II. Law and the Settlement of Moral Controversy

What is the principal function of law—all law? I would argue, and have argued,Footnote ² that law's principal function is to settle what we are obligated to do. More precisely, law's principal function is to settle what our moral obligations are that are properly subject to coercive enforcement. Even if we were all motivated to fulfill our enforceable moral obligations, our disagreements about what those obligations are and what they entail in particular situations would produce huge moral costs. Because of our moral disagreements, we could not predict with any certainty what others were going to do. And because what is the morally correct conduct for me often depends upon what others are doing, my inability to predict and coordinate with others will lead me to do what is morally wrong even if I am trying to do what is morally right. Legal settlement of what I and others are obligated to do allows me to coordinate my acts with others' acts and avert the moral costs of colliding good intentions.

Legal settlement also can improve moral decision-making through giving decision-makers the benefits of others' expertise. I may know that polluting is wrong, but I may not know whether the substance I am about to dump in the river is a pollutant. I may know that I should not operate equipment that endangers others, but I may be ignorant of the fact that this particular piece of equipment is dangerous. Good intentions without factual expertise can be as problematic, in terms of morally significant consequences, as bad intentions are. Legal settlement can improve our moral decision-making if the law makes use of expertise.

Legal settlement also makes moral decision-making more efficient in terms of time and other resources. Even if I could eventually acquire expertise and the ability to predict how others will act, my decisions regarding what I am morally obligated to do, if unaided by law, may require a great deal of time and energy that could be put to better use if there were a legal settlement of the issue that was clear and easily accessible.

Thinking of law in terms of settlement of what we are enforceably morally obligated to do explains how law and morality coexist on the same terrain. Both purport to tell us what we are obligated, forbidden, or permitted to do. So the question might be asked, “Why do we need law, given that we already have morality?” Why do we have many volumes chock-full of laws rather than only the one, Spike Lee law, “Do the right thing”?Footnote ³ The answer is that although the Spike Lee law is in one sense perfect, and would be appropriate for a society of omniscient gods, in another sense, in our world of less than omniscient humans, “Do the right thing” is not “the right thing.” Although we are neither gods nor angels, law is a response to the fact that we are not gods—not to the fact that we are not angels.Footnote ⁴ If we were omniscient gods, but some of us were not angels, “Do the right thing” would suffice. If one of us did the wrong thing, others, being omniscient, would know the right thing to do in response.Footnote ⁵ But if we were angels but not gods—always motivated to do the right thing, but uncertain of and disagreeing about the right thing to do—specific laws settling what we should do would effect a moral improvement over the Spike Lee law. It is paradoxical perhaps, but true nonetheless, that the perfect Spike Lee law is morally inferior to a regime of quite specific and likely imperfect laws.

Of course, to fulfill its settlement function and improve upon Spike Lee, law will have to consist of determinate rules. It is determinacy that produces settlement and its moral benefits of coordination, expertise, and efficiency.Footnote ⁶ Standards—those legal norms that are not determinate rules but rather instruct us to do what is “reasonable,” “fair,” or “just”—leave matters unsettled. In effect, they tell us to “do the right thing” within the region of decision-making subject to them. They fail to fulfill the settlement function. Rather, they defer settlement.

III. The Nature and Functions of Constitutional Law

If law's function is to settle what ought to be done through determinate rules, constitutional law's function is to settle the most basic matters regarding how we ought to organize society and government.Footnote ⁷ One can think of a constitution's being basic law through at least two different prisms. Looked at in terms of legal validity, constitutional law is the law that is highest in the validity chain. It validates lower-level law. Just as an administrative law may be valid only if it is authorized by a statute, a statute may be valid only if it is authorized by the constitution. The constitution is the highest law there is, and is neither valid nor invalid but just accepted.Footnote ⁸

The second way to think of a constitution's being basic law is to think of it in terms of relative entrenchment against change. Typically, a constitution is more entrenched against change than a statute, an administrative rule, or a common law court decision.Footnote ⁹

Usually the law that is highest in the validity chain is also the most entrenched. The reason for this is obvious. If lower-level laws like statutes were more entrenched than the higher-level constitutional laws that authorized such statutory entrenchments, one could repeal the entrenched statutes by the easier route of repealing the less entrenched constitution's authorization of the statutory entrenchment. So it makes sense to have higher-level law be more entrenched than lower-level law. Nevertheless, it is theoretically possible to have a lower-level law be more entrenched than the higher-level law that validates it.

Constitutions, as I said, lay down the ground rules for governance. They “constitute” the government. They set up structures, offices, and lawmaking procedures. Constitutions such as the U.S. Constitution and the Canadian Charter also entrench “rights.”

When a constitution promulgates the structures, offices, and procedures for governance—the rules about how laws are to be made, how offices are to be filled, and which level of government and which officials within government are to have jurisdiction over which matters—the constitutional provisions in question are similar to assembly instructions that accompany gadgets or toys that we buy. They are instructions for “how to assemble a government.” But if this is how we should conceptualize the structural provisions of the constitution, how should we conceptualize those provisions that constitutionalize certain rights?

