Derivative Recognition and Intersystemic Interpretation

Abstract

In intersystemic cases, a court applies the law of a foreign system. Scholars have argued that the court ought to use the interpretive methodology of the foreign system’s courts. I argue against that intuitive position. First, interpretive methodology is not bound up with primary rights and duties such that it constitutes substantive law for conflict of laws purposes. Second, although interpretive methodology has epistemic value and may affect case outcomes, a given methodology might not have the same epistemic value or the same effect on outcomes for differently situated interpreters. Further, the approach that the foreign judges take to interpreting their own law is necessarily anchored to the foreign system’s rule of recognition, which is not true of the approach of external judges. Descriptive facts might align such that external interpreters would have to use the internal methodology to identify the applicable law, but that’s an empirical question the answer to which will vary from case to case.

I. Introduction

In the ordinary context of adjudication, a court applies the law of its own jurisdiction to the dispute before it. Under a Hartian positivist theory of law, the norms applied by the court are legally valid just in case they satisfy the criteria of validity provided by the jurisdiction’s rule of recognition, whether directly or indirectly through another rule. The law is thus ultimately determined by the content of the rule of recognition. This ordinary context involves what H. L. A. Hart called original recognition, because the law that is recognized is original to the jurisdiction where the recognition occurs.¹

But Hart also identified a different kind of recognition, which he called derivative recognition. ² This comes into play in the context of intersystemic adjudication, where a court has the task of applying the law of another jurisdiction. This setting is known as conflict of laws or choice of law or, in the case of a court applying a foreign country’s law, private international law. Here, the applicable legal norms are not valid in virtue of the rule of recognition existing in the law-applying court’s own legal system; instead, they are valid in virtue of the law-supplying system’s rule of recognition. In this sense, the validity of the law that is recognized is derived from the foreign system.

Hart said very little about derivative recognition besides pointing out that it is meaningfully different from recognition in the ordinary, original sense and noting that the distinction “needs further elaboration.”³ This article is an attempt to explore the difference between original and derivative recognition in Hartian terms, and to examine the implications of this difference for legal interpretation in the intersystemic context.

When a court is called upon to decide a case arising under another jurisdiction’s law, the prevailing conflict of laws rules require that court to apply the foreign jurisdiction’s substantive law. In the United States alone, this requirement is commonly triggered: state courts are often required to apply the laws of other states, as well as federal law; federal courts are often called upon to apply state law; and both state and federal courts are sometimes required to apply the laws of other countries. In such intersystemic cases, judges interpret and apply the law of other systems. Even when it is clear which system’s law applies, however, it may be far from clear how the court is supposed to go about identifying the content of that law.

Legal scholars have recognized the significance of the intersystemic interpretation problem, but the leading solution that has been offered remains undertheorized. In recent years, American scholars—most notably Abbe Gluck and Aaron-Andrew Bruhl—have maintained that if a court has a legal duty to apply the substantive law of another jurisdiction, and the judges of that other jurisdiction have a prevailing or consensus methodology for statutory interpretation, then the court ought to apply that interpretive methodology.⁴ Focusing on the American federal–state conflict of laws context—where federal conflict of laws rules provide that state substantive law applies to disputes arising under state law in federal court—they claim that federal courts are generally duty bound to use a state’s interpretive methodology to identify the content of the state’s substantive law. While this view has a lot of intuitive appeal, it has not been subjected to sufficient theoretical scrutiny. I aim to do that here. The exercise reveals that the position needs, at the least, to be heavily qualified.

Note that these commentators seem to assume that if judges are under some legal duty in the course of adjudication, they ought to do their best to fulfill that duty. I will do the same. I take Hartian positivism to be a theory of how law is constituted and not a theory of adjudication. The theory does have implications for adjudication, however, once we introduce certain assumptions, such as the one here that judges have a duty to comply with the adjudicative legal norms that apply to them. If they do have such a duty and they are legally required to apply a particular body of substantive law, then (if positivism is true) judges will fulfill the former duty if and only if they identify and apply norms that satisfy the criteria of validity for that body of law.

I do not offer a defense of Hartian positivism; however, something like it is probably the dominant theory of law today, and at least some of the legal scholars against whom I argue have indicated that they are positivists.⁵ At a high level, my argument takes the following form: if one accepts the basic tenets of Hartian positivism, then one should have doubts about the prevailing scholarly view about interpretive methodology in intersystemic adjudication. My central claim that judges applying another jurisdiction’s law need not employ the interpretive methods of the judges in the law-supplying system is, I think, more intuitive and easier to demonstrate under non-positivist theories of law—for example, Ronald Dworkin’s law as integrity or Mark Greenberg’s moral impact theory.⁶ Positivism would seem to cause more trouble than other plausible theories for the conclusion I want to defend, and for the sake of economy I set aside other theories here.⁷

By legal interpretation, I mean the identification of the legal meaning of a source of law or the legal propositions that the source stands for; put differently, legal interpretation is the process of ascertaining the content of the law—the legally valid norms—provided by some legal source, such as a statute or judicial decision.⁸ By statutory interpretive methodology I mean methods, techniques, or means that interpreters may use to identify the statutory or statute-based legal norms that apply to disputes. Examples are the so-called linguistic or textual canons (such as the rule against superfluities and scrivener’s error) and the use of legislative history, as used to ascertain the legal content of statutory provisions. My main claims about interpretive methodology do not apply to so-called “normative” or “substantive” canons such as the rule of lenity, although those canons are sometimes referred to as “interpretive methods.” The latter norms establish or modify the rights and duties of ordinary people in a way that the former ones do not, or so I will argue. Further, “interpretive methodology” sometimes refers to theories about the content of the criteria of validity. Originalism, for example, is like this. It is sometimes called an interpretive methodology even when used to refer to the view that a norm is ultimately valid if it reflects a legal provision’s original meaning (or something to that effect).⁹ I do not mean to include such substantive views about the actual or ideal content of the criteria of validity in my claims about interpretive methodology.

I will argue that, under a Hartian positivist theory of law, a judge who is under a legal duty to apply the substantive law of another jurisdiction may not be bound to apply that other jurisdiction’s interpretive methodology. Perhaps in some cases the judge will have reasons or even a duty to apply the foreign jurisdiction’s methodology because doing so will help enable them to identify the applicable law (again, assuming that judges are duty bound to comply with the legal requirements that apply to them), but whether they do have such reasons or a duty in a given case will depend on contingent descriptive facts.

To preview my argument, it goes roughly like this. First, interpretive methodology does not constitute primary rights and duties and so is not part of a system’s substantive law. Second, the fact that interpretive methodologies have epistemic value and may affect case outcomes does not mean that they travel with substantive law across systems. This is because (1) various methods of interpretation might be capable of ascertaining the very same body of law; (2) the judges of a system might have converged on a suboptimal interpretive methodology, and an external court seeking to interpret that jurisdiction’s law may thus have reason to use an alternative one; and (3) even if the internal judges follow an interpretive methodology that is optimal for them, the effectiveness of a given methodology will depend somewhat on the identity and institutional context of the interpreters, which differ systematically between internal and external judges (e.g., a state’s judges and federal ones). Further, the approach that judges take to interpreting their own law is necessarily anchored to their own system’s rule of recognition in a way that the approach of external judges is not; this is because the internal judges will ultimately appeal to their rule of recognition as justification for their interpretive approach, whereas external judges need not appeal to that rule in the same way. I conclude that courts operating in the intersystemic context are not under any general duty to employ the law-supplying system’s interpretive methodology.

