William Baude and Stephen Sachs argue for the importance of the “law of interpretation” — legal standards that govern how statutes, constitutional provisions, and other legal materials are to be interpreted.1 Their article begins by following and developing a cluster of arguments I have elaborated in recent work — arguments that emphasize the importance of distinguishing between, on the one hand, the linguistic meaning of legal texts and, on the other, the content of the law.2 But Baude and Sachs’s view about the most important implication of these arguments is very different from mine. Their central message is a practical, lawyerly one: we can avoid the abstract and theoretical complexities and normative and linguistic disputes that have typified central debates over legal interpretation by instead looking to law for the answers. Baude and Sachs’s goal here is not to take a position on what the law of interpretation requires with respect to the relevant issues, but to argue that the answers are there to be found in the law.
It is an important and understudied idea that legal interpretation can, to an extent, be governed by ordinary legal standards. And Baude and Sachs develop the idea in fruitful and interesting ways. The dependence of legal interpretation on legal standards is relatively limited and superficial, however. The more general point is that what makes a method of legal interpretation correct is that it accurately identifies the law. Consequently, as I will argue, answers to questions about legal interpretation depend on how the content of the law is determined at a more fundamental level than legal standards.
Once we carefully distinguish between the linguistic meaning of the legal texts and the content of the law, it becomes clear that the main goal of legal interpreters is to find the latter. Facts about the content of the law — legal facts, for short — are not among the most fundamental facts of the universe. They hold in virtue of more basic facts, such as facts about what various people have said and done and decided in the past and about what various texts mean, and perhaps moral or normative facts as well. Because legal interpretation seeks the content of the law, which methods of legal interpretation are correct depends ultimately on the way in which the legal facts are determined by the more basic facts. There can be legal standards that specify how the more basic facts determine the content of the law, but such legal standards play a subsidiary role in the full explanation of how the content of the law is determined. For the legal standards that make up the law of interpretation, like other legal standards, themselves depend on how the content of the law is determined at the fundamental level — the province of jurisprudential theories such as those of H.L.A. Hart and Ronald Dworkin. As a result, one cannot resolve key questions about Baude and Sachs’s law of interpretation — to what extent and why law can alter the way in which the content of the law is determined, what that “law of interpretation” requires, how to find it, and how much of it there actually is — without addressing the central jurisprudential question of how the content of the law is determined. Attempting to explain how legal standards are determined by pointing to further, second-order legal standards only pushes back a step the core question of how the more basic facts determine legal standards.
Part I begins with an important theme on which Baude and Sachs and I agree — that it is a mistake to think that legal interpretive debates are centrally about linguistic matters. I argue, however, that Baude and Sachs insufficiently motivate their central idea — that these interpretive debates are resolved by second-order legal standards. Part II begins the development of a more theoretically complete account, showing that the correct legal interpretive methodology depends on the way in which the content of the law is determined by more basic facts. Part III explains the distinction between the way in which the content of the law is determined at the fundamental level and the way in which it is determined at less fundamental levels. This distinction allows us to see both why legal standards are relevant to questions of legal interpretation and why they are only a limited part of the story. Part IV shows that, contrary to what Baude and Sachs claim, major competing theories of law have very different implications for legal interpretation. Consequently, Baude and Sachs cannot simply set aside the issue of how the content of the law is determined at the fundamental level. Part V argues that Baude and Sachs do not succeed in justifying their confidence that debates over legal interpretation can be resolved by appealing to the law of interpretation — and without entering into the theoretical complexities that typify legal interpretive debates. Finally, Part VI briefly concludes.