As I see it, there are three, and only three, possibilities here. First, the constitutional authors might be creating rights or making certain rights more determinate by means of rules granting various specific liberties and immunities. Second, the constitutional authors might not be creating or specifying rights through rules but might instead be incorporating real moral rights as they exist in the moral realmFootnote ¹⁰—that is, making certain real moral rights legally as well as morally binding on the government and enforceable by the judiciary. Third, the constitutional authors might be inventing or creating rights but without translating them into determinate rules. I shall take up the implications of these three possibilities in turn.

The first possibility—creating a right, or making an existing right determinate, through a determinate rule—is unproblematic and unremarkable. In the U.S. Constitution, the so-called right against self-incrimination was most likely meant by the authors of the Fifth Amendment to be no more than a rule granting defendants an immunity from being compelled, on pain of contempt, to testify in court.Footnote ¹¹ It was an invented right, as there probably is no corresponding moral right against self-incrimination. And the scope of this invented right was coterminous with the rule that embodied it.

Likewise, some scholars believe that the “freedom of speech” referred to in the First Amendment was meant to be a determinate rule forbidding Congress from requiring licensing of speakers and publishers—that is, a determinate rule against prior restraints.Footnote ¹² If there is a general right of free expression, on this view, the First Amendment was making determinate and constitutionalizing only a portion of the more general right. Something similar might be said of the rights against takings of property without just compensationFootnote ¹³ and against impairing the obligations of contracts:Footnote ¹⁴ the constitutional provisions protecting these rights might be said to be rules that make determinate (and constitutionalize) portions of a more general moral right against upsets of legitimate expectations engendered by law.

Other rights, such as the right in the First Amendment to petition the government, might be seen as rules that are corollaries of the type of governmental structure created by the Constitution—a representative democracy—and not as moral rights that exist prior to that structure.Footnote ¹⁵

Let me now turn to the second possibility for constitutionalizing rights, namely, incorporating by reference real moral rights. Ronald Dworkin, for example, believes that this is what the authors of the U.S. Constitution were doing when they wrote into the Constitution a right against cruel and unusual punishments, a right to equal protection, a right to the free exercise of religion, and other rights.Footnote ¹⁶ There are three aspects of such an enterprise worth careful consideration.

First, if one is going to incorporate real moral rights into a constitution, then one is going to have to subordinate those rights to the various rules setting up the structure of the government and also subordinate them to decisions regarding their content that are supposed to be legally authoritative—that is, decisions that are supposed to settle the question of what the content of those moral rights is for purposes of the legal system.Footnote ¹⁷

Let me elaborate briefly on this point about subordinating moral rights. Within morality, moral rights are not subordinate to human institutions and human decisions. Rather, the contrary is the case. Human institutions and decisions are subordinate to moral rights. If moral rights were incorporated into a constitution without making them subordinate to other aspects of the constitutional regime, then they might completely overturn that regime.Footnote ¹⁸ For there is never any guarantee that a constitutional structure is consistent with moral rights as they really are. After all, the constitutional structure is an artifact produced by human beings possessing fallible moral beliefs.

Moreover, because the content of real moral rights is controversial, incorporating real moral rights into a constitution would make the content of constitutional law—and thus all law—incapable of being authoritatively determined. For the content of real moral rights cannot be settled by any institution's decision regarding that content, whether it is the decision of the legislature or the decision of the Supreme Court. Therefore, incorporation of real moral rights into a constitution will undermine the settlement function of law unless it is understood that those rights are legally (if not morally) subordinate to some institution's determination of their content. If that were not true, decisions of the legislature and the Supreme Court regarding moral rights would fail to be law in the eyes of those who disagreed with the decisions, because they would see those decisions as inconsistent with real moral rights.Footnote ¹⁹

Thus, the first point about incorporating real moral rights is that they must be subordinated to constitutional structures and to authoritative determinations of their content. The second point about incorporation of real moral rights is related: A decision must be made regarding which institution's view of real moral rights should be treated as authoritative for purposes of the legal system. To narrow the focus to the usual suspects, should the legislature's view of real moral rights be authoritative, or should the authoritative view be that of the courts?

Keep in mind that everyone—the legislature, the courts, and the people themselves—is subject to the requirements imposed by real moral rights. So the question is never whether the legislature is morally free to disregard real moral rights, or whether the courts are morally free to do so. The question is whose view of what those rights require should be the authoritative view within the legal system.

