I. Introduction

In this “age of statutes,”[1] the canons of statutory interpretation are the common language spoken by legal interpreters. Judges use the language of the canons to explain their reasonings.[2] Advocates rely on the canons to articulate legal arguments. States codify them.[3] Legislative drafting manuals recite them.[4] And for their part, legal scholars endlessly debate them. Scholars not only assiduously study canon use by courts,[5] legislative staff,[6] and agency personnel,[7] they also propose new canons,[8] monitor their vitality,[9] and predict their demise.[10] Canon use among judges and scholars is pervasive in part because it is cross-ideological: although textualists are most well-known for their vociferous support for the canons,[11] purposivists[12] and pragmatists[13] have also taken great care to find a place for the canons in their theories.[14]

Despite the careful attention that legal interpreters have paid to the canons, they have given surprisingly little consideration to a most basic question: How do we identify a canon of statutory interpretation? Although there is tacit agreement about the canonical status of many well-known interpretive principles,[15] there is no agreement about scores—perhaps hundreds—of others. And not only is there no accepted list of canons, there is no accepted test for determining whether any particular interpretive principle is a canon.

The absence of a list of canons, or even a test for identifying them, has created both theoretical and doctrinal confusion. Without a way to distinguish between canons and noncanonical interpretive principles, legal interpreters often resort to intuition or unstated assumptions when debating a canon’s existence.[16] Other scholars conflate the existence of a canon with its merits.[17] The confused terminology surrounding the canons bears this problem out: scholars and courts alternatively suggest that canons without merit are invalid,[18] unreliable,[19] false,[20] faux,[21] or illegitimate.[22]

Moreover, without a way to identify the canons, it is impossible to evaluate judicial opinions that turn on their application. This much was made clear by the Supreme Court’s landmark decision, Bostock v. Clayton County.[23] In this Title VII case, the Supreme Court considered whether the statute’s prohibition on “sex” discrimination included a prohibition on “sexual orientation” discrimination.[24] The employers argued that because Title VII uses the term sex rather than sexual orientation, the statute implicitly excluded sexual orientation from its reach.[25] The Court rejected the employers’ argument, holding that it rested on the erroneous premise that a statute implicitly excludes a specific example of a general category (sexual orientation discrimination) by explicitly prohibiting a general category that includes that example (sex discrimination).[26] Justice Gorsuch, in his opinion for the Court, characterized this erroneous premise as a “canon of donut holes” and summarily declared that there is no such thing.[27] In other words, the Court rejected the employers’ argument because the argument rested on an interpretive principle that the Court does not recognize as a canon of interpretation.

The outcome of Bostock, one of the most watched cases in recent memory, turned on whether the Court accepted the interpretive methodology employed by a litigant as a canon.[28] But, was Justice Gorsuch correct to reject the “canon of donut holes”? The short and unsatisfying answer is that we don’t know—and we can’t know—unless we have some way to distinguish between a canon of statutory interpretation and other, noncanonical principles of interpretation used by judges, scholars, and advocates.

In this Article, I hope to remedy the scholarly and doctrinal confusion surrounding canon identification. To that end, I propose criteria for canon spotting—that is—for determining whether an interpretive principle should be considered a canon of statutory interpretation. This Article makes two contributions to the literature on the canons. First, it helps scholars evaluate whether new interpretive principles are canons. In just the last few years, scholars have introduced a raft of new canons, including Professor Gluck’s “CBO” canon,[29] Professor Stack’s “enacted purposes” canon,[30] Professor Heinzerling’s “power” canons,[31] Professor Krishnakumar’s “unique national institution” canon,[32] and Professors Leib and Brudney’s “belt-and-suspenders” canon,[33] among many others.[34] Each of these interpretive principles has some appeal. But, whether any of them can be considered a canon is another matter. There is no stable list of accepted canons or accepted standards for determining whether a principle is a canon. By proposing and defending three necessary criteria for canonicity, this Article will allow scholars to assess the canonical status of particular interpretive principles.

Second, this Article will help scholars evaluate the canons normatively. There is a robust scholarly debate about whether and how particular canons should be used.[35] But, when scholars and other legal interpreters, including judges, discuss the merits of canons, they often conflate the existence of a canon with its merits. There is no agreement about whether an interpretive principle lacking in merit is invalid,[36] unreliable,[37] false,[38] faux,[39] illegitimate,[40] or something else altogether. The failure to distinguish between the existence and merits of a canon impedes productive discussion about whether a particular interpretive principle should be adopted or abandoned. This Article will help scholars distinguish the normative from the descriptive: not every interpretive principle is a canon; and not every canon is normatively attractive. By disentangling the existence of a canon from its merits, this Article enables scholars and courts to discuss the normative impact of the canons with precision and clarity.

Now is the right time for scholars to seek greater clarity in canon identification. Interest in canons has never been stronger nor opinions about the canons more varied. Even a cursory look at current canon scholarship reveals a breadth of perspectives and methodologies. Some of this research seeks to learn how congressional staff[41] or agencies[42] use canons in their work, challenging long-standing assumptions underlying canon use. Other research tracks the way that courts themselves use the canons,[43] revealing surprising trends in both state and federal courts. Still other research categorizes,[44] justifies,[45] or critiques[46] the canons, or suggests the adoption of new canons.[47] All of these projects, and others, can be improved by greater precision in canon identification. Moreover, as important as canon use is today, it is poised to become even more important in the future. The group of theories called textualism, although varied,[48] share a commitment to canon use.[49] And with the addition of Justice Barrett, the Supreme Court is now dominated by Justices with strong stated commitments to textualism.[50] Although these Justices may differ in their approaches to statutory interpretation, it is safe to anticipate that the canons will play a prominent role in their interpretations for a generation.

In Part II, I demonstrate why canon identification is so important. The work of scholars, judges, and advocates would benefit from greater agreement about how to identify canons. Nevertheless, the project of identifying canons remains unfinished. Indeed, it has hardly even begun: there is no accepted list of canons nor accepted criteria for identifying them.

In Part III, I propose, explain, and defend three necessary criteria for recognizing whether an interpretive principle is a canon. In the formulation I propose, an interpretive principle should be considered a canon if and only if: it reflects use by legal interpreters; it affects an interpretive outcome when applied; and it is supported by a claim of theoretical justification. After proposing these criteria, I explain how they differ from the criteria used, explicitly or implicitly, by other scholars.

II. The Need for Canon Spotting Criteria

When interpreting statutes, legal interpreters rely on interpretive maxims or principles to help them apply statutory language to specific factual situations. These interpretive principles go by different names. Sometimes they are simply referred to as principles of interpretation; other times they are called maxims, tools, rules, presumptions, or doctrines. Sometimes a court employs a principle of interpretation without identifying or labeling it at all. And sometimes courts and scholars use more than one of these labels to describe a particular principle.[51]

Nevertheless, not all interpretive principles are equal.[52] Some, through a combination of their pervasiveness,[53] antiquity,[54] intuitive appeal,[55] or other factors,[56] have taken on extraordinary status. These extraordinary principles of statutory interpretation are often called the canons.[57] While all canons are interpretive principles, then, not every interpretive principle can be considered a canon.[58]

Because of the special place of the canons in the interpretive process, canon spotting—that is, distinguishing canons from run‑of‑the‑mill interpretive principles—helps clarify debates about statutory interpretation. Nevertheless, identifying whether a particular interpretive principle is a canon has proved difficult. As a result, the task of identifying canons remains unfinished. This Part will first demonstrate the value of distinguishing between canons of interpretation and, on the other hand, noncanonical interpretive principles. Next, it will show that the canons are not only important now but likely will increase in importance in the coming years. Finally, it will explain why the task of identifying canons remains unfinished.

