & PHIL. 433–449 (2000).

(I will return to Kretzmann's view below.)

There is another way to respond to the charge. A claim of the form “a——X is not an X” is not self-contradictory—even assuming it to have existential import—if the blank is filled by an alienans, a certain sort of attributive adjective. “Fake” is always an alienans, and so “fake diamonds are not diamonds” is not self-contradictory: “fake diamonds are not diamonds” is not properly analyzed as “nothing that is both fake and a diamond is a diamond.” But there are adjectives that count as instances of the alienans only in certain contexts, that is, as applied to certain nouns. “Glass” is not generally an alienans (a glass slipper is a slipper), but it can be (a glass diamond is not a diamond). A natural law theorist who takes the strong view could hold that “inadequately serving as a rational standard for conduct” is, when applied to law, an alienans, and thus escape the charge that the strong version of natural law jurisprudence is flatly self-contradictory.28

For a brief discussion of the alienans and its status as an attributive adjective, see Peter Geach, Good and Evil, 17 ANALYSIS 33–42, 33–34 (1956).

Even if the strong reading of the natural law thesis can be defended from the charge that its central thesis is paradoxical, it is nonetheless subject to the accusation that it is open to other obvious, devastating objections. Here, for example, is Brian Bix:

The basic point is that the concept of “legal validity” is closely tied to what is recognized as binding in a given society and what the state enforces, and it seems fairly clear that there are plenty of societies where immoral laws are recognized as binding and enforced. Someone might answer that these immoral laws are not really legally valid, and the officials are making a mistake when they treat the rules as if they were legally valid. However, this is just to play games with words, and confusing games at that. “Legal validity” is the term we use to refer to whatever is conventionally recognized as binding; to say that all the officials could be wrong about what is legally valid is close to nonsense.29

Brian Bix, Natural Law Theory: the Modern Tradition, in OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW 72–73 (Jules Coleman & Scott Shapiro, eds., 2002).

The Fugitive Slave Act of 1850, to take one example, required citizens not to hinder and even to aid federal marshals who sought to return runaway slaves to bondage. This act was passed in order to enforce a constitutional provision and was enacted in due form by the federal legislature. It was socially acknowledged and judicially enforced. It seems that as a matter of social practice, the Fugitive Slave Act was law—regardless of the fact that those under it did not have anything like decisive reason to comply with it. It hence serves nothing but obfuscation to deny—as the defender of the strong reading must deny—that the Fugitive Slave Act was law.

It still seems to me an inconclusive objection. We should grant, of course, that if it were a criterion for success in any account of law that it designate as “law” all those things that are designated “law” by citizens or perhaps by officials, then this understanding of the natural law view would be doomed, for it undoubtedly does deny the designation “law” to some of those items. But the question to be raised is whether this general agreement is to be treated as infallible. Consider, as an instructive analogy, van Inwagen's story of the “bligers”:

When the first settlers arrived in the hitherto unpopulated land of Pluralia, they observed (always from a fair distance) what appeared to be black tigers, and they coined the name “bliger” for them…. A few centuries after the settlement of Pluralia, however, a foreign zoological expedition discovered that, in a way, there were no bligers. “A bliger (Quasi-Tigris Multiplex Pluralianus),” their report read, “is really six animals. Its ‘legs’ are four monkey-like creatures, its ‘trunk’ a sort of sloth, and its ‘head’ a species of owl. Any six animals of the proper species can combine temporarily to form a bliger. (Combinations lasting for several hours have been observed telescopically.) The illusion is amazing. Even a trained zoologist observing a bliger from a distance of ten meters would swear he was observing a single, unified animal. While the purpose of this combination is doubtless to protect its members from predators by producing the illusion of the presence of a large, dangerous carnivore, we can only guess as the evolutionary history of this marvelous symbiosis.30

Peter van Inwagen, MATERIAL BEINGS 104 (1990).

Now, it is not perfectly clear what moral to draw from this story. Van Inwagen draws the moral that Pluralians nonetheless spoke truly when they said “there is a bliger in the back field!” and the like; in saying “there is a bliger in the back field!” the Pluralians did not express a view on whether the various objects arranged bligerwise composed an additional object, a bliger. If I understand Trenton Merricks correctly, he would hold that the Pluralians spoke falsely when they said “there is a bliger in the back field!”: Though what they said is nearly as good as true and good enough for normal practical purposes, what they said is nevertheless simply false.31

Trenton Merricks, OBJECTS AND PERSONS 162–185 (2001).

Whichever way one goes in reading this fable, it seems to me that there is room for one to make the sensible claim—as van Inwagen does—that really there are no bligers. By van Inwagen's lights, what needs to be done to give his assertion sense and to distinguish it from the Pluralian folk's way of speaking is to provide a gloss on his claim. By Merricks's lights, the assertion really—if we were uncorrupted—requires no gloss; what requires explanation is why the Pluralians are less confused in saying “there is a bliger in the back field” than in saying “there is a unicorn in the back field.” On either view, one can make sense, in the context of van Inwagen's story, of eliminativism about bligers.

What the strong natural law theorist should claim is that laws unsuitable to serve as rational guides to conduct occupy the role that bligers occupy in van Inwagen's story. Such rules have been recognized as law by citizens and officials and have been treated as binding as a matter of social practice. But that fact does not make citizens and officials infallible with respect to the philosophical problem of whether these rules insufficiently grounded in reasons are really laws. As with the case of the bligers, the strong natural law theorist can go one of two ways here. He or she can claim that while folks are perfectly right when they say that some unjust laws are laws, there is an important sense in which they are not laws. The central task for the strong natural law theorist taking this route is that of explaining what that sense is and showing that this sense is sufficiently interesting. On the other hand, the natural law theorist can claim that in the ordinary sense of law, law that it is not reasonable to comply with is no law at all. The task here is, I take it, that of showing that there are presuppositions of the designation of social rules as laws which could be brought to light by closer analysis and which could nevertheless turn out to be false. Closer inspection yielded the result that a bliger is not one animal but six; and given the centrality to the practice of bliger-talk that a bliger is one animal, a straightforward inference to draw is that there are not really any bligers. Closer inspection may yield the result that laws unbacked by decisive reasons for action lack some feature whose assumed presence is central to any practice that we would count as the practice of law. And if it turned out that this were the case, a straightforward inference to draw would be that there are not really any laws unbacked by decisive reasons for action.

Now, one might respond: Even if it is not incoherent for the strong natural law theorist to claim that laws unbacked by decisive reasons for compliance are no laws at all, there is a key disanalogy between the case of the bligers and the case of laws unbacked by insufficient reasons. We see the situation with the bligers and we recognize that something is amiss, recognize that there is at least potential tension between Pluralians' bliger-talk and what is the case with bligers. But we see the strong natural law theorist's purported claim about law—that it is necessarily a rational standard for conduct—and we are yet unmoved. Officials go on applying unjust laws (or recognizing that they are laws yet refusing to apply them), citizens go on obeying or reforming them, and there is no tension felt. Does this not show that even if there is conceptual room for the sort of claim that a natural law theorist might want to make, there must be in fact no basis for the view that law must be backed by such reasons?