Some democrats believe that the legislature's view should be authoritative.Footnote ²⁰ However, because the legislature is always subject to the constraints of real moral rights, and can always tailor its legislation to the requirements of real moral rights as it perceives them whether or not those rights are incorporated into the constitution, it is pointless to incorporate them unless one plans to make them judicially enforceable against the legislature (whether or not the legislature is able thereafter to override that determination). I repeat: Constitutionalizing real moral rights only makes sense alongside judicial authority to determine their content and enforce them against the legislature. That is not because courts are superior to legislatures when it comes to determining the content of moral rights; rather, it is because legislatures are already supposed to make their legislation consistent with real moral rights, whether or not those rights are constitutionalized. Incorporating those moral rights as legal rights, but then making the legislature's view of their content legally supreme over the view of the courts, thus accomplishes nothing. (I take no view here on whether courts are superior to legislatures in determining the content of real moral rights.Footnote ²¹ I tend to be skeptical of either institution's ability in this regard, though no more skeptical than I am of law faculties' or philosophy departments' ability.) Nonetheless, as I said, if courts are not superior to legislatures in determining the content of real moral rights, either epistemically or motivationally, it makes no sense to constitutionalize those rights. They apply to the legislature whether or not they are constitutionalized.

So if real moral rights are to be incorporated into a constitution, they must be subordinated to the constitutional structures and to some institution's determination of their content. Therefore, an institution must be chosen that will have the authoritative say regarding that content, though incorporation of real moral rights strongly implies that the chosen authoritative institution will be the courts.

The third thing to note about incorporating real moral rights is that there is no guarantee that the moral realm as it actually is will contain the specific moral rights referred to in the constitutional text. There may not be any moral right of equality,Footnote ²² or of freedom of expression,Footnote ²³ or of freedom of religion.Footnote ²⁴ Or those rights may just be aspects of some moral right that is not named in the constitution. Or the correct moral theory might be a consequentialist one, like utilitarianism or egalitarianism, in which the only moral “right” is that all actions conform to the consequentialist norm. If constitutional authors wish to constitutionalize real moral rights, they had better be certain that the rights they name are real moral rights. But, of course, they cannot be certain.Footnote ²⁵ They might be better off just telling the courts to enforce against the legislature whatever moral rights there actually are, without attempting to name them. For in naming moral rights that do not exist, they might lead the courts to be more confused about what rights there are than would be the case if they had left moral rights unnamed.

I have now discussed two of the three possibilities constitutional authors might have in mind in constitutionalizing rights. They might be creating specific rights in the form of determinate rules that define the rights, such as a rule forbidding judicially compelled incriminating testimony or a rule forbidding requiring a license to speak. Or they might be attempting to incorporate by reference real moral rights rather than defining those rights through determinate rules or creating them. The third and final possibility is that in constitutionalizing a right, the constitutional authors are inventing or creating the right, but without giving it any determinate form—that is, without embodying it in a rule or set of rules. Rather, the right is supposed to function as a principle or value, with weight,Footnote ²⁶ not as a specific rule such as the rule against requiring a license to speak.

There is only one problem with this third possibility, but it is a doozy. This possibility is in truth an impossibility. One cannot, however hard one tries, create a right that is not coterminous with a determinate rule. If there is nothing in the world preexisting the constitution to which this right refers—if this right comes into being only by virtue of its being mentioned in the constitution—then its contours and weight cannot be assessed nonarbitrarily, as there is nothing in the world that would make any such assessment true. The courts would be making it up if they were to declare that such a constitutional right applied or did not apply, outweighed the government's interest in its legislation or did not outweigh it.

I shall give one illustration of the problem that will stand in for every possible illustration. Suppose that we wish to create a right that does not exist in the moral realm—say, a right to fine art. We do not, however, specify through determinate rules just what the right entails: Does it entail government funding of art, and if so, how much, and how is this funding to be balanced against other governmental resource needs? Does it entail a legally enforceable obligation of talented artists to produce such art, and if so, how is this obligation to be balanced against artists' right to liberty? And may government suppress fine art that threatens public order, and if so, of what magnitude must the threat be? Suppose that instead of resolving these questions through detailed determinate rules, we simply say that we will let the courts resolve these questions. Well, how will the courts resolve them other than according to judicial preference? There is no right to fine art in the moral realm which they can consult. When they resolve these questions, then, there will be no criteria available to determine if their resolutions are correct or incorrect. We made up the right, and they will have to make up its contours in deciding cases rather than looking to independently existing contours that could render their decisions right or wrong. They will be in the position of one who is told that a nonexistent creature—say, a unicorn—has a horselike body and a horn on its head, but is then asked to give its height, weight, color, and speed. One would protest, “This is an invented creature, and its inventor hasn't given me this information—so how would I know?” The same applies to invented rights not embodied in determinate rules.

My point is this: Rights that do not exist in the moral realm can only be created through and fully embodied in determinate rules. The enactment of nondeterminate “principles” that are supposed to function as weighty considerations cannot create rights and their weights, if those rights and weights have no pre-enactment existence. Put differently, there are no legal principles, legal norms that are neither determinate rules nor references to preexisting moral principles. If incorporation of actual moral principles is not what the enactment of legal principles represents, can legal principles be created merely through enactment? The answer is “no.” For there is no way to “create” by an act of human will a real principle, namely, a norm that is not a determinate rule with a canonical formulation but is instead a norm that possesses weight.Footnote ²⁷

To see that this is impossible, assume that there is, in the moral realm, no free speech principle. Then put yourself in the position of the lawmaker who wishes to create a free speech principle. How can he accomplish this (again, keeping in mind that there is no such moral principle for him to refer to and incorporate by reference). He could, of course, write out a set of instructions for how to apply the “principle,” but in that case he will merely have reduced it to a rule with canonical form that is either applicable or inapplicable but lacks weight. And I see no way that lawmakers can create weight, except by issuing instructions for how the principle applies in every conceivable case—which is not only impossible, but if possible would just be the enactment of a weightless rule, albeit an infinitely lengthy and complex one.