A. The Importance of Canon Spotting

Legal interpreters assign various names to the interpretive maxims or principles that they use to interpret statutory language. At times, they call them, simply enough, “principles.”[59] Other times, they are called “maxims,”[60] “tools,”[61] “rules,”[62] “presumptions,”[63] or “doctrines.”[64] Sometimes interpreters use these labels interchangeably.[65] And often, courts employ principles of interpretation without identifying or labeling them at all.[66] But, despite this multiplicity of designations, not all interpretive principles are considered equivalent by legal interpreters. When they want to highlight an interpretive principle’s persuasive power, legal interpreters, including judges, advocates, and scholars, refer to the principle as a canon of interpretation.

Both practical and theoretical questions hinge on whether a particular principle of interpretation is a canon.[67] First, designating an interpretive principle a canon carries great rhetorical force. Second, scholarly debates about the canons can be clarified with more agreement about which principles are canonical. Third, whether an interpretive principle is considered a canon influences how advocates argue cases and how judges decide them. Fourth, the canons are especially (but not exclusively) valued by textualist theories, which are poised to play an increasingly important role in legal interpretation in the coming decades.

1. The Rhetorical Importance of Canon Spotting.

The word “canon” itself carries great rhetorical force. As a result, even leaving aside the strength of any particular argument, a legal interpreter strengthens it, rhetorically, by declaring that the argument rests on a canon.

The word canon connotes authority and permanence. A canonical text is one whose authority derives, ultimately, from its ancient origin.[68] Canonical texts serve as models against which modern creations are measured—with an underlying premise that the ancient standard is pure while the innovation is liable to be debased.[69] Not surprisingly, then, for centuries, canon has meant “a set of non-negotiable norms with unquestioned legitimacy.”[70] The Supreme Court’s earliest uses of the term canon to describe a principle of interpretation imply as much by linking the canons with what it considered an indisputably reliable source. As early as 1870, the Court justified an interpretation because it relied on “the canons of construction laid down by Chief Justice Marshall.”[71]

A commitment to preserving authority derived from an ancient tradition means that canonical texts are sometimes hard for moderns to decipher. But, far from being a liability, the authority of a canonical text is only bolstered by its use of “archaic or formal language that seems to certify an ancient origin.”[72] Naturally, then, the keepers of the canon must be “scribes, grammarians, teachers, philosophers”[73] or others with special esoteric knowledge about its meaning.[74] This description fits the way many legal interpreters relate to and discuss the canons of interpretation. Consider the confidence Justice Scalia reposed in the so-called Latin canons. He was simply incredulous that anyone could even criticize them,[75] let alone deny their authority. To the keepers of the canons of interpretation, the esoteric nature of the Latin canons only proves their legitimacy.

Finally, it should not be forgotten that the term canon, for many centuries, has carried strong religious connotations.[76] At some level, then, the suggestion that a text is canonical gestures toward the authority of otherworldly inspiration.[77] Like Justice Scalia, who found it “hard to believe anyone could criticize” the Latin canons,[78] other judges also describe the canons with almost religious reverence, referring to canons as “venerable”[79] or “cardinal”[80] principles, terms with deep religious connotations. Indeed, reflecting the religious sense of the term canon, some legal interpreters have placed “almost unquestioning reliance” on the canons of interpretation.[81]

2. The Analytical Importance of Canon Spotting.

Establishing criteria for canon spotting should provide much‑needed clarity to scholars who discuss and debate the canons. First, without a way to identify canons, there is no way to evaluate assertions that an interpretive principle even is a canon. There have been countless scholarly books and articles identifying new canons,[82] most of which are based on extant uses of a particular interpretive principle.[83] But, without a definitive list of canons or accepted criteria for identifying them, it is impossible to evaluate these scholarly claims.

Second, and equally important, there is no way to evaluate claims that an interpretive principle is not a canon. Consider again the Supreme Court’s landmark Bostock opinion, in which Justice Gorsuch rejected the argument that Congress’s silence implies an unstated exception to a generally applicable rule.[84] He called such an argument a “canon of donut holes” and denied that there is “any such thing.”[85] Was Gorsuch correct to reject the existence of a canon of donut holes? He is surely correct that there is no canon called the canon of donut holes. But, Gorsuch probably is not correct if he meant that no canon directs courts to infer, from Congress’s silence, an exception to a generally applicable rule. Indeed, that is a perfectly reasonable way to describe many clear-statement canons, like the presumption against extraterritoriality[86] or the presumption that a generally applicable statute does not apply to the sovereign.[87] Both of these canons suggest an interpretation that excludes a particular, unstated example from a general category simply because it is unstated.

Now, the canon of donut holes rejected by Justice Gorsuch is not identical to these two clear-statement rules, of course. So, it is possible, despite the existence of similar rules, that Justice Gorsuch was correct to reject the application of what he termed a canon of donut holes to Title VII. But, the important point is that we just can’t know whether Justice Gorsuch was correct to reject the canon: absent criteria for identifying canons, there is no way to verify or falsify his assertion that there is no such thing as a canon of donut holes—and his confident statement that it does not exist becomes no more than a bald assertion. When resting an assessment of a canon’s existence on a bare assertion, Justice Gorsuch was not alone. Consider, for example, Justice Scalia’s cross-temporal retort to Professor Llewellyn about the existence of a canon directing courts to inquire into a statute’s real purpose. In response to Llewellyn’s identification of the canon,[88] Justice Scalia glibly responded: “Never heard of it.”[89]

Third, not only does the inability to identify the canons impede the evaluation of their existence, it makes it difficult to evaluate them normatively. In the absence of standards for identifying canons, it is not always clear whether scholars evaluating the canons are making normative arguments (e.g., a bad interpretive principle should not be followed) or descriptive ones (e.g., a bad interpretive principle is not a canon).[90] Professor Dickerson, for example, suggested that some canons are of “doubtful validity.”[91] Conversely, other scholars, like Professors Eskridge, Baude, and Sachs, make reference to “valid canons.”[92] Courts, too, sometimes refer to particular canons as “valid.”[93] But, it is not clear from discussions of valid and invalid canons whether all canons are valid per se or whether they cease to be canons if they are invalid. If an interpretive principle is not a valid canon, then is it an “invalid canon,” an interpretive principle that is not a canon, or something else altogether?

Different formulations present similar confusions. Professors Baude and Sachs assert that canons can be either “real or false.”[94] In scholarly works, Justice Scalia labeled canons with which he disagreed “faux”[95] or “falsities,”[96] and Judge Easterbrook expressed the view that a canon could be “wholly illegitimate.”[97] In opinions, too, other judges have opined on the legitimacy of particular canons.[98] But again, the consequence of calling a canon faux or illegitimate is unclear; are they faux or illegitimate because they are not canons, or are they merely canons that lead to bad results? We just don’t know.