While I focus on federal–state intersystemic adjudication in the United States, in relevant respects this context would seem to mirror other intersystemic contexts, including that of courts in one American state applying another’s law and of courts in one country applying another’s law. If I am right about that, then my conclusions would carry over to those other contexts. In any event, the analytical framework that I develop here is not context-specific, but rather can be used to draw conclusions about how judges in other jurisdictions ought to approach the interpretation of external law.

The remainder of this article unfolds in three sections, as follows. In Section II, I explain how, in Hartian positivist terms, when conflict of laws rules require a forum court to apply a foreign jurisdiction’s substantive law, the validity of the norms that the court is legally required to apply is determined by the foreign system’s criteria of validity. In Section III, I take up the question of how the forum court ought to ascertain the norms that are so validated. Through an analysis of the relationship between interpretive methods and rules of recognition, I argue that, even if a system has an established methodology for statutory interpretation, external judges might effectively apply that system’s law without employing its interpretive methodology. Given that the relationship between interpretive methodology and the rule of recognition is contingent on descriptive facts about the practices of a system’s legal officials, in Section IV I explore other possible forms that the relationship between the two might take and the implications for intersystemic interpretation. I tentatively conclude that, as a conceptual matter, regardless of the form this relationship takes in the law-supplying system, external interpreters could in principle successfully identify and apply the law, and fully discharge their conflict of laws duties, without employing the law-supplying system’s interpretive methods.

II. Conflict of Laws in a Positivist Frame

Roughly, in the intersystemic context, what I call the law-applying or forum court has a duty to interpret and apply another (the law-supplying or target) jurisdiction’s law. Where does the duty come from? That depends on the intersystemic setting, but generally it comes in one way or another from the forum court’s domestic law.

Federal conflict of laws doctrine, with sources in federal legislation, the Constitution, and precedent, provides that when claims arising under state law are adjudicated in federal court, the court must apply state substantive law.¹⁰ State substantive law is understood as the law governing primary conduct and includes state constitutional, statutory, and judge-made law.¹¹ In a positivist framework, this means that federal courts exercising jurisdiction over claims arising under state law legally ought to ascertain and apply norms that satisfy the state’s criteria of validity for substantive law. This is assuming, I should note, that the state law is metaphysically determinate—that is, that applicable, legally valid norms exist and that those norms determine an outcome. This will not be true in all cases; when state law is indeterminate—that is, there is a gap in the law—the federal court’s duties are different, but for the purposes of this article I set aside such cases.

In Hart’s terms, when federal judges adjudicate a dispute arising under state law, we have a case of derivative recognition. Here is how Hart explained the distinction between derivative and original recognition while noting that it “needs further elaboration”:

In an ordinary case where there is no foreign element, for example, where an English court simply applies an English statute, the court does not base its recognition and application of the statute on the fact that courts of some other country have recognized or would recognize it; this is original recognition. But where, as in cases raising questions of private international law, part of the court’s reasons for recognizing a law is that it has been or would be originally recognized by the courts of another country, this is derivative recognition of the foreign law.¹²

I think Hart must have meant that, in the intersystemic context, the law-applying court is to apply the norms that are recognized by the law-supplying system’s rule of recognition. This interpretation would be consistent with Hart’s theory of law under the assumption that, in the intersystemic context, the court’s job is to apply the foreign system’s actual law, which is generally how the law-applying court’s role is understood.

But if the court is to apply norms that are legally valid according to the foreign system’s rule of recognition, this means the court need not necessarily apply the norms that have been or would be recognized by the courts of the foreign jurisdiction. After all, and as Hart acknowledges, judges are fallible: they might sometimes intentionally or inadvertently violate their own rule of recognition, applying an invalid norm or failing to apply a relevant norm that is legally valid.¹³ Instead, the forum court’s legal duty is to apply the norms that should be recognized by the foreign system’s judges according to their own rule of recognition. Those norms, and not necessarily those that the foreign judges have applied or would apply, constitute the valid laws of the foreign system.¹⁴ The correct norms to apply typically will coincide with those the foreign judges have applied or would apply, however, because those judges will successfully follow their own rule of recognition most of the time—that is necessary for it to be a rule of recognition.¹⁵

Further, once a jurisdiction’s high court applies a particular norm to a dispute, then (supposing the jurisdiction has a doctrine of binding precedent) external judges deciding a subsequent similar case will be legally bound to apply that norm.¹⁶ But this is because of the authority of precedent and not because the interpretations of the jurisdiction’s own judges are necessarily legally correct to begin with. The question of which methods of statutory interpretation a federal court ought to apply to a state statutory provision will arise only for statutory interpretation questions that have not been authoritatively resolved by the state’s courts. The scope of argument of those who maintain that federal courts ought to use state interpretive methodologies when interpreting state statutes is limited in the same way. Indeed, Gluck, Bruhl, and I all agree that if the state has already resolved the matter through binding judicial precedent, then the federal court is legally bound to apply the precedent and has no reason to engage in direct or first-impression statutory interpretation.¹⁷

But it is not so uncommon for a federal court to be charged with answering a question of state statutory law that has not already been decided by binding state precedent.¹⁸ We see this not only in diversity jurisdiction cases (which currently represent about 40 percent of cases in federal court)¹⁹ but also when state statutory provisions arrive in federal court by way of supplemental jurisdiction or embedded federal question jurisdiction, or when a federal court is addressing the constitutionality of state law. The federal court would have to interpret the state statutory provision for itself if the state’s judges have not already applied the provision in a binding decision to facts materially the same as those before the federal court. When a court faces such a question, its job is to identify the norms that the jurisdiction’s judges should apply according to their own rule of recognition if they were faced with the same case, which will usually—but not necessarily—be the norms that the state’s courts have applied or would apply.²⁰

Indeed, one of the main reasons—and probably the most cited one—for the existence of diversity jurisdiction in the United States is to protect litigants against possible defects of adjudication in state courts and, to that end, to provide them access to a real alternative forum—not a replica of the state forum that they would have access to anyway, but one more likely to fairly and neutrally adjudicate the case.²¹ The grant of diversity jurisdiction does not confer a right to have federal judges apply the norms that the state’s judges would apply or to reach the outcome that the state’s judges would reach. Nor does the state have a right that the federal judges adjudicate the case in the way the state would want them to, as some scholars maintain.²² Instead, what the litigants and the state have a right to is for the federal court to apply the true state substantive law to the dispute while using federal modes of adjudication.²³

The picture I have been painting here is complicated by the fact that one of the express aims of the landmark conflict of laws case of Erie Railroad v. Tompkins and related decisions is to prevent untoward “forum-shopping,” where litigants would choose between state and federal court based on where they anticipated getting more favorable outcomes. That aim gave rise to the “outcome-determinative” test for distinguishing substantive from procedural law. Recall that federal courts are to apply state substantive law but federal procedural law.²⁴ The outcome-determinative test provided that, if the application of some state legal norm would change the outcome of the case, then the federal court would have to apply the state norm, treating it as substantive, even if it otherwise seemed to be procedural in nature.