If I am correct, then direct enactment of legal principles is not an option. If lawmakers believe that they are enacting legal principles, they are mistaken. If, for example, there is no free speech principle in the moral domain available for incorporation in the legal domain, then enacting a free speech principle is an impossibility, and lawmakers who believe that is what they are doing must be doing something else.Footnote ²⁸

If legal principles cannot be created directly by enactment, perhaps they can be created indirectly. Indeed, indirect creation is precisely the account given by Ronald Dworkin, whose description of legal principles I am employing. For Dworkin, legal principles are not enacted as such. Rather, they arise out of those legal rules and judicial decisions that are directly enacted.Footnote ²⁹

Legal principles—again, legal norms that lack canonical form and have the dimension of weight—are those principles and their weights that “fit” (would justify) a sufficient number of legal rules and decisions and that have a sufficient degree of moral acceptability. Put differently, legal principles are those principles that are the most morally acceptable of the principles that are at or above the requisite threshold of fit.

On Dworkin's account, legal principles may turn out to be less than morally ideal. That is, legal principles will not be moral principles. For legal principles, unlike moral principles, are constrained by the requirement that they fit the legal rules and decisions, at least to a certain degree. That is why Dworkinian legal principles are not just moral principles consulted by judges. (Dworkin's argument for legal principles, which relies on the notion of fit, both synchronic and diachronic,Footnote ³⁰ actually implies that legislatures and constitutional authors, no less than judges, should be bound by legal principles.)

Over a decade ago, Ken Kress and I wrote an article attacking Dworkin's account of legal principles.Footnote ³¹ The article was long and complex, and I shall give only a very abbreviated version of it here. The nub of the argument, however, was that (1) the moral acceptability axis would dictate whatever threshold of “fit” correct moral principles will satisfy and thus make legal principles identical to moral ones, and (2) Dworkinian legal principles, if not identical to moral principles, would be quite unattractive norms by which to be governed.

With respect to the first point, suppose a jurisdiction has a number of legal rules and judicial decisions on the books that are morally infelicitous or even iniquitous. Moral principles would tell us to follow only those legal rules and decisions that were morally sound or to follow unsound legal rules and decisions only when doing so is warranted according to correct moral principles—in other words, follow moral principles.

It might be objected that if we ignored legal rules and decisions that were morally infelicitous, various bad things would happen. People who relied on infelicitous rules and decisions would have their expectations, on which they may have relied in costly ways, dashed. Coordination with others would become more difficult and costly. And so on.

But notice that if those costs are morally cognizable, which is plausible, then application of correct moral principles will have taken those costs into account. Put differently, if a morally incorrect legal rule or decision is enacted, its enactment changes the facts in the world to which correct moral principles apply. So it may be morally correct to follow a legal rule that it would have been morally better not to have enacted ab initio.

Therefore, the moral acceptability axis will always dictate a threshold of fit that is precisely what following correct moral principles would produce. That means that unless the threshold of fit is determined independently of moral acceptability, legal principles will turn out to be identical to moral principles.

The alternative of the threshold's being independent of moral acceptability is quite unattractive, however. Any threshold of fit less than one hundred percent looks arbitrary. More importantly, however, if the threshold is independent of moral acceptability, legal principles will be normatively unattractive. For on this accounting, legal principles will lack the determinacy virtue of legal rules and decisions—they will have all of the indeterminacy of moral principles (because they can be ascertained only by recourse to morality and will therefore be as controversial as morality)—and they will lack, as well, the moral correctness virtue of moral principles (because they must fit morally incorrect legal rules and decisions). Legal principles will be neither determinate and predictable nor morally correct. They will have nothing to recommend them as norms, and thus there will never be any reason to consult them.

If a norm is not a norm of morality, and if it is not determinate and cannot coordinate behavior, then it has no normative virtues. And if a norm lacks normative virtues, then I would argue it does not exist as a norm. Indirectly enacted Dworkinian legal principles do not satisfy this existence condition.

Robert Alexy's conception of legal principles is very similar to that of Dworkin.Footnote ³² On Alexy's conception, principles have no specific canonical form, and they have the dimension of weight.Footnote ³³ When they conflict, one principle can outweigh the other in the circumstances of the case at hand, but the outweighed principle continues to exist and may outweigh the other principle in different circumstances.Footnote ³⁴ (When rules conflict, however, one rule is either invalid altogether or at a minimum invalid in the circumstances of conflict and thus modified.)Footnote ³⁵ Alexy conceives of legal principles as values to be optimized—realized to the greatest extent possible consistent with their weight vis-à-vis the weight of competing values, including the value of following democratically enacted rules.Footnote ³⁶ Alexy departs from Dworkin—who distinguishes principles from collective goals (policies)—by treating both as principles, the realization of which is to be optimized.Footnote ³⁷

How do principles become part of the law on Alexy's account? Alexy nowhere suggests that, as Dworkin contends, legal principles emerge from the mating of legal rules and legal decisions with moral principles—that is, that legal principles are the most morally acceptable principles that “fit” (to an unspecified degree) with the extant legal rules and decisions.