Perhaps most vexingly, scholars and courts sometimes suggest that there is a hierarchy of legitimacy among the canons. Some canons are deemed “more legitimate”[99] or “more democratically legitimate”[100] than others, some “especially . . . legitimate,”[101] and some “less helpful”[102] than others. Some are called “more reliable”[103] or “less reliable”[104] than others, while others are deemed “highly unreliable.”[105] And some canons are deemed “of equal worth”[106] to others. Again, it is not clear from the way these interpreters hierarchize the canons what makes a canon more or less legitimate, helpful, or reliable. And it is not clear what the hierarchy is, whether interpretive principles very low in the hierarchy are still canons, or whether the hierarchy changes with different applications of the canons. In short, without the ability to identify when an interpretive principle is a canon, debates about their normative appeal are apt to be mired in confusion.

3. The Doctrinal Importance of Canon Spotting.

Distinguishing canons from noncanonical interpretive principles is important because courts give more weight to canons than to noncanonical interpretive principles. This point is distinct from the stronger claim that courts actually feel bound by interpretive methodology. Certainly, there are those who argue persuasively that the canons are mere window dressing—that is, that they are used by courts to justify decisions but do not actually affect case outcomes.[107] Others reply that, at least sometimes, courts actually follow—rather than merely cite opportunistically—interpretive principles.[108] I do not try to resolve this disagreement here. Rather, assuming that courts sometimes do genuinely rely on interpretive principles when interpreting statutes, here, I make the more limited claim that courts citing interpretive principles distinguish between canons of interpretation and noncanonical principles of interpretation, relying more heavily on interpretive principles that they consider canons.

Consider again Bostock. There, the Court was faced with the question whether Title VII’s prohibition on discrimination because of “sex” included a prohibition on discrimination because of “sexual orientation.” The employers argued that by including the term sex, but not sexual orientation, in the statute, Title VII implicitly excluded the prohibition of discrimination based on sexual orientation from its reach. The Court rejected that argument, holding that discrimination based on sexual orientation necessarily entails discrimination based on sex. Explaining why it rejected the employers’ argument, the Court noted that it rested on an interpretive principle that the Court did not consider a canon. The Court explained that there is no “such thing as a ‘canon of donut holes,’ in which Congress’s failure to speak directly to a specific case that falls within a more general statutory rule creates a tacit exception.”[109] In other words, the Court rejected the employers’ argument, at least in part, because it was not framed in terms of a canon that the Court recognized. If we take the Court at its word, then the absence of an argument based on a canon influenced the outcome of the case. Counterfactually, if the employers were able to articulate their argument in terms of a canon recognized by the Court,[110] then the Court would not have dismissed it so lightly.

Bostock is not unique; Justices of the Supreme Court also paid close attention to whether an argument was based on a canon—as opposed to a noncanonical interpretive principle—in Yates v. United States.[111] In Yates, the Court considered whether a fish was a “tangible object” for the purposes of a statute that prohibited a person from destroying, concealing, or covering up a record or tangible object to impede a federal investigation.[112] Justice Ginsburg, for a plurality, interpreted the term “tangible object” in light of its position within the U.S. Code.[113] Specifically, she noted that Congress grouped the tangible‑object language “together with specialized provisions expressly aimed at corporate fraud and financial audits.”[114] She concluded that this placement suggests a specialized reading of tangible object, one more narrow than the literal, broader meaning of this language.[115] In dissent, Justice Kagan criticized this move. She opined that according meaning to the section based on its position in the Code was unjustified, in part, because Justice Ginsburg’s interpretive principle was not mentioned in a particular list of canons.[116] Specifically, she noted that “the plurality’s new number-in-the-Code theory” was unattested in a list of canons compiled by Justice Scalia and Bryan Garner.[117] In other words, Justice Kagan discounted the force of Justice Ginsburg’s interpretive principle—not just because she found it unpersuasive—but, crucially, because she did not consider the principle a canon. Similarly, judges routinely reject arguments that rely on an interpretive principle when they do not consider the principle to be a canon. These rejected canons include a “canon against verbosity,”[118] a canon of “administrative convenience,”[119] as well as others.[120]

This conclusion—that courts are more likely to rely on interpretative principles that they consider canons (as opposed to noncanonical interpretive principles)—suggests a few ways that improved canon spotting will affect statutory interpretation doctrine. First, having a test for spotting canons will force opposing parties, or disputing judges, to join issue on interpretive methodology, making it less likely that they will argue past one another. Without a test for spotting canons, parties and judges are apt to make bald assertions about the existence of a canon and then reach legal conclusions based on their assertions. A test for spotting canons will allow disputing sides to demonstrate that the interpretive principle they use is a canon; or alternatively, to demonstrate that an opponent’s interpretive principle is not a canon. Although this extra analytical step will not end all disagreements about canonicity, it will force interpretive disagreements into the open, allowing opposing parties to make more incisive and productive arguments and counterarguments.[121]

Second, more reliable canon spotting could improve statutory interpretation doctrine by improving advocacy.[122] A test for spotting canons, although it will not eliminate all disagreements about the existence of canons, will make the identification of canons more certain in some cases. When canons can be identified more easily, advocates will have less of an incentive to expend resources debating their existence. As a result, advocates’ arguments will be less cluttered because they will be able to make better informed decisions about how to focus their arguments, including knowing which interpretive principles to press and which to leave on the cutting-room floor.

Third, clarifying which interpretive principles are canons could improve statutory interpretation doctrine by improving the Court–Congress dialogue. Statutory drafters are familiar with some of the interpretive principles that courts use to interpret statutes.[123] If statutory drafters can determine whether a particular interpretive principle is canonical, they should have a better idea about how to draft statutes to elicit interpretations in line with their goals.[124]

B. The Increasing Importance of Canon Spotting

The project of identifying canons is important for the rhetorical, analytical, and doctrinal reasons described above. Together, these reasons present a snapshot of how the canons are currently discussed, analyzed, and used. But, in order to capture fully the importance of identifying the canons, it is necessary to consider not only their present importance but also the increasingly important role they are likely to play in the future. Textualists are the most enthusiastic users of many of the canons. And textualism is likely to become increasingly important to statutory interpretation. Combined, these factors suggest that however important the canons are now, the rhetorical, analytical, and doctrinal importance of the canons is likely to increase in the decades to come.

1. The Textualist Turn to Extratextual Sources of Meaning.

Legal interpreters of all stripes refer to, discuss, and rely on the canons.[125] Nevertheless, the modern turn toward the canons is driven in large part by textualists’ stated goal of bringing rule-like objectivity to statutory interpretation.[126] Although there are different strands of textualism,[127] textualist interpreters are largely unified in their criticism of purposivism for allowing interpreters to implement their own preferences.[128] In light of scholarship demonstrating that textualists, too, are capable of importing their own preferences into their interpretations,[129] I do not find this critique, standing alone, particularly compelling. Nevertheless, this critique can have real repercussions for the way that textualists interpret statutory language. Consider the good‑faith textualist who is interested in restraining her biases by limiting or excluding sources that confer too much discretion. Such an interpreter creates a dilemma for herself: by excluding evidence of statutory purpose (most saliently, legislative history), the textualist interpreter greatly limits available sources for interpreting statutes.[130] This dilemma is most pronounced when characterized by the phenomenon that Professor Nourse calls “petty textualism.”[131] As Nourse has described, petty textualism is the practice of pulling “one or two or five words out of a lengthy statute” and ignoring the rest of the statute.[132] A textualist who will not rely on evidence of legislative purpose from legislative history, or evidence of a legislative plan from a statutory scheme, has fewer sources of meaning to use when interpreting statutes than her purposivist or pragmatist counterparts.