But, shortly after announcing the outcome-determinative test, the Supreme Court recognized that it was overly simplistic and qualified it accordingly across a series of cases. The doctrine now provides that federal courts must apply the primary rules of conduct that govern in a state and, because some rules that seem procedural (e.g., statutes of limitations) partly determine primary rights and duties, federal courts must apply some seemingly procedural state rules. But that does not mean that the federal court must necessarily reach the result that the state’s courts would.²⁵ Again, like all judges, state judges are fallible. A legal requirement that federal judges apply the norms and reach the result that a state’s judges would is not the same as, and will sometimes be inconsistent with, the legal duty of federal judges to apply the genuine primary rules of conduct of the state and reach the result dictated by those rules. It is uncontroversial that prevailing conflict of laws doctrine requires federal courts to apply a state’s extant primary rules of conduct—this is widely understood to be the fundamental duty of federal courts in the conflict of laws context, a duty based in the federal Constitution.²⁶ So the idea that a federal court discharges its conflict of laws duties if and only if it does with a case what the state court would do is untenable.

One could nevertheless argue that federal courts should simply reach the same result as the state’s courts would, because only then could forum shopping be eliminated. And the Supreme Court has observed that federal courts may sometimes choose to defer to a state procedural law as a policy choice to help ensure uniform outcomes in federal and state court. But that kind of uniformity is not a constitutional requirement on federal courts, nor is it a right of litigants.²⁷ Further, a rule requiring that litigants get the same treatment in state and federal court would defeat the point of having diversity jurisdiction to begin with.²⁸ Recall that the very point of that grant of jurisdiction is to provide litigants with a true choice of forum; that choice would be hollow if federal courts in diversity cases were just state courts in fancy dress. The Supreme Court’s efforts to limit forum shopping are directed at a specific kind of strategic court selection: litigants choosing the court that would apply the substantive law most favorable to them. That kind of forum shopping was rampant before Erie because before that case was decided federal courts applied “general” common law to state claims and did not recognize the authority of state common law. If a litigant preferred general common law to a state’s common law, then they could choose federal court. Erie and subsequent decisions were aimed at preventing that kind of forum shopping, but not forum shopping for other purposes—for example, for the purpose of getting a fairer forum or one more likely to apply the law correctly. Those would be legitimate reasons to choose one forum over another—indeed, they are the very reasons why litigants are given the option of federal court to begin with.

Although I focus here on federal-state conflict of laws in the United States, my main points are generalizable to other intersystemic contexts, including the American interstate and the private international law ones. The domestic rules that govern when foreign law applies to a case, and what sort of foreign law applies, might diverge in the details, but the basic principles would seem to be the same: the forum court applies forum procedural law but foreign substantive law.²⁹ When a court in one American state adjudicates a dispute arising under another’s law, the court applies the foreign state’s substantive law but its own procedural law and does not follow a pure outcome-determinative test for distinguishing substance from procedure; instead, the law-applying state court aims to identify and apply the actual primary rules of conduct that govern in the foreign state. Likewise, when a court within the United States adjudicates a dispute arising under a foreign country’s law, it seeks to identify and apply the actual substantive rights and duties of the foreign country, and not necessarily to reach the result that a foreign court would.³⁰

It is undisputed that federal conflict of laws doctrine requires federal courts to identify the relevant state’s substantive rules of conduct and apply those rules to the disputes before them. As I try to show in the next section, state interpretive methodologies are not part of the substantive law that federal courts must apply, nor are they necessary to the accurate identification and application of that law.

III. Interpretive Methodology and the Rule of Recognition

Now that we have a forum court’s conflict of laws duty in a positivist frame, as a duty to apply substantive norms that are valid in virtue of the target system’s rule of recognition, we can turn to our main concern: the question of how the forum court ought to ascertain the target law, and in particular whether the court ought to use the target system’s interpretive methodology.

On what has apparently become the dominant American scholarly view, federal courts are generally duty bound to apply a state’s methodology to disputes arising under the state’s law. Scholars defending this view sometimes suggest that interpretive methodology is a type of substantive law—because it is bound up with primary rights and duties, or affects outcomes in such a way that it qualifies as substantive for conflict of laws purposes—and so the federal court’s legal obligation to apply state substantive law extends to it.³¹ At other times they suggest that applying the state’s interpretive methodology is necessary to effectively identifying the state’s substantive law, and that federal courts therefore ought to use the state’s methodology.³² These two claims are not kept separate in the literature, but they are conceptually distinct. I will question both.

Recall that by statutory interpretive methodology I mean methods, techniques, or means that interpreters may use to identify the statutory or statute-based legal norms that apply to disputes. Because some norms that I do not mean to include in this category might be seen as fitting this description, I want to pause to describe the category in more detail and defend the boundaries I draw.

A norm identified through an interpretive methodology, if valid, is so not because of the methodology employed but rather because of other conditions. Consider the rule that an American federal statute must be approved by a majority of the House of Representatives and of the Senate and signed by the President to become law. Those are conditions of validity for federal statutory law in the United States. Legal officials might treat, for example, the appearance of a statute in the United States Code as evidence that the statute was in fact passed by a majority of both houses and signed by the President. But the appearance of the statute in the Code does not constitute the fact that it went through the proper legislative process. And if there was evidence that the Code was unreliable—perhaps it had been tampered with—then that would be a basis for questioning and criticizing reliance on it. Use of the Code as a means of discovering the content of statutory law is the kind of norm that I have in mind when I refer to interpretive methodology.

This conception of interpretive methodology overlaps, but is not completely coextensive, with Gluck’s and Bruhl’s. They operate with a looser and more capacious notion of statutory interpretive methodology, which includes my sense but also extends to norms that are often referred to “substantive ” or “normative,” such as the rule of lenity in criminal law and the analogous rule that ambiguous bankruptcy laws must be interpreted in the debtor’s favor.³³ While the latter rules are sometimes referred to as “interpretive methods,” they seem to operate as substantive common law rules, establishing or modifying the rights and duties of ordinary people. The rule of lenity, for example, provides that if a statute is vague or ambiguous, then, of the plausible alternative norms that could be derived from the statute, the valid one is that which is most favorable to the defendant.³⁴ Differently put, the norm provides that, if a statutory provision can reasonably be interpreted in multiple ways, then one’s conduct must conform to the interpretation that is most favorable to defendants. In contrast, ordinary people do not have legal rights and duties in virtue of, for example, rules concerning the proper use of legislative history.