That leaves on the table the possibility that legal principles are posited by lawmaking acts intended to bring those principles into the law. Again, there are two ways principles might be deliberately made part of the law. First, lawmakers might wish to incorporate real moral principles into the law, either in legal standards, or as necessary or sufficient conditions for legal validity. Second, lawmakers might wish to create principles and enact their own creations. To repeat what I said earlier, real moral principles incorporated into standards are hemmed in by rules and subordinated to those rules and to judicial decisions regarding the content of the moral principles. They are thoroughly domesticated and incapable of undermining the settlement function of law—although they do nothing to further it. However, law's settlement function is undermined when real moral principles are incorporated as necessary or sufficient conditions of legal validity but are not subordinated to posited rules and judicial decisions.

In any event, Alexy provides ample evidence that incorporation of real moral principles is not what he is envisioning in the positing of legal principles. For example, he mentions the possibility of positing morally obnoxious principles, such as the “principle” of racial segregation.Footnote ³⁸ Obviously, that is not a real moral principle, a principle that true morality contains. Nor is the principle, cited by Alexy, of the “maintenance and support of the manual arts”Footnote ³⁹ likely to be a real moral principle rather than, at most, a relevant consideration under some general moral theory. In still other passages, Alexy speaks of enactment of a principle.Footnote ⁴⁰ I conclude, therefore, that Alexy views legal principles and their weights as capable of being created by deliberate enactment (as opposed to being incorporated from an independent moral realm).

If I have read Alexy correctly, then his position is subject to my earlier critique of the direct enactment of legal principles. Weight cannot be enacted. There may be moral principles that have that dimension; that would depend on whether the best moral theory is one in which there is a plurality of principles that are not lexically ordered. And a complex rule with a number of qualifications and conditions on its application may mimic the dimension of weight. But real weight—and hence legal principles—cannot be enacted. Both direct and indirect enactment of principles with weights are impossibilities. Therefore, there are no legal principles. If that is true, then when courts purport to decide cases by alluding to legal principles and their relative weights, they are making it up. There is no independent realm of legal principles and their weights to which they can be referring.

Well, one might ask, what about legal standards—legal norms that are not determinate rules? Do standards not require judges to fill in their requirements? I would answer that standards do require judges to fill in their requirements, but to fill them in by consulting reasons that preexist the legal system, most notably, moral reasons. A standard essentially instructs the judge to do what is morally best within the space left open by legal rules.Footnote ⁴¹ Standards do not create the reasons on which judges are to rely in fleshing them out.

IV. Constitutional Interpretation

I have argued that constitutional provisions creating governmental structures are like assembly instructions for toys and gadgets. They are assembly instructions for creating a government. Constitutional rights provisions, I have argued, can be of two types: determinate rules creating the rights; or incorporation by reference of real moral rights. Finally, I have argued that creating rights other than through determinate rules is an impossibility.

So what do these points suggest is the proper way to interpret constitutions? Let us consider first the interpretation of structural provisions. Consider what the constitutional enterprise is in this regard. We have asked the constitutional authors to tell us how we should construct our government. How should the legislative branch be structured and chosen? How should the other branches be structured and chosen? What should be the respective powers of the federal and provincial governments? And so on. The constitutional authors have communicated their resolutions of these issues to us through symbols—marks on pages, in the case of written constitutions. It would make sense, would it not, to interpret those symbols as meaning what the constitutional authors intended them to mean—that is, by reference to the constitutional authors' intended meaning. After all, we charged them with the task of producing governmental assembly instructions, and these symbols were meant by them to convey what they had produced. It would be decidedly odd to have one group write assembly instructions but then to take the symbols the group used and “interpret” those symbols as if they had been produced by some other group with different intended meanings. Why should we read the Canadian Charter or the United States Constitution as if they were written in English if we do not care about the intended meanings of their actual authors? Why not imagine that the marks on the pages are not English but are in some special code that means what we would like it to mean? No mark or sound has any inherent meaning. Marks, sounds, and other symbols used by someone or some group to communicate their ideas to others have only the meanings that those who produce them intend them to convey. And although for any set of symbols produced by one author with a given intended meaning, we can always treat those symbols as if they were produced by a different author with a different intended meaning, doing so with constitutional provisions makes the whole process of authoring them and adopting them bizarre. When I read assembly instructions for a toy or gadget, I try to discern what those who wrote them intended, not what someone else might have intended by those symbols, much less what I would have intended. Looking for the authors' originally intended meaning is the only thing that makes sense when it comes to constitutions' structural provisions.