No. For, first of all, the claim that the natural law theorist wants to make does not immediately imply that folks—ordinary folk or legal officials—cannot go on using the term “law” very much as they did before. This much is clear from (at least seemingly) revisionary metaphysical theories that do not of themselves include recommendations for changes in ordinary linguistic practice. Second, the fact that folks have remained unmoved by the claims of strong natural law theory does not show that the claims of strong natural law theory are false. Ordinary users of the language do not enjoy a final authority on the correctness of analyses of the terms they employ nor on the presuppositions of the practices that they are engaged in.

Again, think of the bligers. Suppose that a Pluralian is shown the facts about bligers yet continues to think that there are bligers. A philosopher might note that there are features of bliger-talk that show pretty clearly that it was essential to the designation of that mass in the back field (and others like it) as bligers that they be individual enduring objects rather than temporary animal collectives. Yet the Pluralians just might not see it. While how they use the term “bliger” fixes its reference, they do not enjoy some sort of infallibility, either individually or collectively, on how their use fixes its reference and whether on any given occasion they are applying that term correctly. The same holds of law. The starting point for marking out a set of phenomena as law is the practices of human agents, but that does not make those agents infallible about whether they are correct in thinking that any particular instance is a case of law.

For example, even if there were complete agreement among competent users of the language of law that certain instances were cases of laws and all other instances were not, this would not be sufficient to show that all of those instances are in fact laws. For there might be some platitude about law that is accepted, either explicitly or implicitly, by all of those users and that is absolutely central to the practice of law-talk but that, nevertheless, some of those instances fail to satisfy. Suppose it were true, for example, that all competent users of the language of law believe that A, B, and C are laws and that nothing else is, but all such competent users of the language of law also accept as a deep and crucial platitude about law that compliance with laws is morally obligatory. (This corresponds to the platitude about bligers that they are individual enduring objects, not temporary animal collectives.) If we are able to show that compliance with A is not morally obligatory, this would give us a basis to say that A really is not a law.

I acknowledge, of course, that Bix's point places a weighty burden on the strong natural law theorist. But this burden is no greater than that which falls on any philosopher when his or her view runs contrary to common opinion. The strong natural law theorist bears the burden of showing that it is central to law that it be backed by decisive reasons, and this burden is made weighty by the fact that this view commits him or her to the thesis that a number of socially sanctioned rules called by consensus “laws” are not really laws at all. But we knew this already. It is no criticism of a controversial philosophical position that the defender of that position needs a good argument for it.

Here is another way to respond to Bix's claim of incoherence. If Bix were right, then it would be a condition of the eligibility of a jurisprudential theory that, necessarily, if all of the legal officials in some society hold that X is law in that society, then the theory implies that X is law in that society. But neither unsophisticated Austinian nor sophisticated Hartian positivism satisfies this constraint. So Bix's argument fails through proving too much.

Re Austin's view: in Austin's general jurisprudence, every law is a command issued by a sovereign and backed by a sanction.32

John Austin, THE PROVINCE OF JURISPRUDENCE DETERMINED 21 (Wilfrid E. Rumble, ed., 1995) (1832), Lecture I.

A sanction is a credible threat of harm to a subject attendant on a violation of the order.33

Austin, supra note 32, at 22.

It follows from Austin's view that there is no law that is not backed by a sanction. But, possibly, all of the legal officials in some society might hold that some particular norm, a norm unbacked by a sanction, is law. If Austin's view is true, law without sanction is no law at all. Hence Austinian positivism violates Bix's constraint.

Re Hart's view: in Hart's view, whether something is law in a given society depends on whether it is recognized as such by the rule of recognition, the usually tremendously complex rule that guides legal officials in making, identifying, and applying law.34

H.L.A. Hart, THE CONCEPT OF LAW 94–95 (1994) (1961).

It follows from Hart's view that there is no law that is not acknowledged as such by the rule of recognition. But, possibly, all of the legal officials in some society might hold that some particular norm, a norm not acknowledged by the rule of recognition, is law. The rule of recognition might hold that if norm N was part of the originally adopted constitution, then it is law; but they might all hold a false view about whether some particular norm n was part of the originally adopted constitution. If Hart's view is true, law unacknowledged by the rule of recognition is no law at all. Hence Hartian positivism violates Bix's constraint.

It might be objected that while all of the legal officials could be confused about what is acknowledged as law by the rule of recognition, they could not all be confused about what the rule of recognition is. I think that this is false. We may grant that the practice of legal officials makes the rule of recognition what it is. But because the rule of recognition is not something that legal officials need to be able to make explicit—the rule of recognition is typically shown rather than said35

Hart, supra note 34, at 101.

—it is possible for all legal officials to be deeply confused in their explicit judgments of what the rule of recognition is. All that is really justified in the end by the doctrine of the rule of recognition is that there is something about the actual practice of legal officials that fixes the content of the rule of recognition. But it is consistent with this position to hold that there is a necessary truth about whatever activities that we would be willing to call the “actual practices of legal officials” that would commit us to affirming the view that the rule of recognition cannot confer legal validity on any rule that is insufficiently backed by reasons for action. It may be false that there is any such necessary truth, but it is by no means incoherent to hold this.36

Indeed, it would suffice to show that Hartian positivism runs afoul of Bix's constraint that the Hartian doctrine of the rule of recognition implies either that all legal officials could be wrong about what the rule of recognition is or that all legal officials could be wrong about whether a particular instance is a law. Again, imagine a society in which all legal officials seem to be guided by the rule that if norm N was part of the originally adopted constitution, then it is law; all of them explicitly accept the rule “if norm N is part of the originally adopted constitution, then it is law”; but all of them falsely believe that some instance, n, is part of the originally adopted constitution; thus all of them believe that n is law; and all of them would, were they to learn that n was not in fact part of the originally adopted constitution, cease to say that n is law. (Assume that n is not recognized as law by any other element of the rule of recognition.) We have to say one of the following: either all of the legal officials are confused when they say that “if norm N is part of the originally adopted constitution, then it is law” is part of the rule of recognition, or all of the legal officials are wrong when they say that n is law. Both of these seem to run afoul of the constraint that, according to Bix, rules out the strong natural law thesis.

We will return to this strong formulation of the natural law thesis below; I will suggest that we ought to reject it, but not on account of its incoherence or its obvious falsity.37

It is sometimes put forward as a sign of the obvious falsity of the strong natural law thesis that it confuses law as it is with law as it ought to be. Now, if this were true, this would count as a serious criticism. But the strong natural law thesis commits no such confusion. For it is perfectly consistent for a defender of the strong natural law thesis also to hold that there ought to be a law on some matter but there is in fact not; the legislature might have neglected its responsibilities and failed to pass the requisite law. Perhaps, then, the objection in its slogan form is not to be taken seriously; perhaps it is better read as the objection that according to the strong natural law theorist, what ought not to be law is a subset of what is not law; if a measure ought not to be law, then the strong natural law view denies it that status. But this chastened objection misses the mark as well. For there may be norms that satisfy the strong natural law criteria and thus count as law but nevertheless ought not to be law. It could be that there is a norm that is a slight deviation from justice, either in its content or in its manner of adoption, but with which nevertheless there is now decisive reason for compliance. (Imagine a tax law that gave slightly more to the worse-off than was their due.)