This is not to say that something like the rule of lenity couldn’t be used as a means of identifying norms that satisfy some distinct criteria of validity, such as reflecting legislative intent.³⁵ My distinction depends on how different rules are in fact perceived and practiced by the legal officials in the system of interest. When litigants in their arguments and judges in their opinions rely on the rule of lenity, the rule is treated as a substantive common law rule. A statutory norm is valid only if it satisfies the requirements set out in the rule of lenity; that rule, in turn, is validated by precedent. In contrast, statutory norms uncovered through nonsubstantive interpretive methods are not validated by the methods themselves, but rather by independent criteria of validity; the methods are means to identifying norms that satisfy those criteria. Judges could also treat nonsubstantive interpretive methods as a matter of precedent; however, the precedent would be seen as defective and would be subject to criticism and revision to the extent that and on the ground that following it led judges to misidentify valid statutory norms. The rule of lenity could, of course, also be revised, but its revision would constitute a change in primary legal rights rather than an attempt at a more effective means of identifying preexisting ones.

The rule of lenity, then, like other so-called substantive canons, is constitutively bound up with primary rules of conduct in a way that what I refer to as interpretive methods are not. This picture follows Benjamin Eidelson and Matthew Stephenson’s description of a substantive canon as one that “purports to speak to a statute’s proper legal effect in a way that is not mediated by its evidentiary bearing on what a reasonable reader would take a lawmaker to have said in enacting the statute” (where what a reasonable reader would take a lawmaker to have said constitutes the criteria of validity of statutory law in the system of interest).³⁶ So-called linguistic or textual canons, in contrast, such as ejusdem generis (providing that, if the initial terms of a statutory provision “all belong to an obvious and readily identifiable genus,” interpreters should assume that the rest of the provision also applies exclusively to that genus),³⁷ the rule against superfluities, and scrivener’s error, are invoked for and justified in terms of their evidentiary bearing on some further criteria—such as legislative intent or ordinary public meaning.³⁸ A norm derived from a statute is not valid because it complies with, for example, the scrivener’s error canon; rather, it is valid because it satisfies some independent criteria of validity (such as reflecting legislative intent or ordinary public meaning). Judges commonly and publicly justify the use of this and other nonsubstantive canons in terms of their utility in helping to identify norms that satisfy separate criteria of validity for statutory law.³⁹

On my view, courts operating in the intersystemic context are generally legally required to apply the substantive canons of the law-supplying jurisdiction, just as they are required to apply other substantive norms of the system. To this extent, I agree with Gluck and Bruhl. My disagreement with them concerns nonsubstantive methods. And again, it is this kind of method that I have in mind when I refer to interpretive methodology.⁴⁰ Further, I do not mean to include substantive views about what the criteria of validity in fact look like or ought to look like. If, say, original public meaning is a condition of validity for the substantive law in a jurisdiction, then an external court interpreting that law ought to identify norms that reflect original public meaning. I will argue, however, that it does not follow that the external court ought to use the same means of identifying that meaning as courts in the law-supplying jurisdiction would use.

In the remainder of this section, I examine how interpretive methodology operates in federal and state systems in the United States to show how interpretive methodology does not travel with substantive law across systems in the way that scholars have concluded it does.

A. Intrasystemic Interpretive Disagreement

First, let’s examine the federal system in the United States, where judges have long displayed widespread disagreement on statutory interpretive methodology.⁴¹ This will show how interpretive methodologies can play an epistemic role in identifying the law as opposed to a constitutive role in determining it, and how multiple methods of interpretation may be capable of identifying the very same law. American federal judges disagree, for example, on the role of legislative history in statutory interpretation—whether it should be used at all and, if so, under what circumstances and which types.⁴² Yet few would deny that we have a system of federal statutory law. If such a system exists, then, despite methodological disagreement, judges must be following the same rule of recognition. (Recall that we are assuming Hartian positivism is the correct theory of law.)

What ultimately divides federal judges who disagree on interpretive methodology, then, seems not to be disagreement about how statutory law is constituted but rather disagreement about what kind of tools and empirical assumptions—what modes of identification—will best enable them to ascertain and apply statutory law. This indicates that interpretive methodology is not part of the statutory law and that multiple interpretive approaches may be capable of identifying the very same body of statutory law.

Other theorists have argued that, although federal judges disagree on statutory interpretive methodology—for example, on the question of whether and when consulting legislative history is appropriate—they are generally after the same thing, which is to ascertain the enacting legislature’s intent.⁴³ Now, legislative intent is a matter of controversy. Some contend that the idea that a multi-member legislature has shared intentions is untenable, as the attitudes of individual members are varied and non-aggregable.⁴⁴ In response, theorists have developed conceptions of legislative intent that do not depend on the aggregability of individual intentions and that would make sense of judges’ common appeals to legislative intent as the guiding aim of statutory interpretation. Ryan Doerfler makes a compelling case for the position that we should understand legislative intent as a fiction: “Statutory interpretation as practiced,” he observes, “involves widespread attribution of legislative intent”; and he argues that such claims are “best understood as involving a useful fiction,” which posits “that legislation is written by a generic author.”⁴⁵ On that supposition, “a claim about legislative intent is apt if and only if one would make the claim about a generic author on the basis of her having written the legislation at issue in the context of enactment”; this is how the truth of claims regarding intent is to be evaluated.⁴⁶ Judges probably do not have such a sophisticated conception of legislative intent in mind. But that doesn’t matter. Fictional intent could serve as a condition of validity for statutory law so long as judges have something like it in mind when they engage in statutory interpretation, the condition serves as a shared standard for evaluating the validity of statutory norms, and the norms derived from statutes and taken to be valid generally meet the condition.⁴⁷

There does seem to be a turn afoot among federal judges with growing reluctance to invoke legislative intent and increasing appeals to the ordinary meaning or public meaning of statutory text as the true content of statutory law.⁴⁸ If federal judges are coming to embrace ordinary meaning as the objective of statutory interpretation (and if that refers to something substantively different than legislative intent), then the criteria of validity for statutory law may be changing. But this would not affect the point I am trying to make in this section, which is just that alternative interpretive methods can be, and in fact are, used for the same ends. Indeed, judges also disagree about how best to identify the ordinary meaning of statutory text. For example, in a recent oral argument, Chief Justice Roberts suggested that, if the “objective [of statutory interpretation] is to settle upon the most natural meaning of the statutory language to an ordinary speaker of English,” then “the most probably useful way of settling [the matter] would be to take a poll of 100 … ordinary speakers of English and ask them what [the statute] means.”⁴⁹ Perhaps that kind of poll would be an effective way to identify ordinary public meaning (at least on some understanding of what that means), but it’s hard to imagine that it would be a singularly effective method.

My argument does not rest on any particular view of the content of the criteria of validity for statutory law, but rather on the claim that, for many conceivable criteria of validity (such as fictional intent and ordinary public meaning) multiple methods of interpretation are plausible candidate means of identifying norms that satisfy those criteria. I do not mean to suggest, however, that there is no necessary connection between a system’s rule of recognition and the interpretive methodology used by its judges. For one, judges will criticize and defend methods in terms of their value in surfacing norms that satisfy the criteria of validity—in this sense, those criteria will serve as the subjective, public justification for the methods used. (Methods will be objectively justified to the extent they succeed in surfacing valid norms.) Further, the methods the judges use must actually serve (in general, or most of the time) to identify norms that satisfy the criteria of validity—and so the methods used might affect the criteria of validity in a causal way.