Someone might object that I have overstated my case. They might argue that interpretation of structural constitutional rules can properly depart from the authors' intended meaning so long as the symbols in the constitution can bear an alternative meaning. But this is a confusion on several levels.

First, to repeat, giving some people the job of coming up with structural rules but then disregarding the meanings they intended their chosen symbols to convey seems quite bizarre and verges on unintelligibility.

Second, it is a mistake to imagine that a given set of symbols can bear only a limited number of meanings. Any symbol can convey an indefinite number of meanings. Even if we artificially decree that a certain set of symbols must be treated as if it were English (and no set of symbols itself declares what standard language, if any, it is in), words and phrases in English can bear an indefinite number of meanings, again depending on the intentions of their users and the understandings of those intentions by the audience. At some point in time, the word “bad” came to convey the meaning “good,” as in, “That's a really bad car you're driving.” At some point in time, “He's a really cool cat” came to refer not only to the tabby by the air conditioner but to a jazz musician in sweltering New Orleans. Symbols, languages, codes, idiolects—they can convey any meaning so long as the intended meaning is understood by its audience.

Moreover, it is only the intended meaning of its authors that gives a constitutional text its unity, so that the Canadian or American constitution reprinted in one book is the same constitution as that in another book, and the same as that in the national archives. What I mean is this: If a constitution contains an ambiguous word or phrase, and we can resolve the ambiguity either by reference to the authors' intended meaning or by reference to other possible meanings (meanings someone other than the authors could have intended to convey), then the different possible interpretations represent different constitutions. For suppose the ambiguous word or phrase in English has no counterpart in, say, Italian. Rather, there is a separate Italian word for each of the possible meanings. In that case, when translators translate the constitution into Italian, if they resolve the ambiguity differently from one another, they will produce Italian versions that are symbolically different one from another. The only thing that makes a given set of symbols the same text as a different set of symbols is that they have the same authorially intended meaning.

Someone might object to my assertion that seeking the authorially intended meaning of structural rules is the only thing that makes sense of the enterprise of asking some group to author such rules. She might claim that I have conflated the Constitution as a set of structural rules drafted and ratified by identifiable people at a specific time in history with “The Constitution” that is our fundamental law. The latter refers not to the specific instructions communicated to us by their drafters/ratifiers but to our practice of constitutional argument. Within that practice, we might invoke the intended meanings of the actual authors. But we might also invoke the plain meaning of the text, historical developments, morality, prudence, and so forth. Each argumentative modality is part of our constitutional practice.Footnote ⁴² Originally intended meanings occupy no preferred position among such modalities.

There is a sense in which this claim is true, and there is a sense in which it seems deeply confused. It is true that one can find constitutional arguments framed in all of these ways, and one can find Supreme Court opinions that exhibit this modal diversity as well.

Nonetheless, each modality corresponds to a different constitution. The originally intended constitution is different from the textualist one, which is in turn different from the historical and moral ones, and so on. If rendered in Italian, each would be symbolically distinct. So different modalities entail different constitutions.

But then, when a litigant invokes one modality in an argument and her opponent invokes another, what exactly are they arguing about? To answer “the Constitution” is wrong on the modalities account, as there are different constitutions, not the Constitution. The litigants appear to be arguing past each other and thus not in fact disagreeing. They are each engaged in “the practice” of constitutional argument and cannot be said to be disagreeing about that. Footnote ⁴³

What, then, is the argument about, on the modalities account? Perhaps it is not an argument over what the Constitution means but rather an argument about which constitution the Supreme Court should endorse as the Constitution. Perhaps this is true—although those endorsing the modalities approach tend not to see it as one in which litigants offer up alternative constitutions from which the Supreme Court is to choose. For one thing, if the Court does, as it must, choose one modality to resolve the case, that choice is not supposed to establish the supremacy of that modality in any other case. All of the losing constitutions are supposed to remain available for use in future cases.

More importantly, on what basis is the Court to choose among the modalities? There is no meta-modality to govern choice among the modalities. Any choice among them will thus be an exercise of judicial will and nothing more.

There is only one “modality” that makes sense of the enterprise of drafting and ratifying a constitution. Seeking any “meaning” other than the meaning that the actual drafters/ratifiers intended to convey through whatever symbolic medium they employed renders the project of constitution-making quite bizarre. We might as well make anagrams out of the letters in a constitution (though in what language?)!

When it comes to a constitution's structural rules, therefore, those rules have a fixed meaning, the meaning their authors intended at the time the rules were promulgated. Any other interpretive approach is tantamount to reauthoring the structural rules. The “dead hand” of the authors' original intended meaning must control, lest the enterprise of designing a constitutional structure be upended.

How do these remarks about interpreting structural rules bear on the interpretation of constitutional rights? If the rights are embodied in determinate rules, then what I have said about interpreting structural rules applies equally to interpreting rules that create rights. If those rules are given meanings as if they had been authored by someone other than their actual authors, someone who did not mean by them what their actual authors meant by them, then the constitutional rights will have been amended by the courts. If the rules meant on day one what their authors meant by them—and, to repeat, it would be an odd enterprise to have some group come up with rights-creating rules but then assign meanings to those rules that are different from the authors' intended meanings—then the rules continue to mean the same thing on day two and day three and today. Any “living tree” or other approach is as inapplicable to rules specifying rights as it is to rules structuring the government.