I. Legal Interpretation, Linguistic Meaning, and Legal Standards
I begin with the part of the argument on which Baude and Sachs and I are in agreement. First, the term “legal interpretation” is often used in a way that is ambiguous between ascertaining the meaning of legal texts and using the relevant texts to ascertain what the law is. Theorists often slip back and forth between these two usages. For example, it is ubiquitous in discussions of constitutional interpretation to talk about “the meaning” of the Constitution or of a constitutional provision. It is often left unclear whether the question is the meaning of the words of the provision or the provision’s contribution to the content of the law.3
This slippage is likely both encouraged by and supportive of a widespread implicit assumption that I have labeled the Standard Picture.4 According to this picture, a provision’s contribution to the law is its ordinary linguistic meaning.5 If the Standard Picture were accurate, there would be little difference between figuring out the meaning of legal texts and ascertaining the law. Conversely, using the term “legal interpretation” in a way that slips back and forth between the two activities is likely to blur the difference between the meaning of the texts and the content of the law. I have argued that the Standard Picture is a controversial substantive position, one that is very much in need of argument. And I have offered a variety of arguments that the Standard Picture is not correct.6
Up to this point, Baude and Sachs follow and build on my arguments. They recognize the slippage in the use of “legal interpretation” and the close relation between that slippage and the widespread assumption of the Standard Picture. And they potently develop two of my arguments that that picture is problematic. First, the Standard Picture does not seem to square with the way in which our legal system works.7 Second, and more importantly, the contemporary study of language reveals that there are multiple kinds of meaning.8 Baude and Sachs use “author’s intent” and “reader’s understanding” as their examples, but many fine-grained distinctions can be made in this area.9 Given the multiplicity of types of linguistic content, the idea that the content of the law is determined by the linguistic content of the authoritative legal texts is badly underspecified. Moreover, once it is recognized that we have to bring in nonlinguistic considerations to adjudicate between different types of linguistic contents, there is no ground for excluding candidates for a provision’s contribution to the content of the law that are not linguistic contents at all.10
Baude and Sachs next point out that some have reacted to these problems with the Standard Picture by claiming that there is no right way to engage in legal interpretation. Rather, interpreters must make all-things-considered normative judgments on a case-by-case basis.11 Baude and Sachs argue that this reaction is a mistake. There is an additional resource that the skeptics have neglected — there are legal standards that govern legal interpretation. It doesn’t matter whether there are multiple kinds of meaning; second-order legal standards tell us how legal texts contribute to the content of the law.12 Baude and Sachs go on to develop this central idea in insightful and interesting ways. To take just one example, they convincingly argue against the common assumption that standards governing legal interpretation must outrank the sources of law to which they apply.13
Baude and Sachs are right that legal standards are an additional resource that can help to answer the skeptic. But to point out that there may be second-order legal standards specifying how the content of the law is determined is not a theoretically satisfying response to the skeptic. Because the skeptic thinks that controversy about how sources contribute to the content of the law leads to indeterminacy about what the law requires, appealing to second-order standards simply pushes the problem back a step. For concerns about the determinacy of legal standards will apply at least as strongly to second-order legal standards as to first-order legal standards.
More generally, appealing to legal standards governing how the content of the law is determined immediately raises more basic questions: Are such legal standards the only additional resource? What makes it the case that these legal standards exist? How does one figure out what they require? To what extent do these legal standards reach the issues that are the subject of the ongoing debates in legal interpretation? Are there limits on the extent to which legal interpretation can be controlled by legal standards?
II. The Dependence of Legal Interpretation on How the Content of the Law is Determined
To see the way to a more theoretically satisfying account, let us return to the point that Baude and Sachs recognize clearly and emphasize throughout — that ascertaining the linguistic meaning of a legal provision is a different activity from ascertaining the contribution of that provision to the content of the law. The content of the law consists of obligations, rights, powers, and the like. By contrast, linguistic meaning is information represented by symbols. These two things are not even of the same general category. And, as a general matter, the content of norms need not be determined by information represented by symbols. After all, there are systems of norms in which texts or other symbols have no constitutive role. Morality is one example. And whatever is the truth about the actual norms of etiquette, a society could have norms of etiquette that obtain in virtue of aspects of practices other than the symbolic representation of information.14
Once we make the crucial distinction between linguistic meaning and the content of the law, it is clear that the primary goal of judges, lawyers, and other legal interpreters is to figure out what the law is, not what the meanings of the legal texts are.15 Judges must ascertain the law in order to follow it. Working out the meanings of legal texts is an important means to the end of ascertaining the law, but just a means. And other activities that courts engage in — such as deciding how to resolve a case when there is no decisive legal standard or, in extraordinary circumstances, deciding whether to depart from what the law requires — cannot be reached until the court has ascertained the law. Like Baude and Sachs, I will use the term “legal interpretation” for the activity of ascertaining what the law is, not the activity of working out what the texts mean, though nothing turns on this choice of terminology.16