Turning back now to intersystemic adjudication, what interpretive methodology should external judges—for example the judges of some state—use when adjudicating a question of federal statutory law? They should use whatever methodology would best enable them to identify norms that satisfy the criteria of validity for substantive federal statutory law. Given the widespread disagreement among judges about the proper methods of statutory interpretation and the different methodologies in regular use, federal practice indicates that no single methodology is constitutive of federal statutory law and that multiple methods could in principle identify it.

We cannot say what kind of methodology a state’s judges should use on federal law—for example, a strict textualist approach or one that embraces the use of legislative history, or some other approach entirely—without knowing more about the epistemic position and institutional setting of the state’s judges. Descriptive facts such as the interpretive resources available to the judges, and their experience and expertise—which differ systematically and considerably between federal and state judges, and between the judges of different states too—will determine, in part, how effective a particular methodology will be at identifying the applicable legal norms.

B. Intrasystemic Interpretive Agreement

Federal judges have not converged on an interpretive approach for statutory law, so one might interject at this point that of course external judges applying federal statutes are not required to apply a particular methodology dictated by federal practice. But in some states, the skeptic might continue, judges have converged on interpretive methodology and in some states certain methodologies are even seen as authoritative. When charged with applying the law of such a state, our skeptic might conclude, external judges would have to use the state’s interpretive methodology to identify the law.

This is precisely the position of scholars like Gluck and Bruhl, who argue that federal judges ought to use prevailing state interpretive methodologies when adjudicating questions of state law. These norms, they maintain, are bound up with primary rights and duties and so form part of the state’s substantive law for conflict of laws purposes.⁵⁰ They also suggest that these norms are necessary means to the accurate identification of the state’s substantive law.⁵¹ If either of these claims is true, then federal judges would be duty bound to apply state interpretive methodology. I will argue that the former is false as a conceptual matter. While the latter could nevertheless be true, I’ll argue that we should reject it. I take the state of Oregon as a primary case study to demonstrate these points. My reason for doing so is that the state would seem to represent the strongest case against my position: scholars who claim that interpretive methodology is intersystemically binding rely on Oregon as the clearest example of a jurisdiction with an authoritative methodology that, they argue, external judges charged with applying Oregon law ought to follow.

In the past, Oregon state judges subscribed to a textualist approach that directed judges to ascertain legislative intent from the statutory text and context (text of related provisions) alone and to consult legislative history if and only if intent is not clear from the initial plain text inquiry. (If we think that ascribing an actual intent to the state legislature is problematic, we can understand the judges to mean the kind of fictional intent described above.) This norm was first expressly delineated in the Oregon Supreme Court case of Portland General Electric Co. v. Bureau of Labor and Industries (PGE).⁵² There the Court explicitly and repeatedly asserted that the objective of the interpretive methodology was to ascertain the legislature’s intent. In subsequent decisions where Oregon courts followed PGE’s interpretive approach, the judges continually and emphatically invoked this objective, claiming that disputes about statutory interpretation are ultimately disputes about “the intended meaning of the statutes” at issue.⁵³

Some sixteen years after its decision in PGE, the Oregon Supreme Court expounded a modified version of the PGE methodology in the case of State v. Gaines. ⁵⁴ The court took guidance from a state statutory directive providing that courts may consider legislative history regardless of textual ambiguity when interpreting statutes; that directive had been issued eight years prior but had apparently remained inert until the Gaines decision. Even if the statutory text was clear, the court announced, judges were free to consult and put weight on legislative history as they saw fit. Despite the influence of the legislated directive here, the justices maintained that, “This court remains responsible for fashioning rules of statutory interpretation that, in the court’s judgment, best serve the paramount goal of discerning the legislature’s intent.”⁵⁵ According to the justices in Gaines, the PGE framework was flawed because it directed judges to ignore legislative history in cases where that history might be probative of legislative intent.⁵⁶ The methodology was revised in an attempt to better ensure that statutes “are interpreted in the way that the legislature intended.”⁵⁷

The court in Gaines did not alter (or even attempt to alter) the criteria of validity for statutory law in Oregon—the judges reiterated their ongoing commitment to ascertaining the norms that the legislature intended to establish in enacting the provision at issue—but the court did seek to alter, and in fact ended up altering, the way Oregon judges would go about ascertaining legislative intent. Both the state legislators and Supreme Court apparently believed that judges would be more likely to accurately identify legislative intent if they followed the new methodology. Their problem with the previously embraced methodology was that it was a flawed means of ascertaining the law; in positivist terms, it was not ideal for identifying norms that satisfied the system’s criteria of validity.

The state judges’ references to legislative intent (and the way they described intent) remained constant across the shift in interpretive methodology. It seems, then, that the methodological shift was not accompanied by a change in the criteria of validity for statutory law. If that is right, then the earlier methodology must have been good enough at identifying norms that satisfy the state’s criteria of validity. (Otherwise, those would not have been the criteria!) But the judges may have been right that the new methodology was more effective.

The Oregon example shows how, although interpretive methodology is distinct from the rule of recognition, it is not completely independent of it. A jurisdiction’s methodology will inevitably be based to some extent on the content of its rule of recognition because the main purpose of the methodology, from the judges’ point of view, is that it helps them to ascertain norms that satisfy the criteria of validity.⁵⁸ Methods are criticized and found to be defective to the extent that they do not work to identify norms that satisfy those criteria, and methods are defended and found to be justified to the extent that they do serve to identify such norms.⁵⁹

In the state of Oregon, judges treat what they refer to as the enacting legislature’s “intent” as constitutive of their state’s statutory law. They do not appeal to any further law that would ground or validate that practice; it would seem to represent a social rule and a rule of recognition for Oregon judges. The rule of recognition, Hart wrote, “will be manifest in the general practice, on the part of officials … of identifying the [legally valid rules] by this criterion”: a system’s “rules of recognition specifying the criteria of legal validity … must be effectively accepted as common public standards of official behavior by its officials.”⁶⁰ Under both PGE and Gaines, Oregon judges continually and publicly (in their judicial opinions) appealed to legislative intent as the ultimate basis for assessing and defending the validity of the legal norms they applied to disputes.

To return to our intersystemic inquiry, we want to know whether Oregon’s interpretive framework is part of its substantive law for conflict of laws purposes such that federal judges have a legal duty to apply Oregon’s interpretive framework when adjudicating questions of Oregon statutory law. Or, even if the framework is not part of the state’s substantive law, whether federal judges nevertheless ought to apply it because it is a necessary means of identifying the state’s substantive law.

Consider the following hypothetical. Shortly before the Oregon Supreme Court’s decision in Gaines, a federal court sets out to interpret an Oregon statute. It studies the practices of Oregon judges and discerns, correctly, that Oregon judges take a proposed statutory norm as legally valid only if it reflects the norm that the legislator intended to establish in enacting the statute at issue. Suppose that, from the point of view of the federal judges, the statutory text is unambiguous and the legislative intent is clear based on the text alone. Nevertheless, the judges review the relevant legislative history to see if it might shed further light on legislative intent. In doing so, they discover that they made an incorrect assumption about the intended meaning of some key term in the statute, and they adjust their interpretation of the legal norms accordingly. According to the prevailing view among legal scholars, the federal court would be in dereliction of its duties here, because it did not follow the interpretive methodology delineated in PGE (which prohibited the use of legislative history in the event that the court finds the text to be unambiguous).