The objection is groundless. The natural law theorist is interested in asserting a connection between the law's existence and the law's prescriptive force. The objection assumes that the natural law theorist is interested in asserting a connection between the law's existence and the desirability of its existence. The connections between a would-be rule's prescriptive force and the desirability of its existence are contingent. There is no way to transform the objection so that it applies to a recognizable version of the natural law view.

We ought to reject it just because the key argumentative strategies employed by natural law theorists fail to establish that thesis. These lines of defense of this strong understanding of the natural law thesis notwithstanding, it is obvious that contemporary advocates and friendly critics of natural law theory have, by and large, treated the strong natural law thesis as hopeless and perhaps even as a thesis that no natural law theorist has ever really been concerned to affirm. Soper writes that it is confusion to think of the classic natural law theorists as concerned with the tasks of analytical jurisprudence; they were instead concerned to provide a theory of political obligation and of the subject's moral relationship to law generally.38

Soper, supra note 24, at 1181.

Bix, in a number of accounts of natural law theories classic and contemporary, has affirmed much the same position.39

See, e.g., Bix, supra note 29, at 63.

This reading of the history of natural law jurisprudence is open to doubt,40

Again, the reason to doubt that this is a correct interpretation of the natural law tradition as a whole is that it is false of Aquinas' view, and Aquinas is the paradigm natural law theorist. Aquinas' conclusion about the status of lex iniusta is the result of not a primarily practical investigation but a primarily speculative one—it is a straightforward inference from the fact that human law is a kind of law, and law in general (including the eternal law, which is for the most part unknowable and as such of little practical interest to us) is a rational standard. No doubt the discussion of law is embedded in a section of the SUMMA THEOLOGIAE on ethics, but that does not make the propositions about law in the Treatise on Law practical propositions, any more than it makes the propositions about the nature of virtue (SUMMA THEOLOGIAE, IaIIae QQ. 49–56) practical propositions.

but even if it were acknowledged as true, it would leave behind the task of providing a clearer understanding of what it is that the natural law theorist wants to assert.

Robert George has proposed that “What is being asserted by natural law theorists [is]… that the moral obligatoriness which may attach to positive law is conditional in nature.”41

Robert P. George, Preface, in THE AUTONOMY OF LAW viii (Robert P. George, ed., 1996).

All that the natural law theorist wants to do in affirming a connection between law and reasons is to issue a reminder that adherence to some laws would constitute such a departure from reasonableness that there could not be adequate reason to obey them; the only law that merits our obedience is law that meets a certain minimum standard of reasonableness. We can call this the moral reading of the natural law thesis, and Bix and Soper have agreed with George in holding that this is the point that classical natural law views meant to emphasize. The main problem with this reading is, as Bix notes, that it makes the natural law thesis excruciatingly uninteresting.42

See Bix, Patrolling the Boundaries: Inclusive Legal Positivism and the Nature of Jurisprudential Debate, 12 CAN. J. L. & JURIS. 17–33, 30 (1999); see also Bix, On the Dividing Line Between Natural Law Theory and Legal Positivism, 75 NOTRE DAME L. REV. 1613–1624, 1620, n. 34 (2000).

It is not merely that the natural law theorists would have no basis to disagree with the legal positivists, for whom it has been a central point to emphasize that the rightness of compliance with law depends on an evaluation of the law's merits.43

See Jeremy Bentham, A FRAGMENT ON GOVERNMENT ch. iv, § § 18–22 (Ross Harrison, ed., 1988); see also Hart, supra note 16, at 50–56.

If the moral reading were all there is to the natural law thesis, the natural law theorist would have almost no one to disagree with in the entire history of philosophy.

Given the reluctance of contemporary natural law theorists to affirm a strong reading of the basic natural law thesis, and given the trivial dullness of the moral reading of the natural law thesis, it is of course worthwhile to ask whether there is a third formulation—one that grants that the Fugitive Slave Act really was law without saying merely that it was a law that ought not to be obeyed. There seems to be. Recall again the basic natural law thesis: necessarily, law is a rational standard for conduct. The defender of the strong reading understands this thesis as of the same sort as necessarily, triangles have three sides. From necessarily, triangles have three sides we can deduce that if X does not have three sides, then X is not a triangle; and from necessarily, law is a rational standard for conduct we can deduce that if X is not a rational standard for conduct, then X is not law. The defender of the weak reading of the natural law thesis, by contrast, does not hold that necessarily, law is a rational standard for conduct is a proposition of the same sort as necessarily, triangles have three sides: rather, it is of the same sort as necessarily, the duck is a skillful swimmer. From necessarily, the duck is a skillful swimmer we cannot deduce that if X is not a skillful swimmer, then X is not a duck; we can deduce no more than if X is not a skillful swimmer, then X is not a duck or is a defective duck. The necessity attaches not to individual ducks but to the kind duck; and while it is possible for a duck-instance to lack the feature of being a skillful swimmer, the absence of that feature marks it as defective.44

Cf. Michael Thompson, The Representation of Life, in VIRTUES AND REASONS 247–296 (Rosalind Hursthouse, Gavin Lawrence & Warren Quinn, eds., 1995); and Philippa Foot, NATURAL GOODNESS 20 (2001).

Finnis suggests this sort of move in Natural Law and Natural Rights: His preferred way of putting the point is that some law is law in the focal sense, whereas some law is law in a secondary, peripheral sense.45

Finnis, supra, note 3, at 364.

Hence Finnis writes that attention to the principles of practical reasonableness that govern human conduct “justifies regarding certain positive laws as radically defective, precisely as laws, for want of conformity to those principles.”46

Finnis, supra, note 3, at 24, emphasis in original.

It also seems to be in the spirit of Kretzmann's reading of the natural law thesis. Kretzmann, in order to make the lex iniusta non est lex claim interesting, must hold that “lex” is not used merely equivocally here. One way to pull this off is by making the claim, following Finnis, that the latter sense is somehow primary, whereas the former sense is derivative, truncated, or incomplete. Some of Kretzmann's examples tend to distract one from this point: “a badly disobedient son is no son at all” seems to lean far more toward metaphor than does “an entirely incompetent doctor is no doctor at all.”47

Kretzmann, supra, note 21, at 102–104.

This gives support to Russell's charge against Kretzmann that his interpretation of the lex iniusta slogan is not just available for natural law theorists but open to adoption by anybody who wants to trade on positive connotations of “law” to use the slogan to criticize unjust legal systems or unjust individual legal norms.48

Russell, supra, note 27, at 446.

But while it does seem right that Kretzmann's official statement of his position does provide some ammunition for Russell's criticisms, the overall thrust of his view seems to be toward the weak reading of the natural law thesis, though he does not himself go very far in showing why the weak reading should be taken as correct.