Because legal interpretation seeks to ascertain the content of the law, a method of legal interpretation is correct if it accurately identifies the legal facts. Given this point, it is but a short step to recognize that the correct method of legal interpretation depends on how the content of the law is determined. As noted above, legal facts are high-level facts, which obtain in virtue of more basic facts. In general, in such high-level domains, the correct method of ascertaining the high-level facts will depend on how the more fundamental facts make it the case that the high-level facts obtain.17
To take a simple, concrete example, suppose that we want to know the U.S. trade deficit for 2016. Trade deficit facts are not basic facts about the universe. So we need to know the more basic facts that determine the U.S. trade deficit. The trade deficit is the excess of imports over exports. Thus, in order to figure out what the trade deficit is for 2016, we need to figure out the total of U.S. imports and the total of U.S. exports for 2016 and then subtract the latter from the former. Or, to take a different kind of example, suppose we want to know whether a particular position on a chessboard is checkmate. As checkmate facts are not basic facts, we need to know the more basic facts that make the case that a position is checkmate. Checkmate is a position in which the king is in check and there is no move that will take the king out of check. Thus, in order to figure out whether the position is checkmate, we need to figure out whether the king is in check and, if so, whether there is any move that will take the king out of check. Similarly, in the case of law, whether a method of interpretation is correct — whether it will accurately deliver legal standards — depends on what the more basic facts are that determine the legal facts and how those more basic facts combine to yield the legal facts. To put it schematically, if the content of the law is determined by a certain function f of the x, y, and z facts, then ascertaining the content of the law will require ascertaining the x, y, and z facts and working out f (x, y, z).18
Suppose, for the sake of illustration, that a statute’s contribution to the law is determined by the legal intention19 of the legislature unless that intention is unclear — in which case the statute’s contribution is determined by the semantic content of its text.20 In that case, ascertaining a statute’s contribution will require ascertaining whether the legal intention of the legislature is clear, and, if so, what that intention was, and if not, what the semantic content of the statutory text is. The point is that, just as we cannot figure out the trade deficit or whether a position is checkmate without understanding how more basic facts constitute the trade deficit or checkmate, we cannot figure out the content of the law without understanding how more basic facts make the legal facts obtain.21
Baude and Sachs’s central thesis is that there are legal standards specifying how the content of the law is determined and that these standards resolve many of the central debates over legal interpretation. In the next section, I explain why legal standards are only part of the story about how the content of the law is determined.
III. Different Levels at Which the Content of the Law Is Determined: The Law of Interpretation Distinguished from the Fundamental Level
We need to make an important distinction between the way in which the content of the law is determined at the most fundamental level and the way it is determined at less fundamental levels. To say that the content of the law is determined in a particular way at the most fundamental level is to say that the content of the law is determined in that way and that it is not the case that it is determined in that way because of some further determinant. Jurisprudential theories like those of Hart and Dworkin offer accounts of how the content of the law is determined at the fundamental level.22 (I will use the term theories of law as a technical term for such accounts.)
We can use Hart’s theory to illustrate the distinction between the fundamental level and higher levels. On Hart’s theory, the content of the law is determined at the fundamental level by convergent practices of judges and other officials.23 It is not the case that the content of the law is determined by those convergent practices because of, say, a convention to that effect, a text that so specifies, or a truth about democracy. Rather, it is simply in the nature of law that the content of the law is, at the fundamental level, determined by the relevant practices. On Hart’s theory, then, roughly speaking, sources contribute to the content of the law in whatever way all or most judges treat those sources as doing so. So if judges treat statutes as contributing to the law according to the communicative intentions of the legislature, then that practice makes it true that that is how statutes contribute to the content of the law, and, absent some overriding factor, the corresponding version of intentionalism is the correct method of legal interpretation.
Legal standards are found at higher levels. If Hart’s theory is correct, the fact that the content of the law is determined by convergent practices of judges is not something that is required by a legal standard.24 Rather, this fact is more basic than, and part of the explanation of, the legal standards. Given the basic fact, statutes, judicial decisions, customs, and so on can yield, at higher levels, legal standards that further specify the way in which various factors contribute to the content of the law.
In sum, there are two parts to the story about how the content of the law is determined. First, there is how the content of the law is determined at the fundamental level. Second, because of the way in which the content of the law is determined at the fundamental level — plus the actual statutes, judicial decisions, and so on of a particular legal system — there can be, at higher levels, legal standards that specify how the content of the law is determined.
We can now see why second-order legal standards are indeed a good place to look for answers to questions about legal interpretation. They are one aspect of how the content of the law is determined. But we can also see that legal standards are only part of the story — and the less basic part.