Faced with this loss of evidence of statutory purpose or plan, textualists frequently turn to other extratextual evidence to interpret statutes, like their own intuition about the “ordinary meaning” of the statutory language,[133] dictionaries,[134] nonlegal texts (like novels[135] and newspaper articles[136]) and, increasingly, large bodies of texts designed to reveal how language is typically used.[137]

Their rejection of evidence of legislative purpose also explains textualists’ enthusiastic turn to canon use.[138] Canons are attractive to textualists as replacements for the loss of meaning attributable to their rejection of legislative purpose both because the canons are powerful determinants of meaning and also because they can be used without explicitly requiring the interpreter to consider the purpose of the statutory text.[139] Canons are powerful determinants of meaning because they are generally applicable to a large number of statutory texts,[140] and because, at least for substantive canons, they allow interpreters to import amorphous extratextual values into their interpretations.[141] But, although they are powerful inputs into the interpretive process, most canons can be applied without explicit reference to statutory purpose. Most canons apply across different statutory schemes[142] and are written at a level of generality that allows an interpreter to apply them without referring to the underlying purpose of the statute.[143] Moreover, the linguistic canons are usually described as value-neutral,[144] suggesting that their application is objective.[145] For these reasons, the canons have been, and likely will continue to be, indispensable to interpreters who embrace textualism and find themselves searching for statutory meaning in the absence of evidence of legislative purpose.

2. The Continuing Rise of Textualism.

That canons are essential to textualism as practiced is only part of the story. It is the fact that textualism is poised to increase in importance, perhaps dramatically, that will make canon identification increasingly important in the future.[146] Although the Supreme Court has not embraced a full-throated textualism that rejects extrinsic evidence of statutory meaning,[147] nor applied the canons in a way that leads to predictable results,[148] the Court continues to move from purpose-positive to purpose-skeptical, thus making the canons relatively more important to the interpretive process. Gone are the days when the Court might remark, as it famously did in Overton Park, that because legislative history is ambiguous “it is clear that we must look primarily to the statutes themselves to find the legislative intent.”[149] Instead, under the influence of Justice Scalia’s long-held conviction that legislative history is illegitimate,[150] the modern Court is increasingly likely to shy from relying overtly[151] on legislative intent or statutory purpose, euphemistically referring instead to a “legislative plan”[152] or “Congress’s expressed policy.”[153] Indeed, the Court occasionally almost apologizes for mentioning legislative history, gently reminding readers that there are “[t]hose who deem legislative history a useful interpretive tool.”[154]

Most importantly, the increasing influence of textualism on the Court is heralded by the recent shift in Court personnel. Although the Court lost its most longstanding critic of legislative history with the passing of Justice Scalia, since 2017, the Court has added three Justices with strongly stated commitments to textualism. Take Justice Kavanaugh’s recent dissenting opinion in Bostock: there, Justice Kavanaugh paid homage to Justice Scalia and to textualism by liberally quoting Justice Scalia’s aphorisms on textualism.[155] Justice Kavanaugh assured the reader that his interpretation truly was the textualist one, implying that he would be carrying on in the Scalian tradition in the late Justice’s absence.[156] For his part, Justice Gorsuch has also made it clear that he views himself as part of the textualist tradition. In his book, A Republic if You Can Keep It, Justice Gorsuch dedicates a chapter to textualism, using a personal anecdote to deride the use of both legislative history and purposivism more generally.[157] He concludes by asserting that textualism, but not purposivism, is a legitimate form of interpretation and that the canons are “neutral[] and objective interpretive tools.”[158] Justice Barrett has well-developed public opinions on textualism and the canons. Although she notes that textualists avidly use both linguistic and substantive canons, she argues that some substantive canons supplant a more natural reading of the text.[159] The extent of her use of the canons on the Court remains to be seen, but early indications suggest that she will be a reliable user of, at least, the linguistic canons.[160]

In sum, because textualism is central to the identity of the Court’s newest members, and because the canons are central to the practice and rhetoric of textualism, the canons are poised to become even more important to statutory interpretation in the years to come. Identifying them, therefore, is likely to be an even more important task in the future than it is today.

C. The Unfinished Task of Canon Spotting

Legal interpreters, including scholars, judges, and advocates, would benefit from being able to identify whether a particular interpretive principle is a canon. And there is, in fact, widespread agreement about the canonical status of some well-established interpretive principles. Few, if any, legal interpreters would gainsay the canonicity of expressio unius[161] or the rule of lenity.[162] Nevertheless, determining whether less well-established principles of interpretation are canons, including the raft of new putative canons regularly introduced by scholars, turns out to be more complicated. There is no agreed-upon list of canons—the corpus of interpretive rules considered canonical changes over time—and there are no generally accepted criteria for determining whether a principle is a canon. As a result, there is no established way to determine whether a particular interpretive principle should be classified as a canon.

First, as Professor Eskridge has noted, there is “no canonical collection of valid canons.”[163] Professor Eskridge’s list of canons, which numbers them in the hundreds,[164] is considerably longer than the parsimonious list of fifty-seven canons assembled by Justice Scalia and Professor Bryan Garner in their competing treatise.[165] The collection in Sutherland’s Statutes and Statutory Construction falls somewhere between these two poles.[166] As a result, legal interpreters searching for a definitive resolution to a dispute about the existence of a canon will not be able to settle it by recourse to a universally accepted list.[167]

Second, uncertainty over the existence of a particular canon is exacerbated by the fact that the corpus of possible canons is unstable over time. In just the last few years, scholars have introduced a host of new interpretive principles, including: Professor Gluck’s “CBO” canon,[168] Professor Stack’s “enacted purposes” canon,[169] Professor Heinzerling’s “power” canons,[170] Professor Krishnakumar’s “unique national institution” canon,[171] Professors Leib and Brudney’s “belt-and-suspenders” canon,[172] and many more.[173]

Just as new canons are born, others pass away: Professor Bruhl, tracking canon use in the lower courts, found that the Supreme Court sometimes abandons canons, leading lower courts to follow suit.[174] Other once-vital canons are not yet dead but show signs of decline. Professor Mendelson, reviewing canon use, concluded that some “traditionally recognized canons . . . are used rarely and may soon evaporate altogether.”[175] Bruhl, similarly, observes that canons once in use can be retired or demoted.[176] And adding to the confusion, courts sometimes use an existing canon in a novel way,[177] obscuring whether and when new canons have been created or retired.