The weak reading of the natural law thesis is clearly distinct from the strong reading, allowing that there can be laws with which it is unreasonable to comply. But one might wonder whether it is really distinct from the uninteresting moral reading. Hence Bix, who puzzles a bit over why anyone would think that the moral reading of the natural law thesis is anything but “banal,” immediately identifies the moral reading with the view that immoral law is a perversion of law or defective as law.49

Bix, Patrolling the Boundaries, supra, note 42, at 1620, n. 34; see also Bix, On the Dividing Line, supra, note 42, at 30.

But this identification is illegitimate. The weak reading of the natural law thesis does not say simply that some laws might fail to be adequate rational standards and that this is in some way objectionable; it takes the further step of saying that this way of being objectionable counts as a defect in law. The standards for counting something a defect are far more stringent than those for counting something objectionable. To count the absence of a feature as a defect in some thing, one must show that it is intrinsic to the kind to which that thing fundamentally belongs to possess that feature. We might object to a particular coloring pattern in a duck's feathers on aesthetic grounds, but that objection would not suffice to show that the pattern counts as a defect in the duck. We might find the duck's propensity to leave its droppings around ponds objectionable, but that would hardly count as making ducks that leave their droppings around defective ducks. On the other hand, a duck that cannot fly or swim is defective, regardless of whether a duck's inability to fly or swim suits our own purposes.

To affirm that the moral reading is the proper understanding of the natural law thesis would be the end of natural law theory as an interesting jurisprudential view. The strong reading, while often quickly dismissed, even by those sympathetic to natural law theory, possesses adequate resources to fend off the most straightforward objections and is hence worthy of further scrutiny. And the weak reading, since it is entailed by the strong reading and distinct from the moral reading, must be worthy of further scrutiny as well. We may hence turn to the task of seeing what sort of arguments have been put forward for the fundamental natural law thesis, seeing what success those arguments have had, and seeing whether any such success militates only in favor of the weak reading or also in favor of the strong reading as well.

DEFENDING THE NATURAL LAW THESIS

There are three interesting and initially plausible50

I put to the side the defense of natural law theory, popular in the mid-twentieth century, of arguing on behalf of natural law theory and against positivism that only natural law theory can serve as a bulwark against abuse of law. Cf. Gustav Radbruch, Fünf Minuten Rechtsphilosophie (1945) and Gesetzliches Unrecht und übergesetzliches Recht (1946), both reprinted in RECHTSPHILOSOPHIE 327–329, 339–350 (8th ed., Erik Wolf & Hans-Peter Schneider, eds., 1973); cf. also Lon Fuller, Positivism and Fidelity to Law: A Reply to Professor Hart, 71 HARV. L. REV. 630–672 (1958). The proper criticism and reform of statutes can intelligibly take place regardless of whether one is a positivist or a natural law theorist; whether one is better able to do so if one is a natural law theorist or a positivist is of merely psychological interest. For criticism of this approach to defending theories of the nature of law, see Philip Soper, Choosing a Legal Theory on Moral Grounds, 4 SOC. PHIL. & POL'y 31–48 (1987).

lines of argument toward the central natural law thesis: Finnis's “internal point of view” argument, Moore's functional-kind argument, and Raz's self-image-of-law argument.51

It is undoubtedly prima facie bizarre to think that Raz, who is a hard positivist, offers a defense of natural law jurisprudence. But Raz's hard positivism is a hard positivism only about legal validity—that is, for law's existence conditions—and, as is clear already and will be discussed further below, the weak natural law thesis is compatible with at least the canonical formulations of the hardest such positivisms out there.

I want to make clear the structure of each of these arguments and offer a brief assessment of the prospects of each. While Finnis's argument is ultimately unsuccessful, it does bring out a point that can be exploited for natural law purposes by the sorts of arguments that Moore and Raz employ. The Moorean and Razian arguments, however, do not provide support for the strong natural law thesis, though they suggest bases for affirming the weak thesis.

Finnis's Internal-Point-of-View Argument

Finnis's argument for the natural law thesis is that the natural law thesis drops out of Hart's and Raz's jurisprudential method, a method that has shown itself to be fundamentally sound. Finnis writes in praise of Hartian and Razian jurisprudence that Hart's and Raz's views were able to advance so far beyond earlier positivist views by their more or less self-conscious employment of three methodological features: attention to the practical point of legal systems, use of a focal-meaning approach to definition, and adoption of the viewpoint of those who take an insider's point of view.52

Finnis, supra, note 3, at 6–18. Finnis develops the importance of each of these separately, noting with respect to each that he is simply following themes explicit in Hart and Raz. Hence Finnis takes himself to be answering the same questions that Hart and Raz are trying to answer, and using a basically similar methodology. The difference is that on Finnis's view, Hart and Raz arbitrarily stop short of fully embracing that methodology's relevant set of implications.

Hart had argued against earlier positivist views that such views had failed to take into account the point of view of the person who takes the internal point of view with respect to a legal system, treating it as a standard by which he or she guides his or her conduct. So Hart's view privileges the internal point of view, and it is the point of view of one who treats the law as a standard for conduct. Hart is insistent that no further differentiation of the internal point of view is called for. People who treat the law as a basis for their conduct out of a calculation of long-term advantage, on a whim, out of altruistic concerns, out of the demands of morality, to please one's parents, to conform to time-honored tradition, and so on are all taking the internal point of view, and Hart is not interested in taking their different motives as shaping his theory of law.53

Hart, supra note 34, at 203.

Finnis's argument for the natural law thesis is to take Hart's starting point—that analytical jurisprudence must adequately take into account this insider's point of view—and to try to show that its characterization of the internal point of view is too undifferentiated, that it fails to take into account that some of these insiders' points of view are more paradigmatically insiders' points of view than others. By Finnis's lights, there is a clearly most central internal point of view with respect to the law:

If there is a point of view in which legal obligation is treated as at least presumptively a moral obligation…, a viewpoint in which the establishment and maintenance of a legal as distinct from discretionary or statically customary order is regarded as a moral ideal if not a compelling demand of justice, then such a viewpoint will constitute the central case of the legal viewpoint.54

Finnis, supra note 3, at 14–15.

But even within this central legal viewpoint we should recognize that

Among those who, from a practical viewpoint, treat law as an aspect of practical reasonableness, there will be some whose views about what practical reasonableness actually requires in this domain are, in detail, more reasonable than others. Thus the central case viewpoint itself is the viewpoint of those who not only appeal to practical reasonableness but also are practically reasonable.55

Finnis, supra note 3, at 15.

Law that fails to be morally obligatory will be viewed, from this central legal viewpoint, as defective, deficient, falling short. And since the central legal viewpoint is the proper vantage point from which to do analytical jurisprudence, we have a basis for holding that law that fails to serve as a mandatory requirement of practical reasonableness is defective precisely as law. So the weak natural law thesis is true.56

Finnis emphatically rejects the strong natural law thesis; see Finnis, supra note 3, at 363–365.