IV. Baude and Sachs’s Attempt to Set Aside Theories of Law
Baude and Sachs focus on legal standards throughout. Although they note in passing that the existence and content of the legal standards depend on the way in which the content of the law is determined at the fundamental level, for most of the paper they simply ignore the issue. Their most explicit attempt to address it runs as follows:
Whether our system is textualist, intentionalist, purposivist, or something else is a legal question, to be answered by our sources of law — and, in the end, by the appropriate theory of jurisprudence. (We assume in this Article something like Hartian positivism, partly for ease of exposition, though much of our framework should hold true on any mainstream theory.)25
Here, Baude and Sachs are acknowledging that, ultimately, which method of interpretation is correct depends on which theory of law is true — that is, on how the content of the law is determined at the fundamental level. But their parenthetical suggestion is that they can afford to set aside the issue because their position (or at least much of it) would be true on any “mainstream theory” of how the content of the law is determined at the fundamental level.
This is a surprising claim. Let’s begin with Hartian positivism, which Baude and Sachs say that they assume in their article and which is the most widely held theory of law, at least in law schools. Hart’s theory is an unfortunate choice for theorists who claim that longstanding debates over legal interpretation can be resolved by looking to the law of interpretation, for Hart’s theory implies that theories of legal interpretation can be true only if they are already widely accepted by judges and other officials or validated by a criterion that is itself grounded in such a consensus. Either way, with certain exceptions that are not helpful to Baude and Sachs’s project, Hartian positivism makes little room for a law of interpretation that goes beyond what is already widely accepted.26
According to Hartian positivism, in order for it to be the case that a particular theory of legal interpretation — textualism, say — is true, it must be either incorporated in the rule of recognition or validated by the rule of recognition. Let’s consider these two possibilities in turn. In order for textualism to be part of the rule of recognition, there must be a convergent practice among at least a large majority of judges of treating statutes as contributing to the law in the way that textualism requires. (For simplicity, I restrict attention to judges.) Specifically, if judges: (1) regularly treat statutes as contributing to the law in the way that textualism specifies; (2) are disposed to criticize other judges who fail to do so (or threaten to fail to do so); and (3) regard such criticisms as justified, then textualism is, for that reason, the correct theory of legal interpretation.27
The second possibility is that textualism is validated by a criterion of validity that is treated as correct by at least a large majority of judges.28 For example, textualism could be correct because a constitutional provision or a customary practice, when interpreted as specified by the rule of recognition, requires textualism (and that requirement is not overridden by some conflicting requirement validated by the rule of recognition). As these are the only two possibilities that Hart’s account allows, to the extent that there is no consensus on a theory of legal interpretation and no consensus on a criterion that validates a particular theory of interpretation, it is indeterminate which theory of interpretation is correct.
The first possibility is hopeless for Baude and Sachs’s purposes. Because of the straightforward way in which the rule of recognition is determined by the convergent practice of judges, nothing that is uncertain or controversial can be part of the rule of recognition. Thus, to the extent that judges lack consensus with respect to whether a theory of legal interpretation is correct, that theory is not part of the rule of recognition. In other words, no theory of legal interpretation that is not widely accepted can be part of the rule of recognition. For example, an intentionalist or textualist position on constitutional interpretation cannot be part of the rule of recognition, for there is anything but a consensus among judges that intentionalism or textualism is correct.
A theorist might argue that at a high level of generality judges agree on the theorist’s preferred method of interpretation, and that the disagreement is in the application of that consensus. For example, it might be argued that there is a consensus that the communicative content of the legislative texts is determinative, but that many judges have mistaken views about what constitutes the communicative content. Or that there is a consensus that a statute’s contribution to the content of the law is whatever the best understanding of democracy says that it should be, but that there is a disagreement about what democracy requires. Or, differently again, that there is a consensus that the Constitution should be interpreted solely in accordance with what a reasonable member of the audience at the time of ratification would have taken the Constitution’s semantic content to be, but that judges disagree about what such a person’s understanding would have been.
It is not at all plausible, however, that there is such a consensus. Judges do not all agree that a specific type of meaning such as communicative content or semantic content is controlling. They don’t even agree that the linguistic content of the relevant texts is controlling. Nor do they agree that democratic considerations are the relevant ones. The only kind of consensus is on bland platitudes such as that original meaning matters or legislative intention is important. Such platitudes are too underspecified to yield a uniquely correct application. Which kind of original meaning? Semantic content? Communicative content? “Public meaning”? Exactly how does original meaning matter, and what other factors matter and in what way? Which kind of legislative intention? Communicative intentions, legal intentions, application intentions, or some kind of hypothetical or constructed intention?29 At the level of the rule of recognition, the Hartian account has no criterion for correctness other than the convergent practice of judges. Consequently, the account has no resources with which to argue that, say, giving a specific role to a particular kind of legislative intention is the correct way of understanding the importance of legislative intention.