Now suppose, alternatively, that the federal court proceeded in exactly the same way but just after the Gaines decision, in which the Oregon Supreme Court relaxed the PGE framework to permit consideration of legislative history even in the absence of textual ambiguity. Under the prevailing scholarly view, here the federal court would have properly discharged its duties.

But the modification of the PGE methodology in Gaines, by the lights of both the state legislators and Supreme Court justices, was supposed to help enable Oregon’s judges to correctly identify legislative intent, which was already the objective of statutory interpretation well before that decision (and perhaps all along). It’s hard to see how the federal judges could have erred in their application of Oregon law by using an interpretive approach that Oregon legislators and judges were about to acknowledge was superior to the PGE methodology for the purpose of identifying the state’s statutory law.

Now suppose one further thing: that if the litigants had chosen to have their dispute adjudicated in Oregon state court, the Oregon judges would have followed the PGE framework; taking the meaning of the statute as clear based on the text alone, they would have derived the same norm as the federal court derived initially (before it consulted legislative history and realized that the norm identified from the text alone did not accurately reflect legislative intent). The difference in interpretive methodology, then, would mean that a different norm would be applied in state and federal court in the very same case, perhaps leading to different outcomes. In that sense the choice would be “outcome determinative.” But the state court is the one that would have applied the wrong legal norm—wrong by the lights of its own rule of recognition. If the federal court’s duty is to apply the state’s true substantive law, then here it should not follow the methodology that the state courts would use.

Even if a state’s judges have a consensus interpretive methodology and that methodology affects which primary rules of conduct will be identified as legally valid, a federal court is not necessarily duty bound to apply the state methodology. It might even be required to depart from the state’s interpretive approach, because that approach could interfere with its ability to correctly identify the substantive legal norms provided by the state statute at issue—and it is legally required to apply those norms.

It is possible that the state legislature is aware of the state judges’ preferred interpretive methodology and might legislate with that methodology in mind, in which case following it might be a particularly effective means of identifying norms that satisfy the state’s legislative intent.⁶¹ To the extent that this is the case (which is an open empirical question),⁶² when dealing with the statutory law of a state like Oregon, where legislative intent seems to be determinative of substantive norms, external judges might do well to take into account and even follow the interpretive methodology used by Oregon judges. This may be why some commentary suggests not that the federal court is necessarily bound to apply the interpretive methods “employed in the enacting state,” but rather that those methods “apply most naturally.”⁶³

Further, because it is the practice of the state’s judges that constitutes the relevant rule of recognition, the federal judges should be careful not to impose their own conception of the state’s criteria of validity (for example, legislative intent) when interpreting the state law; it is the criteria of validity as understood by the state’s judges that determine the content of the state’s law. And so, when interpreting Oregon law, the job of the federal judges is to make out legislative intent according to the conception of intent embraced by the state’s judges. That might require studying Oregon judicial opinions to figure out what the judges mean by legislative intent. But it does not mean that the federal judges must mimic the interpretive methodology of those judges.

To recap this section so far, even when a state has a prevailing interpretive methodology, as in the case of Oregon, that methodology can be distinguished from its substantive law. Further, a state’s methodology might be a suboptimal means of identifying its substantive law, even for the state’s own judges; in that case, federal judges might better discharge their conflict of laws duties by using an alternative methodology.

Further still, even if a state’s methodology is ideal for the state’s own judges, it might not be so for federal judges. For example, suppose the state’s judges follow a strict textualist approach, which prohibits or severely restricts the use of legislative history. Federal judges might be more dependent than the state’s judges on legislative history to identify state legislative intent. This is because the state’s judges probably have a better understanding of their own state’s legislative process and drafting conventions; they also likely have some background knowledge of their state’s legislative history that the federal judges lack.⁶⁴ These epistemic differences might mean that, given the very same case, federal judges would have to consult legislative history to successfully identify the applicable legal norms even if the state’s judges could identify those norms without doing so. And so a state norm restricting the use of legislative history might be compatible with a rule of recognition that recognizes legislative intent as a ground of statutory law, even if external judges would need to consult legislative history to successfully identify intent in many cases. On the other hand, perhaps using state legislative history is infeasible for federal judges or would be likely to lead them astray, given their lack of familiarity with the state’s process of enactment and sources of legislative history.⁶⁵ In that case, federal judges might do best to avoid relying on such sources when interpreting a state statute and instead use, for example, a strict textualist approach, even if the state’s own judges would use legislative history.⁶⁶

Crucially here, the extent to which consulting legislative history is necessary to identifying legislative intent might be contingent on descriptive facts that differ across jurisdictions, like the experience, expertise, and institutional setting of the judges. Identification is an epistemic matter, and we should not assume that the epistemic value of a given means of identification will be the same across differently situated groups of officials.

It is possible that an interpretive methodology serves as an epistemic rule of recognition in a jurisdiction. As Jules Coleman explains, an epistemic rule of recognition comprises “modes of identification” and is the means “by which particular propositions of law are verified.”⁶⁷ This is in contrast to the rule of recognition in what Coleman calls the ontological or semantic sense and I prefer to call the constitutive sense.⁶⁸ The rule in the constitutive sense determines, as a metaphysical matter, what the law is in a legal system or lays out the system’s criteria of legality or the existence conditions for the system’s law. It tells us what makes a proposition a legally valid norm, whereas the rule of recognition in the epistemic sense tells us how legal officials make out or discover legally valid norms. Unlike a constitutive rule of recognition, the existence of a legal system does not depend on the existence of an epistemic rule. Although Hart was not always clear on this, he must have meant the rule of recognition in the constitutive sense and not the epistemic one.⁶⁹

A norm would make an appropriate or effective epistemic rule of recognition to the extent it would succeed at identifying legally valid norms or correctly classifying norms as valid or not. Whether a norm would succeed in that way depends on the context in which it is employed and on the identity of the individual or institution employing it. As Coleman explains, an epistemic rule of recognition governs “how, in a particular community, one [goes] about learning the law”; the content of an epistemic rule depends accordingly “on the available sources of legal knowledge” in a community.⁷⁰ But the available sources of legal knowledge are not the same across jurisdictions, so even if a system has an epistemic rule of recognition, a law-applying court in the intersystemic context does not necessarily have to follow it to make out the applicable law; the court might even do better not to follow it.

The criteria of validity or legal existence conditions that determine the substantive law applicable to a particular case are not contingent on empirical facts about the context in which the case is adjudicated. Whether the case is adjudicated in the intra- or intersystemic context, the criteria of validity will be the same. The appropriate modes of identification for those criteria of validity, however, are contingent on empirical facts about the adjudicative context—facts that likely differ systematically between different legal systems.