If we take for granted the fundamental soundness of Hart's approach, the key questions are first, whether Finnis has taken up what is in fact Hart's method, and second, whether Hart's method thus understood admits of arbitrariness if it stops short of the natural law thesis. The answer to the first, it seems to me, is that he has gone much further than Hart, who holds merely that analytical jurisprudence must characterize law in such a way that exhibits how it is possible for persons to take the internal point of view with respect to it. Finnis seems to want to make the further claim that the internal point of view is the privileged point of view with respect to the description of law. But it seems that this massive privileging of the internal perspective carries Finnis beyond the descriptive jurisprudence that he takes himself, along with Hart and Raz, to be practicing and into a more straightforwardly normative jurisprudence.57

One might wonder whether Finnis did take himself to be doing descriptive jurisprudence. The answer that he did could not be clearer. As I note above (n. 52), Finnis takes his views to be rival answers to the same questions posed by Hart and Raz, who are undoubtedly doing descriptive jurisprudence. It is also worth looking closely at sections 1.4 and 1.5 of Finnis, supra note 3, along with the accompanying notes. There Finnis holds his natural law jurisprudence to be, just as Hart described his own view (supra note 34, at vi), part of descriptive social science (Finnis, supra note 3, at 21); he distinguishes the fundamentally descriptive jurisprudence of Hart and Raz from the normative jurisprudence of Dworkin and aligns his own jurisprudential project with the former rather than with the latter. For Finnis, the fact that the descriptive theorist “needs the assistance of a general normative theory” (Finnis, supra note 3, at 21) does not render the resulting theory simply an exercise in normative jurisprudence.

Second, it seems that Finnis's argument here for the centrality of the point of view of the party who treats the law as presumptively obligatory is, to say the least, not self-sufficient. Hart repeatedly compares legal rules to rules of games, and the comparison is useful here as well. In understanding the rules of cricket, one needs to understand them not just from a third-person perspective but from the perspective of a participant in the game; one needs to understand how the rules of cricket function in the decision-making of players and officials in a cricket match. But it is not relevant why the cricketer takes the rules of cricket as a guide to his conduct. All the descriptive theorist need do is to provide an account of those rules that shows how it is possible for one to take such a stance with respect to the rules.

Now, law is not cricket, and so it might be that there is reason to privilege the view of one who treats the law as morally obligatory while there is no reason to privilege the view of one who treats the rules of cricket as governing his conduct for some particular reasons. But the explanation will have to be driven by something other than remarks about point of view; it will instead have to be driven by some features of law that distinguish its rules from the rules of cricket. Perhaps these will be further facts about the function of law in contrast to the function of cricket; or perhaps these will be further facts about the claims made by legal officials—those in a privileged position to speak on behalf of the law—that contrast with claims made by cricket officials. And both of these do seem to be fruitful points of departure for defenses of the natural law thesis.

Moore's Functional-Kind Argument

Michael Moore has suggested that the most promising route to the natural law jurisprudential thesis is through an argument concerning the function of law.58

See Moore, supra note 18; Moore, supra note 22.

When characterizing the nature of law, writers have often thought that law is to be defined in terms of some set of distinctive structures. But Moore wants to say that it is far more likely that law is to be defined in terms of its function, by its serving some end. In order for this to be the case, Moore claims, we would have to find some distinctive goal that law serves—otherwise we would not be able to define law in terms of the function of serving that goal. If it turns out that there is such an end, then if it can be shown that law must be moral-obligation-imposing in order to promote this goal, we have a basis to say that there is a necessary dependence of law on moral obligation; law must be morally obligatory, and any norm that cannot be morally obligatory cannot be law. Indeed, Moore thinks that if the premises of this argument can be established, the conclusion would be the strong natural law thesis.

As Moore notes, there are all sorts of difficulties involved in making a plausible argument that fits this schema. His constraint on definition by functional-kind-membership generates, on his view, a dilemma for the natural law theorist: Either the attempt to define law in terms of its serving some end will fail or the attempt to show that the law must be moral-obligation-imposing in order for it to serve this end will fail. The problem is this: In order for law to be defined by its serving some end, that end must be distinctive—it must be an end that is served only through or by law. So the goal that law serves cannot simply be “everything that is worth pursuing and promoting.” But, Moore wonders, how can anything short of “everything that is worth pursuing and promoting” be the source of the moral obligation that is, on the natural law theorist's view, essential to human law?59

Moore suggests a tentative response to this dilemma: that the end that law serves, while not identical to “all the values there are,” is so connected to their realization that moral obligation must result. Moore takes Finnis's understanding of the common good, which is the sum total of those conditions that individuals can draw upon in order to realize their own choiceworthy conceptions of the good (Finnis, supra note 3, at 154), to be potentially such an end (Moore, supra note 18, at 223).

As the argument is laid out, the second horn of Moore's dilemma strikes me as unproblematic. Given the way that Moore has set out the law-as-functional-kind argument for the natural law thesis, it is not necessary that the source of the moral obligation to obey the law be identical with the goal that law serves. So one might hold that while there is some distinctive goal G that law serves, it is not the law's serving G that is alone sufficient to make law morally obligatory. It might be, however, that for G to be served, or to be served properly, folks must be under a moral obligation to obey the law; and this moral obligation might arise from various sources—consent, fairness with respect to the promotion of G, gratitude to the law for helping us to promote G, and so forth. If there is a genuine difficulty to which Moore's formulation of the second horn of the dilemma points, it is that of finding some end of law that can be promoted only through obligatory norms, regardless of the source of that obligation. Suppose, for example, that serving the Finnisian common good were the function of law. It is obvious that this can be served other than through means that impose obligations. Moore mentions that a regime of sanctions might do the trick.60

Moore, supra note 18, at 225.

Or a set of common standards that did not impose obligations might be sufficient in a community where citizens were extremely public-minded and extremely conformist. Their public-spiritedness and conformism could be sufficient to lead them to act on a common standard.

More troubling is the difficulty that it just seems obvious that there is no good “distinctively served by law” in Moore's sense. There is no good that is served only by institutions that could by any stretch of the imagination be thought of as legal systems. The Finnisian common good, Dworkinian integrity, whatever—all of these can be served by institutions that are obviously pretheoretically extralegal. Moore sees the problem and thinks that if this is true, then the upshot is that law cannot be a purely functional kind.61

Moore, supra note 18, at 223.

While I agree there is a real worry in the vicinity, part of the fault must lie with Moore's overly strict understanding of what makes something a functional kind. There is, so far as I can see, no reason to think that for something to be a functional kind it must be adequately marked off simply in terms of its serving some goal. Functional kinds are typically marked off by serving some goal through some characteristic activity. Hence functional ascriptions involve both ends and means; to say that X is a member of functional kind F is to say, in part, that its characteristic activity tends toward the realization of some particular end. Not every X whose characteristic activity tends toward the realization of the same end E belongs to the same functional kind, for their characteristic activities may be of such different sorts that they could not be placed in the same kind. Moore is obviously right that heart is a functional kind, that there could be hearts of various structures and made of various materials. But while the end of the heart is to circulate the blood, it is pretty clear that only objects whose characteristic activity is that of pumping can be classified as hearts.