Moreover, even if, implausibly, some less bland formulation could be found to which all or most judges would give lip service, that is not the kind of agreement that the Hartian account requires. As we have seen, the rule of recognition is supposed to be constituted by the actual practices of judges — by their regularly treating sources of law as contributing to the law in a particular way, being disposed to criticize other judges for not so treating those sources, and so on. Being disposed to accede to a form of words — to give lip service — does not even amount to genuine belief, let alone a convergence of practice. There is neither a consensus of belief nor of practice on any of the issues on which theories of legal interpretation disagree.
The second possibility — that the relevant legal standards are validated by a criterion of validity that is itself accepted by consensus — is a little more promising for a law of interpretation that goes beyond consensus. How could there be a consensus criterion that points to a legal answer, yet the participants in the consensus have failed to notice that their criterion yields that legal answer? There could be a consensus on a criterion the application of which is controversial. The most promising type of candidate would be a normative one — what democracy requires, for example. But there is no consensus on a criterion of this sort that would plausibly hold out the promise of yielding answers to controversial questions of legal interpretation; certainly Baude and Sachs do not offer such a consensus criterion. Moreover, Baude and Sachs make clear in various places that they are not sympathetic to the idea that the relevant criteria of validity in the U.S. legal system are normative.30
If we restrict attention to descriptive or factual criteria of validity, it’s harder to see how the relevant legal standards could have remained unrecognized. It would have to be that there was a consensus that pointed to a source of law, but that there was an empirical mistake about what that source of law specified, perhaps because the source had somehow been lost. Baude and Sachs focus their discussion on unwritten rules of interpretation that have their source in customs or practices other than judicial decisions. When we turn to their discussion of such unwritten law below, we will see that they do nothing to point to a consensus among judges on a criterion for recognizing how customs contribute to the law — let alone one that could plausibly pick out a right answer (though one that has largely gone unnoticed) as to whether, for example, textualism is correct. In sum, Baude and Sachs provide no reason to doubt that, if Hartian positivism is true, there is not much law of interpretation in the U.S. legal system that goes beyond what is widely accepted by judges.
The more general point is that, pace Baude and Sachs, an inquiry into legal interpretation cannot afford to set aside the question of how the content of the law is determined at the fundamental level. The reason is that prominent theories of law have extremely different implications for legal interpretation. In sharp contrast with Hartian positivism, Dworkin’s “law as integrity” theory yields a highly determinate account of the correct method of legal interpretation that goes far beyond any consensus. If Dworkin’s theory is true, then — at least if Dworkin is right about the consequences of his theory — the correct method of legal interpretation seeks the set of principles that best fit and justify the practices of the legal system, including the Constitution, statutes, and judicial decisions.31
Differently again, Scott Shapiro argues that his Planning Theory has complex consequences for theories of legal interpretation, including in particular that there are different criteria for the correctness of a method of interpretation depending on why the current officials of the legal system accept, or purport to accept, the rules of the system.32 To illustrate, in an “authority” system — one in which the officials accept the rules because they believe “that these rules were created by those having superior moral authority or judgment” — which method of interpretation is correct depends on the planners’ attitudes about the trustworthiness of interpreters.33
On my own Moral Impact Theory, roughly speaking, legal interpretation requires ascertaining the all-things-considered normative consequences of the ratification of constitutions, enactment of statutes, and other actions of legal institutions.34 The Moral Impact Theory does not imply, however, that every interpretation of a legal text requires direct moral or normative reasoning. It might be that it is all-things-considered required to follow a certain method of legal interpretation that involves no normative reasoning — or, differently, to follow this method unless there is a defeating consideration.
In sum, different accounts of how the content of the law is determined at the fundamental level have importantly different implications for legal interpretation in the United States and similar legal systems, both for how much law of interpretation there is and for what it requires. Moreover, many of the most important implications for legal interpretation are not specified by legal standards, but determined at a more basic level. The implications for legal interpretation of the different theories sketched above are not required by the law of interpretation; rather, they follow from the theories’ claims about the nature of law or about how the content of the law is determined. Consequently, even if the law of interpretation does not resolve a dispute over interpretive methodology, that question may nonetheless be resolved by the way in which the content of the law is determined at the fundamental level. Another reason the distinction between the fundamental level and legal standards may matter is that it is controversial to what extent legal interpretation is subject to control by legal standards. For example, many scholars and judges have concerns about whether federal courts’ decisions about interpretive methodology may be restricted by law.35 By contrast, it is difficult to dispute that courts should interpret in accordance with the way in which the content of the law is determined at the fundamental level.36
V. Do Baude and Sachs Succeed in Showing That the Law of Interpretation Resolves Real Interpretive Controversies in the U.S. Legal System?