Third, there are no generally accepted criteria for spotting canons. The absence of a stable list of canons would be less consequential if there was an accepted standard or accepted criteria for determining when an interpretive principle is a canon. Nevertheless, most scholars discussing the canons, including those introducing new canons, fail to articulate standards for identifying them. A few scholars have implicitly identified criteria without elaborating on them or explicitly justifying them.[178] Still other commentators eschew criteria altogether, implying instead that the canons simply reflect common sense and need no other justification.[179] And others blur the line between the descriptive and the normative, proposing aspirational interpretive principles as canons without making any claims about their current canonical state.[180]

Notable exceptions include Professors Krishnakumar and Nourse,[181] and (separately) Professor Sinclair, who have more explicitly identified and explained their preferred criteria for determining when an interpretive principle is a canon. Krishnakumar and Nourse evaluate possible criteria, including frequency of use, historical pedigree, theoretical justification, and the explicitness of the adoption.[182] They conclude that “the basic thread connecting the canons is (or should be) established convention” as measured by “regular Supreme Court use” by Justices “across ideological divides.”[183] Professor Sinclair adds that the formulation of a canon must be relatively stable. Even if “statute-like determinacy” is too demanding, he suggests that “some stability or limit on variety in expression, some canonical form, is necessary.”[184] Moreover, he adds, an interpretive principle receives canonical status, in part, “from the robustness of the reasoning associated with it.”[185]

In sum, there is no settled list or stable body of canons. And although a few scholars have suggested criteria for evaluating whether an interpretive principle is a canon, these criteria have not been widely adopted. As a result, the task of canon spotting, important for rhetorical, theoretical, and doctrinal purposes, is unfinished. In Part III, I take up the task of proposing and defending criteria for the identification of canons.

III. Canon Spotting Criteria

Despite the rhetorical, theoretical, and doctrinal benefits that would flow from greater certainty in canon identification, there is no settled, stable list of canons and no accepted criteria for identifying them. This Part takes up the unfinished task of canon spotting by proposing three criteria for the identification of canons of interpretation. An interpretive principle or maxim should be considered a canon if and only if it: (1) reflects actual use by legal interpreters; (2) affects an interpretive outcome when it is used; and (3) is supported by a claim of theoretical justification.

These criteria are meant to distinguish between principles of interpretation that are canons and other, noncanonical principles of interpretation. That is, these criteria are meant to be descriptive rather than normative. I do not suggest that every interpretive principle that meets these proposed criteria is a good principle that courts should adopt; nor does a principle that meets these criteria cease to be a canon simply because it leads to bad results.

An interpretive principle should be considered a canon of interpretation only if it reflects use by legal interpreters vested with authority to interpret statutes. This criterion, in turn, can be divided into two separate questions: who are “legal interpreters,” and what does it mean for them to “use” a principle of interpretation?

On the first question, I would include the Supreme Court in the category of legal interpreters. Here, I agree with scholars like Professor Eskridge, who would consider use by the Supreme Court sufficient to establish canonicity.[186] As Professors Krishnakumar and Nourse argued, it makes sense to pay close attention to usage by the Supreme Court because lower courts and advocates nationwide naturally are aware of the way that the Supreme Court adopts or rejects principles of interpretation.[187]

Also, like Professor Eskridge[188] and other scholars,[189] I would include institutions other than the Supreme Court in my definition of legal interpreters. Specifically, I would include lower federal courts, state courts, and administrative agencies as legal interpreters for the purpose of canon spotting. On this score, my test is broader than the test proposed by Professors Krishnakumar and Nourse, who would attribute canonical status to principles used by the Supreme Court but not by these other institutions.[190] Specifically, they argue that use by the Supreme Court creates trends in a way that lower court usage does not; moreover, they argue that state court use can reflect state idiosyncrasies.[191]

The Supreme Court undoubtedly has the unique ability to influence statutory interpretation trends. Nevertheless, lower federal courts, state court, and administrative agencies also should be considered legal interpreters for canon identification. First, lower federal courts, state courts, and administrative agencies all adjudicate large numbers of statutory interpretation disputes every year and, indeed, orders of magnitude more than the Supreme Court. Even a rough comparison of the number of cases they decide demonstrates how much information about the use of interpretive principles is lost by excluding lower federal court, state court, and administrative agency data. For the past decade, the Supreme Court has disposed of an average of about seventy cases per term with written opinions, not all of which have involved statutory interpretation.[192] By contrast, lower federal courts resolve hundreds of thousands of disputes each year, a large proportion of which include statutory claims.[193] Federal administrative agencies[194] and state courts,[195] too, dispose of countless disputes each year, many of which require the resolution of statutory meaning.

Second, because the Supreme Court is not representative of interpretive principle use throughout the country, omitting lower federal court, agency, and state court data skews the picture of interpretive principle use. With different institutional constraints, lower courts are apt to take the canons more seriously than the Supreme Court. Compared with disputes that reach the Supreme Court, the “ideological stakes” are lower in many district court and court of appeals opinions.[196] As a result, lower courts have less of an incentive to use canons opportunistically to cover for motivated reasoning. Instead, lower courts are more likely to cite canons in a genuine attempt to follow them. Accordingly, as Professor Bruhl has discovered, the Supreme Court and lower courts display different patterns of canon use.[197] As a result, excluding lower court data paints an incomplete picture of how the canons are used in practice.[198] Similarly, both state courts[199] and administrative agencies[200] also have interpretive practices that differ from those of the Supreme Court. As a result, excluding data from these institutions also skews the picture of how different interpretive principles are used.

Third, Krishnakumar and Nourse are surely correct that the Supreme Court is uniquely positioned to influence interpretive methodology in courts nationwide;[201] however, it would be wrong to conclude that state interpretive methodologies are idiosyncratic and therefore of limited influence on other jurisdictions. Interpretive principles are all part of the same transjurisdictional common law ecosystem.[202] As Professor Eskridge has argued, the canons are “America’s common law of statutory interpretation—the nation’s ongoing experience with applying statutes to concrete circumstances.”[203] And indeed, just as substantive common law principles evolve over time, the application of individual canons can change over time.[204] Moreover, just like substantive common law rules migrate among state courts relatively freely,[205] so, too, do state courts adopt interpretive principles from other states. As Professor Gluck discovered, for example, there is a growing consensus among state courts to pursue what she has called a “modified textualism” that seeks greater constraints than traditional purposivism but without an unbending refusal to consider legislative history.[206] An emerging example of the phenomenon of migrating interpretive principles is the judicial reliance on corpus linguistics tools to interpret statutes.[207] Corpus linguistics is a set of procedures or methods that help researchers analyze linguistic information in bodies of text.[208] The use of this methodology for statutory interpretation was advocated for many years by a judge of a single state’s high court.[209] Corpus linguistics tools subsequently were adopted by courts in other states.[210] And most recently, corpus linguistics has made the jump from state to federal courts, now claiming adherents at all levels of the federal judiciary.[211] Because principles of interpretation migrate across state boundaries and between state and federal courts, excluding state court methodological decisions omits a large amount of interpretive methodology data and understates the influence that state methodology has on federal interpretive decisions.