What causes unnecessary trouble for Moore's argument is his spartan understanding of functional kinds in which such kinds are individuated entirely by the ends they serve. Given an understanding of functional kinds in which such kinds are individuated also by the characteristic activities of the members of that kind, it could be that it is law's characteristic activity for the sake of its end that provides the needed support for the natural law thesis. So one might say that while legal systems might promote various ends, all of these involve the imposition of order; but one might say that it is the characteristic activity of law to realize this end through the provision of rules with which agents have decisive reason to comply. This would give us reason to say that the or a function of law is to impose order by laying down rules with which agents have decisive reasons to comply. And hence the natural law thesis would take its warrant not from the end that law serves (as in Moore's view) but from the characteristic activity of law in serving this end.

How would one show that this is law's characteristic activity? As Moore suggests: Look at the various particular ways that systems pretheoretically designated as “legal” operate and see whether their activities tend to be explicable in terms of and regulated by the giving of dictates backed by decisive reasons for the sake of imposing order. Look at the features of legal systems to which Raz has drawn our attention, that is, that they claim to be authoritative62

See Raz, THE AUTHORITY OF LAW 30 (1979).

and that, characteristically, their dictates go with the flow of normative reasons rather than against them.63

This is Raz's “service” conception of authority: see Raz, THE MORALITY OF FREEDOM 56 (1986).

Look at the way in which law characteristically ties sanctions to certain activities in order to give agents further reason to abstain from them. Look at Fuller's eight ways to fail to make law; each of them indicates some way in which law can fail to serve as a reason for action for those living under it.64

See Fuller, supra note 1, at 39.

On the basis of such considerations, one might well come to the conclusion that it is part of law's characteristic activity to lay down norms with which agents will have sufficient reason to comply.

Moore sets up the functional-kind argument as an argument for the strong natural law thesis (though he does not consider the weak natural law thesis as an alternative). But it seems false to suppose that, whether on Moore's functional-kind argument or on the emendation I suggested, the strong natural law thesis would be the result. The law cannot carry out its function if it is not backed by decisive reasons for compliance, on this view, but why would we think that there is no law unbacked by decisive reasons for compliance rather than merely that all such law is defective? There is, after all, nothing more ordinary than things that have the function of Φ-ing but which at the moment are not Φ-ing and in their present condition cannot Φ: witness broken alarm clocks, broken arms, and so on. A broken alarm clock is an alarm clock; it is just a defective alarm clock. To have one's arm broken in a skiing accident is not to lose (or even just temporarily misplace) an arm in a skiing accident. The functional-kind argument should aspire to no more than the weak natural law thesis.

Raz's Self-Image Argument

Raz is a positivist, but it seems to me that his work can be conscripted for natural law causes. (I am not alone in this suspicion; Goldsworthy65

Jeffrey D. Goldsworthy, The Self-Destruction of Legal Positivism, 10 OXFORD J. LEGAL STUD. 449–486 (1990).

and Kramer66

Matthew Kramer, IN DEFENSE OF LEGAL POSITIVISM: LAW WITHOUT TRIMMINGS (1999).

have made similar suggestions.) The way that Razian jurisprudence can be co-opted for natural law theory is by appeal to the cornerstone of Raz's legal theory, that is, that the law's self-image is that it is authoritative. On Raz's view, law necessarily claims to be a practical authority. For law to be authoritative would be, in part, for law's dictates that those in class C Φ to provide protected reasons67

A reason to Φ is a protected reason if it is a reason to Φ and a reason to disregard reasons not to Φ. See Raz, supra note 62, at 18.

for those in class C to Φ. Now, it seems to me that this view provides evidence for the weak natural law thesis, that is, the thesis that law that fails to serve as a standard of conduct for those rational agents under it is defective. For a thing to be defective is for it to fail to satisfy a standard that is internal to the kind to which it belongs. But that law essentially makes a claim to authority suffices to indicate that providing a particularly important sort of reason for action is a standard that is internal to the kind law. And so Raz's view that law necessarily claims authority entails that law that is not authoritative is defective and hence that something like the weak natural law thesis is true.

Why think that the fact that law necessarily claims authority would show that being authoritative is a standard internal to legality? Well, if the fact that law necessarily claims authority shows that being authoritative is an appropriate standard by which legal norms are measured, that standard is surely internal to the kind law rather than imposed on it from without, for after all, it is law's self-image that it is authoritative. That it is making this claim for itself suffices to meet the condition that the standard be relevantly internal, the sort of standard of which the failure to attain it would count as a defect. But why think that being authoritative is shown to be a standard for law at all? Because law's claim to authority is not just an interesting fact that it is reporting about itself, just as I might claim to be able to slam-dunk (which is false) or might claim to be from Texas (which is true). Rather, law's claim to authority is made in the context of justifying its other activities, its activities of laying burdens on citizens and punishing those that fail to comply, of rendering decisions on allocations of goods and putting to the side rival ways to allocate those goods. Hence its being authoritative is not just a feature it has self-reported but a standard to which it has held itself accountable. Because it holds itself to this standard, it can rightly be treated as a defect in law if it fails to be authoritative. Raz's views about law's essential claim to authority, which underwrite Raz's hard positivism about legal validity,68

See Raz, Authority, Law, and Morality, 68 MONIST 295–324, 315 (1984).

can therefore also be used to underwrite a weak formulation of the natural law view.

I will not here enter into the debate on whether Raz is right that law necessarily makes this claim to authority.69

See, e.g., Kramer, supra note 66, at 83–89; see also Philip Soper, Law's Normative Claims, in THE AUTONOMY OF LAW: ESSAYS ON LEGAL POSITIVISM 215–248, 229–240 (Robert P. George, ed., 1996).

I will, however, note that it is not at all clear that one need go as far as Raz in order to show that laws that are not decisively backed by reasons for compliance are defective as laws. Suppose that we allow that Kramer is right that law does not necessarily claim for itself practical authority, for there is, Kramer writes, a possible legal system in which all mandatory norms are stark imperatives, simply the demands of law.70

Kramer, supra note 66, at 83–89; see also Hart, Introduction, in ESSAYS IN JURISPRUDENCE AND PHILOSOPHY 1–18, 10 (1983): “It seems to me unrealistic to suppose that judges in making statements of legal obligation must always either believe or pretend to believe in the false theory that there is always a moral obligation to obey the law. It seems to me that such statements may be better construed as stating what may be properly demanded of their subjects by way of action according to the law which the judges accept as setting the correct standard of legal adjudication and law enforcement” (emphases in original).

Even in such a case we might think that demands of law that are not backed by decisive reasons for compliance are defective. Here is why: It is standard in speech-act theory to distinguish between success conditions, the conditions under which a speech-act is performed, and nondefectiveness conditions, the conditions under which all the presuppositions of a speech-act are satisfied.71

See John Searle and Daniel Vanderveken, FOUNDATIONS OF ILLOCUTIONARY LOGIC 12–13 (1985).