In light of this discussion, we might wonder how successful an approach that focuses on the law of interpretation — to the exclusion of the way in which the content of the law is determined at the fundamental level — could be. I now turn to a brief consideration of how Baude and Sachs’s approach fares: to what extent are they able to show that ongoing debates over legal interpretation can be resolved in a lawyerly way — and without entering into the theoretical complexities typical of those debates?
They mostly focus on unwritten law of interpretation. One reason is presumably that, though there are some statute-based interpretive rules, they are largely well known and uncontroversial and do not resolve the deep, ongoing debates. Also, Baude and Sachs suggest that unwritten law of interpretation has been to some extent overlooked because theorists have wrongly assumed that it has to outrank the sources of law whose interpretation it governs. In the case of constitutional interpretation, moreover, Baude and Sachs suggest that unwritten law may be pretty much all we have, at least in part because statutory law cannot affect the way the Constitution contributes to the content of the law.37
They claim that unwritten law includes, in addition to canons of interpretation, law on fundamental questions: which practices or materials are the relevant ones and how those materials contribute to the content of the law. For example, they suggest that the question of whether various aspects of legislative history, such as committee summaries of bills, are among the determinants of law “is a question of law.”38 Along the same lines, in response to Judge Katzmann’s suggestion that Congress incorporates into statutes the materials that it deems useful to their interpretation, Baude and Sachs insist that “[w]hat statutes ‘incorporate’ — and who has power to ‘deem’ things ‘useful’ — is a legal question, and it has to be answered by reference to our law of interpretation.”39 Even more centrally, the question of whether textualism, intentionalism, or some other method is correct is a legal one: “[a]rguments about the approaches used in our legal system should be conducted as legal arguments, based on legal materials and not (or not primarily) on pure interpretive theory.”40 Indeed, that we can and should resolve such questions by looking to law is perhaps the central theme of the article.
But why are Baude and Sachs so confident that there is law on the relevant questions?41 Why think that there is an existing legal standard that specifies whether committee summaries are a determinant of law or whether textualism is the correct method of legal interpretation? As we have seen, the extent to which there are determinate answers to interpretive questions on controversial issues depends on which theory of law is true — on how the content of the law of interpretation is determined. Hartian positivism has the consequence that there will be relatively little in the way of determinate answers — at least in the circumstances of the U.S. legal system. But other theories, such as Dworkin’s, have very different implications. (Moreover, even when basic issues about legal interpretation have a determinate resolution, that resolution may be not at the level of legal standards but at the more fundamental level of a theory of law.) So it seems rash, without addressing the level of theory of law, to claim that there is law of interpretation that resolves the ongoing disputes over legal interpretation.
In order to justify their confidence that legal standards resolve the disputes, Baude and Sachs could explain their account of what makes it the case that unwritten law includes a particular norm. We would then be able to evaluate whether that story would in fact yield extensive norms on the relevant questions. (Providing such an explanation of how the unwritten norms are determined would also illuminate how we could go about ascertaining the unwritten norms.) Short of such an account, Baude and Sachs could develop convincing examples of ways in which unwritten law resolves controversial debates.
Baude and Sachs don’t take either of these routes. As far as offering examples, they make occasional assertions about what the unwritten law requires, though they don’t do much to back up these assertions. Most notably, they assert in passing that our law rejects pure textualism and pure intentionalism in favor of a method that searches for a kind of idealized artificial intention: “We read a statute as if it had been written by a sole legislator, as if that legislator were aware of the whole code, . . . and so on. The legislative intent we impute might be ‘apparent,’ . . . or ‘constructed,’ but it’s still proper according to law.”42 The only support that they offer for this claim is that this is how lawyers actually proceed in our system. “Lawyers don’t actually cast aside any statutes after learning that some (or even many) legislators disagreed about their meanings. . . . They proceed instead to the artificial intent, meaning, and purpose to which the law points.”43 This quick suggestion does not come close to showing the correctness, on any prominent theory of law, of the kind of idealized intention theory of interpretation that the authors gesture at. (To begin with, it is far from obvious that lawyers generally proceed in the way that Baude and Sachs claim.) The presentation of the example seems to assume, instead, that whatever contemporary lawyers do is, for that reason, correct. Such a quietist position seems poorly suited, however, to Baude and Sachs’s view that existing law resolves disputes on which contemporary lawyers have deep disagreements.