But, just as I would go further than Professors Krishnakumar and Nourse to include interpretations by lower courts, state courts, and administrative agencies, I agree with their hesitation to canonize “aspirational rules of interpretation that have been advocated by scholars.”[212] In the absence of use by courts or agencies, it is difficult to know whether and how an interpretive principle would be applied in real-life disputes. Furthermore, a principle that has not been adopted by courts or agencies lacks an official decision-maker’s imprimatur. Accordingly, I would exclude legal scholars from my definition of “legal interpreters” for the purpose of this criterion. As a result, if an interpretive principle, even a very sensible one, has not been adopted by the kind of legal interpreters described above, I would not consider it a canon.[213]

In addition to the question of who is a legal interpreter, there is the question of what it means for interpreters to “use” an interpretive principle. There is a strong temptation to identify the frequency with which an interpretive principle must be cited in court opinions before it is considered a canon. Despite their acknowledgement that any number of cases chosen is arbitrary, for example, Krishnakumar and Nourse suggest that somewhere in the range of seven to ten Supreme Court cases is usually appropriate, provided that the number chosen ensures that “loose judicial commentary is not labeled a canon on the basis of one or two (or even three) stray utterances by the U.S. Supreme Court.”[214] I share their concern that any minimum number of citations is arbitrary and that identifying a particular number can be misleading. In my view, the many qualitative factors that influence whether the use of an interpretive principle is appropriate make the precise quantity of opinions that adopt a particular principle unimportant. Some principles of interpretation may be very powerful, but only appropriate for a limited class of statutes.[215] Other principles may arise only in a limited class of factual situations.[216] Still others may be deployed—or omitted—by a court for strategic or rhetorical reasons.[217] As a result, I would put little weight on the number of citations attributable to an interpretive principle when deciding whether it is actually used by legal interpreters.

In lieu of a particular number of citations, I suggest the following qualitative test: legal interpreters use an interpretive principle within the meaning of this criterion if they rely on it sufficiently frequently so that it would be considered effective advocacy for a party to invoke it in an actual legal dispute. As skilled litigators know, the cost of making unpersuasive arguments can be high. One federal court of appeals judge described the effect of advancing unpersuasive arguments in the following way: “Weak issues in a brief are more than surplusage. They have a definite tendency to infect the entire brief because they detract from the efficacy of your good issues.”[218] Advocates, therefore, have a strong incentive not to assert an interpretive principle if it is so rarely used that it would not be worth invoking in an actual dispute. Accordingly, if an interpretive principle is so rarely used by courts or agencies that it would not be effective advocacy for a party to rely on it, then I do not consider it “used by legal interpreters” within the meaning of this criterion.

B. Affects Interpretive Outcomes

In addition to actually being used by legal interpreters, an interpretive principle should be considered a canon only if following it affects interpretive outcomes by making a possible interpretation more or less persuasive.[219] To clarify, I do not mean that the interpretive principle must finally resolve a case to be considered a canon; that would be an unrealistic expectation because a case’s outcome may turn on many things other than the resolution of a particular interpretive dispute.[220] Rather, an interpretive principle makes a possible interpretation more or less persuasive if following it, in a case in which it applies,[221] gives that interpretation more or less persuasive force. The work of several scholars, including Professors Bruhl, Gluck, and Rosenkranz, suggests the connection between an interpretive principle’s influence on interpretive outcomes and its canonicity.[222]

The requirement I propose would easily be satisfied by most well-accepted linguistic and substantive canons. For example, consider the application of a common linguistic canon, expressio unius est exclusio alterius (the expression of one thing excludes all others)[223] to Federal Rule of Civil Procedure 9(b). FRCP 9(b) expressly requires a heightened pleading standard for allegations of fraud and mistake.[224] Expressio unius, therefore, suggests that FRCP 9(b) does not require heightened pleading for allegations other than those of fraud or mistake. Indeed, this is how the Supreme Court applied expressio unius in a case alleging civil rights violations under § 1983.[225] The Court concluded that the specific mention of fraud and mistake in FRCP 9(b) precludes courts from requiring heightened pleading for allegations of civil rights violations.[226] Because the application of expressio unius made one possible interpretation of 9(b) more persuasive—that there is no heightened pleading requirement for civil rights allegations—it satisfies the condition stated above.[227]

Most common canons, both textual and substantive, make one possible interpretation more or less persuasive; but, this is not true of every interpretive principle, even principles that have been identified as canons. In particular, this criterion would exclude principles that are merely reminders and principles that are not sufficiently rule-like. First, if a statement is merely a reminder, it does not affect the outcome of an interpretive decision. For example, consider the interpretive principle called the “pet fish” canon, which provides that “[p]hrases and word clusters might have an ordinary meaning different from the aggregation of the separate words.”[228] That is, the term “pet fish” might mean something different than the words “pet” and “fish” interpreted successively. But the pet fish canon merely states that a phrase might have a different meaning than the disaggregated words of that phrase; it doesn’t actually identify circumstances in which a phrase does mean something different than its words disaggregated. What’s more, the pet fish canon does not help an interpreter identify a group of words as a phrase in the first place. Rather, it simply asks the interpreter to keep an eye out for multiword phrases in statutes. Because the pet fish canon merely serves as a reminder to think about phrase meaning, it does not make a possible interpretation either more or less persuasive. As a result, it does not affect the outcome of any interpretive decision as described above.

Consider the application of the pet fish canon to a statute that prohibits “willful and wanton” conduct, a common statutory formulation. Without context, this phrase could be read either one of two ways. On one hand, it might prohibit one type of conduct: “willful and wanton” conduct. On the other hand, it might be read to prohibit two types of conduct: both “willful” conduct and “wanton” conduct. Indeed, although “willful and wanton” is a common statutory phrase,[229] there are also many statutes that prohibit willful[230] conduct or wanton[231] conduct separately. As a result, this is a classic pet fish canon scenario: willful and wanton could either be a single phrase or it could be two separate terms that each modify conduct. However, while the pet fish canon reminds the interpreter that willful and wanton can be read either as one term or two, it does nothing to suggest that one of these interpretations rather than the other is correct in any given circumstance. Context could provide clues about which reading is more appropriate to a particular statute, of course, but it is this context, not the pet fish canon, that is contributing to statutory meaning by making one possible interpretation more or less persuasive.[232]

Second, in line with some other scholars, I would not consider a principle a canon if it is not sufficiently “rule-like.”[233] Professor Mendelson suggested that, to be sufficiently rule-like, an interpretive principle must be “applicable across a range of statutory settings.”[234] A principle fails to be a canon under this formulation, it would seem, if it is confined to the interpretation of a single statute. Professor Mendelson’s formulation is at once broader and narrower than the suggestion of Professors Krishnakumar and Nourse, who propose that a canon must reflect “the agreement of Supreme Court Justices appointed by different parties and across ideological divides.”[235] This latter formulation is rooted, perhaps, in the concern that an ideologically bound interpretive tool is less a principle than an excuse; or, in other words, that it is not a principle at all. Professor Sinclair adds that the formulation of a canon must be relatively stable; even if “statute-like determinacy” is too demanding, he argues, “some stability or limit on variety in expression, some canonical form, is necessary.”[236]

I agree that an interpretive principle must be sufficiently rule-like to be a canon.[237] But, I would frame the analysis differently. I propose that a principle is sufficiently rule-like only if it can be generalized into a prescription that applies to an open‑ended class of potential future disputes. This proposed test follows from Professor Schauer’s observation that the statement of a reason for an outcome implies the existence of a general principle that controls the particular outcome.[238] Consider Schauer’s example: “You ask why I am carrying an umbrella, and I respond that the weather forecast predicted rain. Although the response is not explicitly prescriptive, it embraces the mandate, ‘Carry an umbrella when rain is forecast . . . .’”[239] In other words, by providing a reason for carrying the umbrella, the speaker implicitly states a general rule that controls future cases: any time rain is forecast, we would expect to see the speaker with an umbrella.