Now, it is plausibly a presupposition of the illocutionary act demanding that A Φ that, upon receiving the demand, A has decisive reasons to Φ. The most straightforward way of arguing for this claim is the paradox test: It seems to be pragmatically inconsistent for one to demand that A Φ while allowing that A might perfectly reasonably refrain from Φ-ing.72

I make this argument at greater length and for a different purpose in AN ESSAY ON DIVINE AUTHORITY 24–29 (2002).

So even if mandatory legal norms were mere demands, we would have some basis to think that a mandatory legal norm insufficiently backed by reasons for compliance is defective precisely as law.

It is pretty clear that neither the Razian argument nor the illocutionary-act variation on it would provide any basis to affirm the strong natural law thesis. So none of the plausible routes to natural law theory leads to the strong natural law thesis; they lead to but not beyond the weak thesis. While the initial objections to strong natural law theory can be avoided, the view fails simply for lack of evidence in its favor. But the weak natural law view is distinctive and defensible. The kind law may well be necessarily connected to reasons for action even if individual legal systems and individual laws can be unreasonable in the extreme.

IS THERE SUBSTANTIAL DISAGREEMENT WITH THE POSITIVISTS?

It is obvious that the strong reading of the natural law thesis is incompatible with legal positivism; it is the strong thesis that the positivists were concerned to deny. It is obvious that the moral reading of the natural law thesis is compatible with legal positivism; the positivists have taken as a central part of their program the emphasis on the need to scrutinize the merits of laws to determine whether they are worthy of obedience. What, then, about the weak reading? Is it contrary to the letter or spirit of positivism to affirm that law that is insufficiently backed by reasons for compliance is defective precisely as law?

It is not contrary to the letter of positivism. For if there is any canonical understanding of positivism, it is a thesis about legal validity; but the weak reading does not call into question the claim that whether law is valid is a matter of social fact. It claims that just as there are straightforward truths about when a duck or a heart is defective, there are straightforward truths about when laws are defective; and it claims that just as it is straightforwardly true that an adult duck that cannot fly is defective and that a heart that is fibrillating is defective, it is straightforwardly true that law that is not backed by adequate reasons for compliance is defective as well. Neil MacCormick, a positivist, describes and endorses this combination of positions in discussing Finnis's view:

Of course there may be legislation properly enacted by competent authorities which falls far short of or cuts against the demands of justice. The validity of the relevant statutory norms as members of the given system of law is not as such put into doubt by their injustice. The legal duties they impose, or the legal rights they grant, do not stop being genuinely legal duties or legal rights in virtue of the moral wrongfulness of their imposition or conferment. They are, however, defective or substandard or corrupt instances of what they genuinely are—laws, legal duties, legal rights.73

Neil MacCormick, Natural Law and the Separation of Law and Morals, in NATURAL LAW THEORY 105–133, 108 (Robert P. George, ed., 1992), emphasis added.

The weak natural law thesis may, however, be contrary to the spirit of positivism. For if the weak natural law thesis is true, it follows that one cannot have a complete descriptive theory of law without having a complete understanding of the requirements of practical reasonableness. For one cannot have a complete descriptive theory of law without an exhaustive account of the ways that law can be defective; and one cannot have an exhaustive account of the ways that law can be defective without having a complete understanding of the requirements of practical reasonableness.

That one cannot have an exhaustive account of the ways that law can be defective without having a complete understanding of the requirements of practical reasonableness is a pretty straightforward inference from the weak natural law thesis. But what is the warrant for claiming that there cannot be a complete descriptive theory of law without an exhaustive account of the ways that law can be defective? In any of those cases that are not at issue here, we would find very peculiar a theory of Xs that claimed to be a complete descriptive theory of Xs but did not offer an exhaustive account of the ways that Xs can be defective.

Suppose, for example, that I claimed to have an exhaustive descriptive theory of automobiles but freely acknowledged that I did not have an exhaustive account of when automobiles are defective and when they are not. Or suppose that I claimed to have an exhaustive theory of the kidney but freely acknowledged that I did not have an exhaustive account of when kidneys are defective and when they are not.74

One has not said nearly enough in providing a theory of automobiles if one has not provided an account of their function. But to commit oneself to a view of automobile function is to commit oneself to a view of automobile defectiveness. And so if one is to give a complete descriptive account of automobiles, it must include a complete account of the automobile's function, and to provide a complete account of the automobile's function, one must include a complete account of the automobile's defectiveness conditions. The same argument obviously applies mutatis mutandis to kidneys. (One might, of course, deny that the notion of defect really applies to natural objects like kidneys. But that is not what is at issue here. What is at issue is, given that the notion of defect nonvacuously applies to kidneys, whether a complete descriptive theory of kidneys could fail to provide a proper account of kidney defect.)

It seems perfectly obvious in these cases that a complete descriptive theory of the automobile or kidney would include a correspondingly complete theory of automobile or kidney defect. The burden of proof, then, seems to be on one who would hold that one can have a complete descriptive theory of law without a complete account of when and how law can be defective. But I have not the slightest idea how one would meet this burden.

Insofar, then, as positivism has presupposed a methodology that allows one to proceed further in jurisprudence without commitment to a particular conception of how agents ought individually and collectively to act, defenders of the weak natural law thesis must reject positivist methodology. As I noted at the beginning of this article, the natural law thesis in analytical jurisprudence can be formulated and initially defended without appeal to the particulars of a moral or political theory. But the weak natural law thesis, once defended, implies that a rich jurisprudence—even one that aims to be just descriptively adequate—cannot forego moral and political theory.75

Cf. Finnis, supra note 3, at 15–19.

ON NATURAL LAW THEORIES OF ADJUDICATION

I have claimed that there is an interesting natural law view in analytical jurisprudence that can be formulated in abstraction from the details of any natural law moral or political theory. It seems to me that while there are a number of plausible and promising views on offer that advertise themselves as natural law theories of adjudication, there is no basis to take any of them as the privileged statement of a distinctive natural law theory of adjudication. There is no theory of adjudication that can be both fairly called a natural law view and formulated in abstraction from the details of any natural law moral or political theory.

Consider the following sets of claims about adjudication.

  1. The meaning of a legal text is to be grasped by way of the real referents of the terms employed. So when moral terms such as “reasonable” or “cruel” or “due” are used, judges ought to look to the true nature of reasonableness, or cruelty, or “due-ness” in order properly to interpret the statute or constitutional provision.
  2. Whether judges ought to interpret constitutional or statutory provisions in accordance with legislator's understanding or the real referents of the terms employed is itself a legal matter to be determined by the particulars of the legal system in question and its understanding of the judicial role.
  3. The meaning of a legal text is to be grasped by way of some conventional meaning of the terms employed—in particular, the understandings of the legislator. So when moral terms such as “reasonable” or “cruel” or “due” are used, judges ought to try to give life to the legislator's intention in order properly to interpret the statute or constitutional provision.