With respect to how the relevant legal norms are determined, Baude and Sachs do offer tentative but suggestive remarks about how they understand unwritten law. Their basic picture seems to be that unwritten rules are found in practices other than judicial decisions, rather than being laid down by courts in accordance with stare decisis. (In any case, it would be perverse to suggest that judicial decisions, in particular decisions of the U.S. Supreme Court, resolve the debates over, for example, which aspects of legislative history matter and which methods of interpretation are correct.44) They contrast Abbe Gluck’s view that the law of interpretation is judge-made law and therefore open to judges “to revise as they see fit,”45 with their own view that judges “might be recognizing elements of an existing general-law tradition . . . that makes its appearance in judicial decisions, but isn’t merely their creature.”46
The first court decision recognizing a given principle wasn’t necessarily making it up. Courts might have a judicial obligation to find common law rules in other sources (including customary sources), not merely to make them to fit one’s will. As those sources evolve by slow accretion, or as understandings of the same materials change over time, eventually some court will be the first to say so.47
This is an intriguing picture, but in order for it to justify Baude and Sachs’s confidence that unwritten law will answer our questions, we would need to know which practices are the relevant ones and, even more importantly, how those practices contribute to the content of the law. From what they say, the practices in question are not judicial decisions (and, as noted, judicial decisions would not be helpful). Customs are the only example of unwritten sources that Baude and Sachs offer, and they do not say anything more specific. Telling lawyers to look to customs, without more, doesn’t take us very far. Which customs matter? Do customs of ordinary people count for anything? Of ordinary lawyers or only certain elite lawyers? Must customs already be regarded as legally binding in order to be relevant (as in the standard account of customary international law)? And exactly how do the customs in question contribute to the content of the law? There are familiar problems about how customs or practices more generally determine norms. For one thing, any set of customs is consistent with indefinitely many conflicting norms. Is the relevant norm the one that most practitioners actually have in mind? (Of course, this option requires that practitioners actually have a shared understanding of the relevant norm.) Is it the one that best justifies the customs? Or is it something else? Baude and Sachs don’t say anything about these issues, simply exhorting us to look to unwritten law.48
More importantly, in order to support their thesis that unwritten law provides the answers, Baude and Sachs would have to do more than just give us additional detail about which customs are the relevant ones and how they contribute to the content of the law. Any such account would be controversial, and they would have to provide an argument that their account is correct. But it is difficult to see how they could provide such an argument without entering into the question of how the content of the law is determined at the fundamental level.
At one point, Baude and Sachs sketch a view of which rules of interpretation apply. They distinguish “adoption rules,” which “determine the legal content of a written instrument upon its adoption,” from “application rules,” which “are framed as instructions to future decisionmakers, including judges, on what to do at the point of application.”49 They give a quick argument for the view that the relevant adoption rules are the ones that were legally required at the time a provision was adopted, whereas the relevant application rules are the ones in force at the time that a court must interpret the provision. The argument does not look to a Hartian rule of recognition — to the way judges treat provisions as contributing to the content of the law. Rather, the gist of the argument is that a legal instrument makes its contribution to the law at the time it is adopted; furthermore, legal rules persist until a legally significant event alters them.
The argument has intuitive appeal, though there is a worry that the hard work is being pushed back to the decision whether a rule is an adoption rule or an application rule.50 Even setting aside this concern, the argument’s conclusion simply tells us that the way in which statutory and constitutional provisions contribute to the content of the law is the way that was correct at the time they were adopted. We are left without an account of how to determine what adoption rules were correct at that time — and, again, such an account depends on the correct theory of law. (Hartian positivism, it is worth noting, seems to be inconsistent with Baude and Sachs’s view of which interpretive rules apply, for there is no consensus among judges or other officials that the relevant adoption rules are the ones legally required at the time a provision is adopted. Even originalist judges do not typically argue for their originalist positions on the ground that those positions were correct at the time the Constitution was adopted.) Moreover, on Baude and Sachs’s view, which rules of interpretation apply does not depend on what contemporary lawyers actually do. Their view thus undermines the basis for their earlier endorsement of idealized intentionalism and removes the possibility that consulting contemporary practice yields an easy way to find the law of interpretation.