Adopting Schauer’s reasoning, in my formulation an interpretive principle is rule-like only if it is stated generally enough to embrace some unknown future decisions. By contrast, I would not consider a principle to be rule-like if it is a “ticket good for one day only,”[240] that is, applicable only to the decision in which it is announced. Similarly, I would not consider a principle to be rule-like to the extent that it provides a rule of decision that applies only because it is law of the case[241] or that applies only to a closed class of pending cases.[242] Most of the commonly invoked interpretive principles would easily satisfy this condition: they are stated in general terms and could apply not only to the decision in which they are announced but, at least potentially, to future decisions as well.[243]

The definition of rule-like that I propose is broader than the one suggested by Professor Mendelson. For example, in Mendelson’s view, the group of interpretive principles that apply to the Sherman Act alone[244] would probably not be considered canons because they do not apply to a range of statutory situations.[245] By contrast, under the test I propose, an interpretive principle that applies only to the Sherman Act could still be sufficiently rule-like, provided that it is framed to apply to an open-ended class of future cases. In contrast to Sinclair’s formulation,[246] in my view, an interpretive principle is sufficiently rule-like only if it applies to future cases, no matter how definitely it is stated. By contrast, it is possible to imagine a definitely stated interpretive principle that applies only to a closed class of cases,[247] which, in contrast to Sinclair, I would not consider rule-like.

As opposed to Krishnakumar and Nourse, in my view, a principle of interpretation may be rule-like even if it does not reflect “the agreement of Supreme Court Justices appointed by different parties and across ideological divides.”[248] Under their definition, even some well-established interpretive principles, like the major questions doctrine, might not qualify as canons.[249] Despite the fact that it has been around for decades, and despite the fact that it has been used in multiple Supreme Court opinions, the major questions doctrine has been overwhelmingly,[250] if not solely,[251] used in opinions authored by Justices appointed by Republican presidents. Moreover, even if Krishnakumar and Nourse’s definition was viable at the time they wrote it, it is still more questionable now that recent appointments likely have left the Court dominated for a generation by Justices appointed by Republican presidents. Under Krishnakumar and Nourse’s formulation, an interpretive principle like the major questions doctrine might fail to become a canon, even if it is used liberally by the majority of this Court for decades to come.

To be clear, I share the concern implicit in Nourse and Krishnakumar’s definition, which I take to be that overtly partisan or ideological interpretations are unprincipled attempts to hide motivated reasoning. But, here, I believe Krishnakumar and Nourse blur the line between the normative and descriptive. In my view, an interpretive principle that is used for partisan or ideological purposes is likely a normatively unattractive principle of interpretation that should be rejected by courts. But the fact that it is a bad principle of interpretation is different from the determination that it is (or is not) a canon. Although normative critiques of the canons are crucially important,[252] it is work that is outside the scope of this Article, which has a purpose of advancing the discussion related to identifying canons. As a result, the broader descriptive definition I propose surely may encompass interpretive principles that turn out to be normatively unattractive on further analysis.

C. Claims Theoretical Justification

An interpretive principle should be considered a canon only if there are interpreters who claim (or would claim if asked) that it is theoretically justified. Some canons are defended on the ground that they reflect the way that ordinary people use language, including their word choices and grammatical conventions.[253] Most often, this defense is made of the textual canons,[254] but some substantive canons have been defended on this ground as well.[255]

Other canons are defended on the ground that they reflect legislative intentions or behavior, including assessments about how legislatures try to accomplish their goals, how they intend to interact with other branches of government, or how legislators themselves use language.[256] The canon of constitutional avoidance, for example, “has been justified on the ground that Congress does not usually intend for its statutes to provoke serious constitutional questions.”[257] Still other canons are defended on the ground that they reflect the law as it actually is.[258] That is, if canons are, as some scholars have argued, a species of common law,[259] then they require no additional justification other than that they are legal rules for courts to apply.

Moreover, many canons are defended on normative grounds.[260] Federalism canons, for example, are justified on the ground that they protect underenforced federalism values; the major questions doctrine has been defended on the ground that it prevents overbroad delegations.[261] Other canons are defended on the ground that they improve the institutional relationship among the branches of government.[262] For example, Professors Gluck and Bressman describe the “feedback” justification of some canons; by demonstrating to the legislature how they will read statutory language, courts can “teach Congress how to legislate better.”[263] Finally, many of the canons are justified on more than one of these grounds.[264]

Although it is not a difficult requirement to meet, it is essential to canonicity that interpreters claim (or would claim if asked) that an interpretive principle is theoretically justified.[265] An interpretive principle, at bottom, is a reason for interpreting language in a particular way. By contrast, if users of an interpretive principle claim no theoretical justification for its use, then the principle ceases to be a reason and instead reveals itself to be an ipse dixit assertion. An interpretation that rests on an ipse dixit assertion, rather than on a reason, is open to the charge that it is irrational and perhaps even illegitimate.[266] An ipse dixit assertion cannot be generalized to future situations, explained, challenged, or defended.[267] Put otherwise, an interpretive principle that has no claim to a theoretical justification is no better than the least satisfying of all explanations, “because I said so.”[268]

For these reasons, I disagree with Justice Scalia, who, in his scholarly work on the canons, comes very close to asserting that some canons require no theoretical justification. He finds it hard to believe that anyone could criticize the “Latin” canons, for example, because they are “so commonsensical”; and, conversely, he derides a putative canon because he has “[n]ever heard of it.”[269] (In fact, the Latin canons he describes are often justified on a number of grounds: that they are so well-known that Congress legislates in light of them;[270] that they reflect the way that people speak;[271] and that they are so well-established that they are part of the law.[272]) But, if taken seriously, Justice Scalia appears to suggest that no reason other than the interpreter’s intuition (or perhaps his intuition!) is necessary to adopt or reject a canon. Intuition about the existence of a canon, even if widely shared, cannot be effectively challenged or defended. As a result, applying an interpretive principle derived solely from intuition is very close to resting on the bare assertion “because I said so,” providing no more reason for the outcome than a coin-flip.[273]

To be clear, I am not suggesting that the theoretical justification must be undisputed before a principle is considered a canon.[274] There are many interpretive principles that rest on disputable empirical grounds. For example, whether a canon may be justified for its “feedback” effect depends on whether Congress in fact employs statutory language in response to court interpretations.[275] This empirical question is probably impossible to answer: sometimes Congress seems particularly attentive to court interpretations, other times, less so.[276] For most statutes, it is impossible to know whether Congress used particular language because of feedback from the courts, despite this feedback, or wholly independent of it. Moreover, there are many interpretive principles that rest on contestable theoretical premises. Some interpreters, for example, believe that reliance on legislative history is illegitimate. An interpreter holding this view is unlikely to agree with an interpretation that relies on legislative history, no matter how well-articulated. Because there are disputable empirical and theoretical premises underpinning many of the canons, it is too much to ask that an interpretive principle be indisputably justified before it is considered a canon; such an insistence would result in a stalemate over the canonicity of even the most well-known and oft-used interpretive principles.[277] For this reason, I agree with Krishnakumar and Nourse’s conclusion that canonicity does not require actual proof that a principle is theoretically justified.[278]

But, in contrast to Nourse and Krishakumar, I would still require a claim of theoretical justification, even if proof is lacking that a justification is warranted. As noted above, claim of justification is essential because it provides a reason for relying on it—even if the basis is ultimately contestable. By separating the fact that an interpretive principle claims a theoretical justification from the persuasiveness of the justification itself, interpreters are better able to evaluate the strength of the justification without conflating the existence of the canon with its merits. Rather than creating dubious categories like “wholly illegitimate,”[279] “faux,”[280] or “unreliable”[281] to describe unmeritorious canons, an interpreter can more clearly argue why a canon is unmeritorious. For example, if an interpreter concludes that the given justification for the rule against surplusage is based on an erroneous assumption, then she can squarely argue that it is a bad canon and should be rejected rather than implying that it is not a canon at all or that it falls into some gray area, neither canonical nor something else.