Let’s return to Oregon, which formerly followed the interpretive framework delineated in the case of PGE and now (at the time of writing in 2022) follows the modified version of that framework laid out in the case of Gaines. The aim of statutory interpretation in the state appears to be to identify legislative intent. A federal court applying Oregon law would have to use the state’s interpretive methodology if that methodology represented the sole means by which the federal court could successfully identify norms that reflect the state legislature’s intent. It is doubtful that the Gaines methodology represents a singularly effective means of identifying such norms for any group of judges. But it is at least more likely for that to be true for Oregon judges than for federal judges or another state’s judges, who are in different epistemic positions vis-à-vis Oregon’s legal institutions and practices. The Oregon judges, after all, came up with the methodology for their own use, and they are similarly situated to one another in terms of experience, expertise, and resources.

One might object that, because it is the Oregon judges’ conception of intent that constitutes the condition of validity that the norms the federal judges ascertain would have to satisfy, the federal judges will do best to adopt Oregon’s interpretive methodology. That is certainly possible, but I have some doubts.

First, the way Oregon judges describe legislative intent and the kinds of interpretive methods they use to ascertain it suggests that they do not hold an idiosyncratic conception of intent, but rather have the same kind of thing in mind as federal judges do when they refer to legislative intent in their efforts to interpret federal statutes—something like the legal norms that the legislature intends (perhaps in the fictional sense) to establish. So Oregon’s criteria of validity, at least insofar as they include what Oregon judges call legislative intent, should not be obscure to a federal judge. Second, given that state judges adjudicate cases arising under federal law all the time, they will likely be familiar with the federal criteria of validity, so we should expect some convergence between the kinds of criteria operative in the federal and state systems, and we should especially expect convergence in substance where the state judges use the very same terminology as federal judges do, as is the case with legislative intent.

Third and finally, even if a state’s criteria of validity are obscure to a federal judge, that doesn’t mean that the federal judge necessarily ought to adopt the state’s interpretive methodology to identify norms that satisfy those criteria, especially since the methodology itself may be complicated and unfamiliar. If, however, the law-supplying system’s rule of recognition is especially obscure to the law-applying judges, then they might do best to take an approach that allows them to avoid (more or less) engaging in legal interpretation, such as treating the law as a question of fact, trying to predict what the internal courts would do, or even requesting guidance from internal judges (more on this in the next section).

For the purposes of this section, I have focused on the kinds of interpretive norms that make up common interpretive methodologies in the United States and that scholars have claimed are intersystemically binding. These methodologies are distinct from the criteria of validity that comprise rules of recognition. A system’s interpretive methodology will nevertheless be informed by its criteria of validity, since judges adopt a methodology in part because it facilitates their identification of norms that satisfy those criteria. A system’s interpretive methodology might also affect its rule of recognition in a causal way, since the criteria consistently satisfied by the norms that the system’s judges identify through that methodology will constitute the actual criteria of validity in the system. Necessarily, there is an intimate relationship between a system’s interpretive methodology and its rule of recognition.

That doesn’t mean, however, that external judges who are legally required to apply a jurisdiction’s statutory law are necessarily required to use the jurisdiction’s interpretive methodology. The methodology is not bound up with substantive rights in such a way that conflict of laws principles would require deference. Nor will the external judges necessarily be dependent on the jurisdiction’s methodology to fulfill their legal duty to interpret and apply the jurisdiction’s substantive law. Just because norms generated through the chosen methodology tend to satisfy some criteria when used by internal judges, that doesn’t mean an alternative methodology couldn’t be used to ascertain norms that satisfy the same criteria. Further, even if in the hands of internal judges the methodology is singularly effective at ascertaining such norms, it may not be so in the hands of external ones.

Would it be possible, however, for a system’s interpretive methodology to take such a form that, as a conceptual matter, external judges could not successfully interpret and apply the system’s substantive law without using it? I take up that question in the next section, tentatively answering it in the negative. I proceed here by trying to imagine possible forms that interpretive methodology might take such that it would be intersystemically binding. While I fail to come up with any, I certainly may be overlooking some possibility, which is why my conclusion remains tentative.

IV. Epistemic Norms as Constitutive Conditions

First, we could imagine a system’s judges embracing an interpretive norm with content that substantially overlaps with the system’s rule of recognition. Suppose, for example, that a state’s only interpretive norm provides something like seek legislative intent when interpreting statutes. And suppose further that the state’s judges treat legislative intent as authoritative, as a criterion of validity for statutory law. So a rule of recognition in the state would provide something like take a statutory norm to be legally valid if it would effectuate legislative intent. The interpretive norm would effectively direct judges to identify norms that satisfy the system’s rule of recognition. In that event, would the interpretive norm, so far as it goes, be intersystemically binding?

To characterize such a norm as an interpretive methodology would be a stretch, I think; recall that I am using the term to refer to methods or techniques that interpreters may use as means to identifying the statutory or statute-based legal norms that apply to disputes. The seek legislative intent norm would do no more than give judges a sense of what they are looking for without providing any guidance on how to find it. Perhaps the system would best be characterized as one without any interpretive norms, as the judges would be permitted to use any means they see fit to identify norms that reflect legislative intent. What about external judges, then? They need not take such a permissive approach, but could instead constrain themselves in various ways, for the same reasons I gave in the previous section that would justify external judges in departing from the law-supplying system’s internal methodology when it has a circumscribed one.

But would the external judges’ interpretive approach have to be guided by the law-supplying system’s rule of recognition, such that their methods would be chosen with that rule in mind? If so, it might seem that, at a high level at least, the external judges’ interpretive approach would have to match that of the internal judges.

As a conceptual matter, the external judges would not have to be guided by the internal rule of recognition, since their practices do not constitute the rule. As an empirical matter, I’m not sure the extent to which the external judges would ultimately have to be guided by the internal judges’ rule of recognition. That will depend on the legal systems and judges in play. We can say, at least, that the external judges do not need to follow the law-supplying system’s rule of recognition in the same way or to the same extent as the internal judges. After all, the rule is constituted by the practices of the system’s own legal officials in following it—they must in general treat it as a test of legal validity for it to exist. In this sense, the interpretive approach of the internal judges is necessarily anchored to their rule of recognition. Their ultimate subjective justification for their interpretive approach (i.e., what they feel, believe, and claim justifies it) will coincide with the ultimate objective justification for it, which is that it in fact serves to identify norms that satisfy their system’s rule of recognition. In contrast, the rule exists entirely independently of the practices of external judges. The interpretive approach of the latter is, we might say, free floating in this sense, whereas the interpretive approach of judges operating intrasystemically is necessarily anchored to their rule of recognition. The external judges’ awareness of the content of the law-supplying system’s rule of recognition could be quite limited. They would probably need to possess at least some inkling, however, of what the rule looks like to assess the merits of pursuing various interpretive methods and to stand a good chance of getting the applicable law right.⁷¹