What about constitutional rights that represent incorporation by reference of real moral rights? Well, with such provisions, interpretation begins and ends with discovering that the authors intended to refer to real moral rights. Once the courts get into the business of spelling out the content of those rights, they are no longer interpreting the constitution but doing moral philosophy. It would be inapt to describe this enterprise as “living tree” interpretation, or any other kind, because it is not interpretation at all. And based on what I said earlier about incorporating real moral rights and making them judicially enforceable, there is room to doubt that the rights mentioned in the U.S. Constitution or the Canadian Charter were invitations to the courts to consult moral reality and apply their findings as fundamental law.

As I have already noted, the final possibility, the creation of rights other than through determinate rules—that is, through the promulgation of legal principles—is an impossibility.

Therefore, if a court is interpreting a constitution, it is interpreting rules laid down by those authorized to do so and is seeking their authorially intended meaning. Originalism, in the sense of seeking the authorially intended meaning, is the only option in a constitutional system that does not allow judicial amendment.

This is not to say that the courts should never engage in first-order moral reasoning. Even if we put aside the incorporation of real moral rights, even the most determinate set of constitutional rules will undoubtedly leave some matters open; that is, there will inevitably be a realm consisting of standards rather than determinate rules. Moreover, the courts will have to implement even the determinate constitutional rules through development of their own doctrines, and in doing so they will have no choice but to consult first-order moral and prudential reasons. Nevertheless, their doing so will be hemmed in by the rules laid down by the constitutional authors.

Moreover, I am not denying that discerning the originally intended meaning of a constitutional rule will often be difficult. It will often be difficult for the authors themselves, when faced with a rule's application that was not in mind at the time of the rule's promulgation, to distinguish whether they intended the rule to apply but now regret that they did, or did not intend the rule to apply, or possibly had no intention at all with respect to the matter. In other words, it will be difficult, even aided by ordinary norms of conversational implicature, for the authors themselves, as well as their interpreters, to determine what meaning they intended in some range of cases. It will likewise frequently be difficult for the authors themselves, as well as their interpreters, to distinguish what is implied in their rules from what is implied by their rules but not in them.

Despite these difficulties in determining authors' intended meanings, determining those meanings is what interpreting legal directives issued by legal authorities is. Anything else is not interpretation and substitutes some fictitious author for the authors who have the authority to make law, whether constitutional or subconstitutional.Footnote ⁴⁴

V. Judicial Review

What do these observations about law, constitutions, structures, rights, and interpretation suggest regarding the practice of judicial review, particularly the strong and much mooted form of that practice that makes constitutional decisions by the highest court authoritative for all other officials?Footnote ⁴⁵ Can judicial review be justified, and if so, on what assumptions?

First, to the extent that the constitution consists of rules, there is surely a case for making courts' interpretations of those rules supremely authoritative within the legal system. A democratic legislature is, in almost all conceivable circumstances, inferior to a court when it comes to interpreting legal rules, including constitutional rules. The question is not whether the will of a current democratic majority should be subordinated to the will of a nondemocratic body. That question has already been answered in the affirmative by the adoption of a constitution. No matter how democratic the composition of its authors and ratifiers, their rules are not those of current democratic majorities. (And even states like Great Britain without formal constitutions have unwritten “constitutional” rules defining Parliament, its operation, and its powers that even Parliament cannot override.) The only question is which institution is better able to interpret the constitutional rules correctly; and in the case of a written, consciously created constitution, courts seem to be better constructed to get the meaning of rules right. This is true whether the rules are rules about structure or rules creating and elaborating rights. For the meaning of the rules is a matter of historical fact: what their authors' intended meaning was. And ascertaining historical facts is right in the wheelhouse of courts.

Even if the desideratum for rules is that their meaning not be controversial—after all, their function is to settle controversy and avert its costs—authorially intended meanings will often not be clear to everyone and thus will often leave matters unsettled unless authoritatively interpreted. The hope then is that even if the constitutional authors' intended meaning is not clear enough to avert controversy, the courts rendering of that meaning will be.

If a constitution contains standards, matters are different. Standards tell a decision-maker to do “the right thing” within a decisional area cabined by rules. There is no reason to assume that courts are necessarily better reasoners about morality and policy than legislatures or the executive. However, I see no reason of democratic principle against leaving the authoritative fleshing out of standards to courts. Although Jeremy Waldron disagrees,Footnote ⁴⁶ the question for me is what mix of democratic and nondemocratic institutions is most likely to “do the right thing.” If courts are epistemically and motivationally superior to legislatures in fleshing out standards, then good constitutional design will give that task to them.