Near the end of their article, Baude and Sachs briefly analyze a controversy over which adoption rules were in fact correct at the time the Constitution was adopted. They point out that people back then disagreed over which interpretive rules would apply and emphasize that what matters is not which view had the majority or plurality: “But the way to resolve a Founding-era legal disagreement isn’t merely to total up the votes and see who had more; we want to know who had the better of the argument, based on the higher-order legal rules of the era.”51 They go on to suggest that the conflict was between “the lawyerly class” and “the common folk.”52 Their bottom line:
Whether that elite tradition trumped any contrary popular objections isn’t just a historical question: it’s a legal one. Different theories of jurisprudence look to different facts to identify a society’s law. If the practices of lawyers who participate in the legal system have a heightened claim to determining the law, at least as compared to the policy preferences of the general public, then it’s the elite practice that matters.53
Exactly so. How we resolve a disagreement about which method of interpretation was correct ultimately comes down to how the content of the law is determined at the fundamental level — in this case, to what role elite practice has in determining the content of the law. Thus, the most concrete example Baude and Sachs develop of how unwritten law resolves a controversial issue is not really an example of looking to legal standards to resolve which method of interpretation is correct. As they essentially concede, it is an example in which we must look to jurisprudential theories.
Thus, although the advertised message of the article is that ordinary legal methods for ascertaining the law can resolve difficult interpretive questions, Baude and Sachs’s most concrete attempt to illustrate their message points to a different conclusion — difficult interpretive questions can only be resolved by the way in which the content of the law is determined at the fundamental level, that is, by what theories of law seek to specify.
The law of interpretation, like ordinary first-order legal standards, depends for its existence on the way in which the content of the law is determined at the fundamental level by more basic facts. Consequently, appealing to the law of interpretation to explain how the content of the law is determined does not do deep theoretical work. We can put the point in terms of Baude and Sachs’s skeptics. Skeptics who take uncertainty over interpretive methodology to show that the content of the law is indeterminate (to the extent of such uncertainty) cannot be satisfactorily answered by being told to look to the content of the law to resolve interpretive controversies. There is at least as much uncertainty with respect to how to ascertain the content of second-order legal standards as there is with respect to first-order legal standards. In order to answer the skeptics, we need to explain how the content of the law — including the law of interpretation — is determined. Moreover, the way in which the content of the law is determined can answer questions of interpretive methodology even when there are no relevant legal standards.
On Baude and Sachs’s view, a method of interpretation is correct because it is required by law. The more general point, however, is that a method of interpretation is correct because it accurately identifies the law. Consequently, which methods of legal interpretation are correct depends on the way in which legal facts are determined by more basic facts. Second-order legal standards that affect how the content of the law is determined are just a special case of this general point.
If it were somehow possible to identify the law of interpretation on a range of controversial issues without an understanding of how the content of the law is determined, the law of interpretation could do important practical work, resolving disputes over interpretive methodology. Baude and Sachs’s discussion does not justify their confidence in this possibility, however. They have surprisingly little to say about how we are to derive second-order legal standards from practices that themselves involve conflicting interpretive methods.
Their article rightly emphasizes that legal interpretive questions are not linguistic or open-ended normative questions. But the explanation for this truth is not that there are legal standards that resolve the questions but that answers to interpretive disputes depend on how the content of the law is determined at the fundamental level, that is, on which theory of law is true. This point is not widely understood. Most theorists of legal interpretation, rather than appealing to how the content of the law is determined, defend their theories on various grounds: arguments sounding in democracy and rule of law; consideration of which theory would produce the best consequences, for example, which theory would best constrain judges from imposing their preferences; arguments about the very nature of interpretation in general; and linguistic claims. If, on the contrary, as I have argued, difficult interpretive questions depend on which theory of law is true, then relatively ad hoc appeals to moral, instrumental, and linguistic considerations and the like are beside the point. Questions of interpretive methodology ultimately turn on the central legal philosophical question of how the content of the law is determined at the fundamental level.54
* Professor of Law and Professor of Philosophy, UCLA; Faculty Co-Director, UCLA Law and Philosophy Program. I am indebted to Andrea Ashworth, Mitchell Berman, Sarah Burns, Richard Re, and Larry Sager for valuable comments on a draft. I also am grateful to Charles Barzun for sending me his draft paper The Positive U-Turn, which makes related arguments about William Baude and Stephen Sachs’s work. Special thanks to Alicia Solow-Niederman for her thoughtful engagement with this paper and to other editors of the Harvard Law Review for their helpful suggestions.