The main point here is that the practices of judges in the external system do not constitute the relevant rule of recognition, so their subjective justification for the legal norms they identify and apply need not be based on that rule in the same way as the subjective justification of the internal judges needs to be. The less familiar a law-applying court is with the content of the source system’s rule of recognition or the more obscure that content is, the less helpful it is likely to be to aim for norms that satisfy it. Perhaps this is why, in many systems, the law of other countries is approached as a matter of fact, to be ascertained through fact-finding methods such as the testimony of expert witnesses, rather than as a matter of traditional legal interpretation.⁷²

Federal judges often use a predictive approach to identify state law, which might not directly rely on the state’s rule of recognition at all. On this approach, sometimes called an “Erie guess,” the judges seek to anticipate the content of the legal norm that the state’s courts will apply to like cases in the future. One might contend that prediction of this sort inevitably relies on the state’s interpretive methodology, since the federal judges seek norms that the state’s judges would identify as authoritative and the state’s judges would identify norms ascertained through their own methodology as authoritative. Maybe there is something to that. But the federal judges themselves would not need to make use of the state’s interpretive methodology as a means of predicting how the state would resolve the issue (although they certainly might do so).⁷³ In any event, federal judges often use alternative means and consult various forms of data thought to be probative of the state’s law, including extra-legal materials such as how similar states have decided the same question.⁷⁴ This is very different from how the state’s own judges would go about interpreting their law.

Prediction (like certification, where the federal court asks the state’s apex court to answer a question concerning the state’s law) is justifiable under the assumption that the state courts will get the question right, meaning that they will apply the relevant norms that are legally valid according to the state’s rule of recognition. On that assumption, if the federal court is able to accurately predict the norms that the state’s courts will identify, then its decision will satisfy the state’s criteria of validity too, whatever they may be, and the federal court need not have them in mind. Either prediction does not really constitute legal interpretation at all or, if it does, it is a kind of interpretation that differs—often considerably—from the approach that the state’s judges would take to interpreting their own law.

I want to explore one further possibility. Could a system’s judges treat a norm as legally valid because or on the ground that it was identified through particular epistemic means? For example, imagine that a state’s judges embraced an extreme form of public meaning textualism: they would take as legally valid whatever norms the public understands a statute (based on the text alone) to establish, where that understanding is determined by a public opinion poll. Whatever norm a majority of respondents takes the statute to establish would meet the system’s criteria of validity in virtue of the process that generated it. A rule would be legally valid, then, just because it was ascertained via a particular interpretive methodology. Here the content of the rule of recognition would seem to be constituted in part by the epistemic means that the judges employ to identify their law. Note that in this hypothetical it is not public meaning itself that determines legal validity, but rather public meaning as ascertained through a particular polling method. So judges could not possibly identify legally valid norms by using some other means of ascertaining public meaning. If this were possible of a legal system, it would seem that any court—even an external one—would have to use the polling method to accurately identify the system’s law.

I do not think this is possible, however. First, here the system would have no statutory law that pre-existed the judicial act of interpretation, which conflicts with the commonly held understanding that legislatures create statutory law.⁷⁵ I think this would be a problem with any interpretive methodology playing a validating role. Second, some conceivable interpretive approaches could not possibly play a constitutive role in determining a system’s law, given the necessary characteristics of a legal system—in particular the general efficacy (and so certainty), the generality, and the standing nature of legal norms.

For a legal system to exist, ordinary people must be able to identify and follow norms that satisfy the system’s rule of recognition, and must in fact identify and follow them (not in every case, but generally).⁷⁶ Otherwise, people would not really be governed by the law. Such governance is necessary to the existence of a legal system.⁷⁷ Further, legal norms must be general in a double sense: applicable to classes of conduct and to classes of people.⁷⁸ And legal norms must have a “standing” character, in contrast to orders or wishes, which may be one-off.⁷⁹ There is no guarantee that laws ascertained through the polling method would meet any of these conditions.⁸⁰

There are other possible approaches that judges could take that would even more surely fail to meet the existence conditions. For example, suppose judges took the wishes of their supreme leader as gleaned through signs of nature to constitute the law. That process would not generate either general norms or standing ones. And people could not be expected to effectively identify and follow norms generated through that approach. Even if judges were to treat the wishes of their leader as grasped through signs of nature as constitutive of the law, then, there would be no law thus constituted.⁸¹ Because such norms could not constitute the law of a genuine legal system, we would not even get to the question of whether external judges seeking to apply the system’s law would have to use the methodology. There would be no legal system’s law to seek.⁸²

One might respond that epistemic norms nevertheless could constitute criteria of validity in a system, along the following lines. First, some conceivable interpretive methodologies would generate norms that satisfy the necessary conditions for a legal system (and perhaps even the polling method would do so). Second, judges could treat their interpretive methodology as part of their rule of recognition without that entailing that a legal norm would not exist unless and until the method was used to identify it. The valid law of the system would be constituted by the norms that would be ascertained through the method as employed by the system’s judges. And such norms would pre-exist the identification process. But in that case, I see no reason why an interpreter could not come up with the valid norms without going through the method. They could use some other means—in the polling method hypothetical, for example, external judges might use corpus linguistics instead of the survey to identify the applicable norms. If that is right, then the method itself would not truly be constitutive of the system’s law. Instead, something like common public meaning would be constitutive, and that meaning could conceivably be ascertained through multiple different methods.

As a conceptual matter, then, no matter how a system’s judges treat their own interpretive methodology, I do not think external interpreters would necessarily have to follow it. As a practical matter, it may nevertheless be advantageous or even necessary for external interpreters to take up the target system’s interpretive methodology; the extent to which that is so will depend on descriptive facts that vary across cases.

V. Conclusion

I have tried to explain how a federal court might properly apply a state’s substantive law and fully discharge its conflict of laws duties without employing the state’s interpretive methods, even if the state’s judges follow a common methodology. My analysis could explain why, in practice, federal courts charged with applying state law often do not defer to state interpretive methods (although there are certainly other possible explanations—for example, that federal courts are simply not aware of state methods).⁸³

I have focused on the federal–state intersystemic context and the federal court’s duty to apply substantive state law. I have argued that this duty, which is widely understood to be a fundamental federal constitutional duty, does not come with any kind of blanket duty to apply the interpretive methodology of the state.

But the duty of federal courts to apply state substantive law, although central, is only one concern of federal courts and does not exhaust the concerns to which they may reasonably respond when adjudicating cases arising under state law.⁸⁴ There may be reasons that weigh in favor of a federal court following the interpretive methodology of state judges even if doing so is not necessary to accurately identifying the substantive law.⁸⁵ One is that doing so might have a constraining and disciplining effect, making it less likely that federal judges impose their own policy preferences in the process of applying state law. Another reason might be that the federal court wishes to express respect for the state’s own decision-making procedures and a commitment to federalism. Still another might be the possible “unseemliness” of visible differences in state and federal approaches to adjudicating the very same legal questions.⁸⁶ On the other hand, there may be reasons for the federal court to instead use the interpretive methods with which it is most familiar and comfortable, even if it would be more likely to get the substantive law right by using the state methods. After all, the administrative burdens of keeping up with the interpretive methodologies of the various states might be considerable, and the cost of error when a federal court misinterprets a state’s law might not be as great as when one of the state’s own courts makes such an error, since the federal court’s decision will not be determinative of state law going forward.