In any event, even if a constitution were all rules and no standards, courts would have to consult moral and other practical considerations in designing doctrines to implement those rules. For, as Mitch Berman and others have pointed out,Footnote ⁴⁷ judicial enforcement of a constitution's rules will require the creation of doctrines—rules—regarding burdens of proof and other similar matters. These doctrines could themselves be contained in constitutional rules, but often they are not.Footnote ⁴⁸

Thus, courts are likely to need to resort to moral and prudential practical reasoning in fleshing out (rulifying) constitutional standards and in constructing doctrines for implementing both constitutional rules and constitutional standards. If courts are not superior to democratic legislatures in such first-order practical reasoning, then perhaps the judicially created doctrines for fleshing out constitutional standards and for implementing constitutional rules and standards should not be finally authoritative within the legal system, but should be subject to legislative overrides. Again, the principal reason would be one of relative competence, not Waldron's reason that there is a moral right to have controversial moral issues resolved by a democratic vote.

Finally, what if a constitution refers to preexisting moral rights, notwithstanding the perils of incorporating morality to which I have alluded? As I have said, if that is what a constitution does, then this would imply that such moral rights are to be enforced by the judiciary. After all, the legislature and the executive are always supposed to make their actions consistent with moral requirements, whether or not those requirements are also found in the constitution. So explicitly referring to preexisting moral requirements in a constitution signals not merely that such requirements bind the government but also that they are to be enforced by the courts. Whether or not it is wise as a matter of constitutional design to make courts the moral censors of government action is a question about which reasonable minds can disagree. (A body with the responsibility of hearing individual moral complaints about proposed and enacted legislation and advising the legislature on the legislation's moral propriety might be a better design than judicial enforcement of the judicial view of that propriety.) Moreover, judicial moral oversight can vary from de novo consideration to highly deferential review. Once more, the issue for me is not one of democratic right but rather one of wise political design.

VI. Constitutional Change, Organic or Otherwise

Constitutions usually provide for their own amendment. When they are amended, we can describe that as an organic change in the constitution, a change authorized by the constitution itself.

Constitutions change in ways other than organic ones. One of the most obvious of these is through decisions by the highest legal authority—in the United States, the U.S. Supreme Court—that misinterpret some constitutional provisions. Because law's function is to settle what we are obligated to do—or so I have claimed—when there is disagreement over the law's meaning, it is important that the law provide a means of settling that dispute. So any constitution whose provisions' intended meanings are at all disputable should designate some institution as having the authority to settle disputes over those meanings.Footnote ⁴⁹

When the highest legal authority misinterprets the intended meaning of a constitutional provision, we now have a conflict between the constitution itself and the authoritative interpretation of it. Which is the “supreme law”? Although some disagree,Footnote ⁵⁰ the better view is that the authoritative interpreter's interpretation settles what we are obligated to do. What is more vexing, however, is what the authoritative interpreter should do when it comes to believe its interpretation of the constitution was mistaken. This is the issue in U.S. constitutional law of whether the Supreme Court should always, never, or sometimes follow its own constitutional precedents that it now regards as mistaken. My own opinion is that “sometimes” is the correct answer, and that the Court should overrule its own constitutional precedents only if (1) it believes them to be erroneous and (2) it believes that overruling them will not create a great deal more mischief than leaving the incorrect precedent intact. (Overruling a mistaken decision upholding the constitutionality of paper money would produce disastrous consequences;Footnote ⁵¹ overruling Roe v. Wade Footnote ⁵² would not, the plurality in Casey notwithstanding.)Footnote ⁵³

It is also possible that the highest legal authority's mistakes regarding the constitution's intended meaning—or the authority's deliberate refusal to seek the intended meaning and its substitution of a “meaning” it prefers for the intended meaning—will meet with public approval and will be accepted by the public as fundamental law, much in the way the public accepts the constitution qua the originally intended meaning. The public's acceptance as fundamental law of those deviations from the originally intended meaning will result in the constitution's having been changed nonorganically.

In other words, it is always possible for there to be some sort of bloodless constitutional revolution. Constitutions are fundamental law only if they are accepted by the people as fundamental law, and the people may wake up tomorrow and begin accepting as fundamental law some new instrument. The U.S. Constitution was not an organic continuation of the Articles of Confederation. (The constitutional convention that produced it was itself created under the authority of the Articles of Confederation, but the convention was not authorized to produce a new constitution, only to amend the Articles.) The Constitution was just run up a flagpole, and the people saluted. If they had not saluted, the Constitution would have no more authority today than do the original Articles of Confederation or the Constitution of the Confederate States of America.

Therefore, if justices depart from the intended meanings of the constitutional authors, and the people accept these new judicial “amendments” as fundamental law, then we will have had several constitutional revolutions.Footnote ⁵⁴ Several new constitutions, superficially resembling but actually different from one another, will have come into being through successive judicial “amendments” and popular acceptance of those “amendments.” But the real question then is whether the people are actually aware of what is going on. Is their acceptance itself dependent on their belief that the courts are not amending the constitution from the bench but are interpreting it?Footnote ⁵⁵ If so, then a constitutional crisis perhaps awaits.