Each of these views has been defended under the banner of natural law theory. (1) has been defended as a natural law view by Michael Moore in a number of papers;76

For Moore's view, see Michael Moore, The Semantics of Judging, 54 S. CAL. L. REV. 151–294 (1981); and A Natural Law Theory of Interpretation, 58 S. CAL. L. REV. 277–398 (1985). See also, for a similar view (though without the “natural law” labeling), David O. Brink, Legal Theory, Legal Interpretation, and Judicial Review, 17 PHIL. & PUB. AFF. 105–148 (1988). Moore's and Brink's views are critically appraised in Brian Bix, LAW, LANGUAGE, AND LEGAL DETERMINACY 133–177 (1993). It is Dworkin's affirmation of such views on adjudication that tempt some to label him a natural law theorist, for Dworkin has advanced a similar view, though without the metaphysically realist trappings of Moore's position (e.g., Dworkin, LAW'S EMPIRE, supra note 2, esp. 45–86); for Dworkin's critique of the need to say anything metaphysically freighted to account for the truth of the moral claims presupposed in Dworkinian adjudication, see Dworkin, Objectivity and Truth: You'd Better Believe It, 25 PHIL. & PUB. AFF. 87–139 (1996). While Dworkin's parochialism in jurisprudence (see n. 2) is enough to disqualify him as a defender of natural law jurisprudence, his antimetaphysical stance in ethics is enough to disqualify him as a defender of natural law ethics.

(2) has been defended within the natural law theory of Robert George;77

Robert P. George, Natural Law, the Constitution, and the Theory and Practice of Judicial Review, 69 FORDHAM L. REV. 2269–2283 (2001); George, The Natural Law Due Process Philosophy, 69 FORDHAM L. REV. 2301–2312 (2001); and George, Natural Law and the Constitution Revisited, 70 FORDHAM L. REV 273–282 (2002).

and (3) is defended by Hobbes, who describes his moral and political theory in natural law terms, in his account of the proper role of judges in a commonwealth.78

Hobbes, LEVIATHAN, ch. 26, ¶¶20–22 (Edwin Curley, ed., 1994). You do not find much of this today. In order to get Hobbes's strong conclusion, you need a natural law view on which legislative intentions are naturally supremely authoritative—that is, you do not have a proper commonwealth at all unless legislative intentions are supremely authoritative. Hobbes thinks sovereignty is like this. For a central reason for dissension outside of a political society is differing interpretations of the law of nature, and unless one person's (natural or corporate) judgment is taken as authoritative, superseding all other authoritative sources, the deficiencies that folks are trying to avoid by entering into political society will reappear. For more on Hobbes's natural law credentials, see Norberto Bobbio, HOBBES AND THE NATURAL LAW TRADITION (Daniela Gobetti, trans., 1993); my Was Hobbes a Legal Positivist? 105 ETHICS 846–873 (1995); and David Dyzenhaus, Hobbes and the Legitimacy of Law, 20 LAW & PHIL. 461–498 (2001).

Consider next the following sets of claims about adjudication:

  1. Given that a particular interpretation (or range of interpretations) of the law is fixed by the text, it is not the case that judges ought always to apply the law in accordance with that interpretation (or within that range of interpretations).
  2. Given that a particular interpretation (or range of interpretations) of the law is fixed by the text, within one possible legal system it is the case that judges ought always to apply the law in accordance with that interpretation (or within that range of interpretations), and within another possible legal system it is not the case that judges ought always to apply the law in accordance with that interpretation (or within that range of interpretations).
  3. Given that a particular interpretation (or range of interpretations) of the law is fixed by the text, judges ought always to apply the law in accordance with that interpretation (or within that range of interpretations).

(1) has been defended by David Brink as a “natural law theory of adjudication” in a paper attempting to reconcile natural law and positivist views;79

David O. Brink, Natural Law Theory and Legal Positivism Reconsidered, 68 MONIST 364–387 (1985). Heidi Hurd, though refraining from labeling the position a natural law view (because it does not take a position on the nature of legal validity), commits herself to a position like Brink's in MORAL COMBAT (1999).

(2) is affirmed by Robert George in his attempt to clarify the commitments of natural law theories;80

See Robert P. George, Natural Law and Positive Law, in THE AUTONOMY OF LAW: ESSAYS ON LEGAL POSITIVISM 321–334 (George, ed., 1996); see also Christopher Wolfe, Judicial Review, in NATURAL LAW AND CONTEMPORARY PUBLIC POLICY 157–189 (David Forte, ed., 1998).

and (3) is the classic Hobbesian view, on which the natural law directs judges to defer to the sovereign's commands.81

Hobbes, supra note 78, at ch. 26, ¶20. Even the very extreme Hobbes must be willing to allow for some exceptions here—if the mob, which cannot be checked by the sovereign's available forces, is pressing against the courtroom door, and the judge will be torn apart if he or she offers the interpretation of law fixed by the text handed down by the sovereign, the judge must on Hobbesian premises be free to save him- or herself by fudging the interpretation; Hobbes, supra note 78, at ch. 21, ¶¶11–15. If even the severe Hobbes lacks the resources to defend such a stark account of the judge's duty in applying the law, it is surprising that Brink, supra note 79, attributes to “the positivist” an account of judicial duty on which judges are always ultima facie bound to apply the law within the range of meanings fixed by the text.

Because each of the theses in these sets are incompatible with the others in these sets, at most one from each can be true. To assess the prospects of the natural law theory of adjudication, then, we would need to select which of these theses has a better claim to count as the natural law view than the others. The theses themselves do not, however, carry their natural law credentials on their faces. When we look to the arguments that have been offered in support of them, they are from a variety of sources: from philosophy of language, from moral arguments on the relevance of institutional roles, from consequentialist arguments, from political/legal values such as the rule of law and separation of powers, from general theories of the nature and form of political authority, and so forth. None of these seem to be distinctively nor even particularly “natural law” arguments.

It seems to me, then, that the “natural law theory of adjudication” label is useless, save for one purpose: to describe a theory of adjudication that is developed self-consciously with the resources of and within the constraints set by natural law theories of morality and politics. In this respect natural law theory in adjudication stands in stark contrast to natural law theory in analytical jurisprudence. In defending natural law theory in analytical jurisprudence, one can abstract from the details of natural law moral and political theory because one can raise an intelligible and interesting question about the presence of an internal connection between the kind law and the kind reasons for action apart from any specific view of what reasons for action there are and how those reasons shape the deliberation of a practically rational agent. So one can raise and offer considerations that militate in favor of or against the natural law view on the nature of law without committing oneself to affirming or rejecting a natural law (or a Kantian, or a utilitarian, or a Humean) theory of practical rationality.

But a natural law theory of adjudication cannot so abstract. Because a theory of adjudication is unabashedly a normative enterprise, centrally concerned with the asking and answering of normative questions, everything turns on the details of what reasons for action agents generally have and how those reasons for action bear on the person who occupies the judicial role. So there cannot be an interesting view that merits the title “natural law theory of adjudication” apart from a fairly well-worked-out natural law theory of practical rationality. There is nothing informative that I or anyone else can say about the prospects of a distinctively natural law theory of adjudication except that its prospects are those of a distinctively natural law theory of practical rationality.