Finally, when searching for a theoretical justification, a distinction must be drawn between longevity for its own sake and longevity as evidence of authoritativeness. I would not consider longevity, standing alone, to be a justification for an interpretive principle. As Krishnakumar and Nourse articulated, reliance on longevity as a sufficient criterion for canonicity creates the twin risks of entrenching interpretive principles long out of use and of barring entry to new, even well-attested, interpretive principles.[282] But, I would go even further: longevity should not even be considered an indicator of canonicity. It is easy to confuse longevity with correctness, especially in a legal system built on precedent. Nevertheless, the two concepts are distinct. A new rule should not be debarred because of its novelty any more than an old rule should persist from “blind imitation of the past.”[283] Indeed, a dynamic canon of interpretive maxims, capable of adding or dropping principles over time, would be in line with other canons, like literary or religious canons.[284] Just as the keepers of these other canons drop or add texts as cultural needs change,[285] so, too, is it fair to expect that the canons of interpretation will be pruned and supplemented as our understanding of both language and legal and political institutions develops.

By contrast, longevity could be evidence of legal authority, which I do consider to be a theoretical justification sufficient to satisfy the final criterion for canonicity. As some scholars have argued, at least some canons of interpretation are a kind of common law.[286] To the extent that this is true, courts are justified in applying the canons in the same way that they are obligated to apply the substantive law, including, of course, all the limitations and nuance contained within the canons themselves.

IV. Conclusion

What is a canon of statutory interpretation? It is striking that a concept with such rhetorical force, and which is discussed in countless Supreme Court cases and scholarly works, has no settled definition. Without a definition, we cannot properly evaluate whether Justice Gorsuch was right to dismiss the canon of donut holes in Bostock or whether scholars are correct when they propose new canons. Moreover, without criteria for canonicity, we cannot know whether an interpretive principle that lacks merit is a bad canon, an invalid canon, or something else altogether.

This Article proposes and defends three criteria for assessing whether an interpretive principle should be considered a canon. It asks the interpreter to determine whether the interpretive principle is used by legal interpreters, whether using it would affect an interpretive outcome, and whether it claims a theoretical justification. Focusing on these criteria of canonicity, an interpreter will be able to distinguish normative from descriptive arguments about the canons, allowing for clearer and more precise debates about both.


  1. Guido Calabresi, A Common Law for the Age of Statutes 163 (1982) (setting out the role of courts in a society dominated by statutory law).

  2. Anita S. Krishnakumar, Backdoor Purposivism, 69 Duke L.J. 1275, 1277, 1295 (2020) (noting that the court speaks in terms of canons); Aaron-Andrew P. Bruhl, Communicating the Canons: How Lower Courts React When the Supreme Court Changes the Rules of Statutory Interpretation, 100 Minn. L. Rev. 481, 496, 523–24 (2015) (describing how changes in Supreme Court canon use affect lower court canon use).

  3. Jacob Scott, Codified Canons and the Common Law of Interpretation, 98 Geo. L.J. 341, 350 (2010) (describing codified canons of statutory interpretation).

  4. Grace E. Hart, State Legislative Drafting Manuals and Statutory Interpretation, 126 Yale L.J. 438, 455 (2016) (describing legislative drafting manuals’ use of canons).

  5. Anita Krishnakumar, Dueling Canons, 65 Duke L.J. 909, 913–14, 925 (2016) (studying how the Supreme Court opinions use both the same canons in majority and dissenting opinions); Bruhl, supra note 2, at 519–20 (describing Supreme Court and lower court use of the canons).

  6. Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 949, 970 (2013) (reporting results of interviews with congressional staff about their canon use).

  7. Jerry L. Mashaw, Norms, Practices, and the Paradox of Deference: A Preliminary Inquiry into Agency Statutory Interpretation, 57 Admin. L. Rev. 501, 533 (2005) (theorizing agency statutory interpretation); Christopher J. Walker, Inside Agency Statutory Interpretation, 67 Stan. L. Rev. 999, 1032 (2015) (reporting results of surveys of agency statutory interpretation).

  8. E.g., Abbe R. Gluck, Congress, Statutory Interpretation, and the Failure of Formalism: The CBO Canon and Other Ways that Courts Can Improve on What They Are Already Trying to Do, 84 U. Chi. L. Rev. 177, 182 (2017) (identifying the “CBO Canon”); Lisa Heinzerling, The Power Canons, 58 Wm. & Mary L. Rev. 1933, 1944 (2017) (identifying the “power” canons); Anita S. Krishnakumar, The Hidden Legacy of Holy Trinity Church: The Unique National Institution Canon, 51 Wm. & Mary L. Rev. 1053, 1057, 1064, 1068 (2009) [hereinafter Krishnakumar, Holy Trinity] (identifying the “unique national institution” canon); Nina A. Mendelson, Change, Creation, and Unpredictability in Statutory Interpretation: Interpretative Canon Use in the Roberts Court’s First Decade, 117 Mich. L. Rev. 71, 110–15 (2018) (discussing the newly identified “no elephants in mouseholes,” “location of codification,” “veteran’s benefits,” and “jurisdictional rules” canons); see also Anita S. Krishnakumar, Reconsidering Substantive Canons, 84 U. Chi. L. Rev. 825, 855–56 (2017) [hereinafter Krishnakumar, Reconsidering Substantive Canons] (noting frequency with which the Court recognizes new canons); Nicholas Quinn Rosenkranz, Federal Rules of Statutory Interpretation, 115 Harv. L. Rev. 2085, 2143 (2002) (“The Supreme Court may . . . foreshadow[] a new canon with insinuating dicta.”).

  9. Bruhl, supra note 2, at 519–20 (describing that courts sometimes abandon canons).

  10. Mendelson, supra note 8, at 78 (“Some traditionally recognized canons . . . are used rarely and may soon evaporate altogether.”); Bruhl, supra note 2, at 547–48 (describing how lower courts often follow Supreme Court’s use or abandonment of canons).

  11. Antonin Scalia, A Matter of Interpretation 25, 27 (1997) (discussing judicial reliance on the canons); Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts xi–xvi (2012) (explaining fifty-seven canons and other principles of interpretation); John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 705 n.140, 739 (1997) (“[T]extualist judges also use canons of construction, which routinely form part of the context used to interpret statutes.”); Amy Coney Barrett, Substantive Canons and Faithful Agency, 90 B.U. L. Rev. 109, 121 (2010) (“Textualists routinely bring canons to bear on the interpretation of statutes. They are probably the biggest proponents of linguistic canons. Textualists, however, also embrace substantive canons.”); James J. Brudney & Corey Ditslear, Canons of Construction and the Elusive Quest for Neutral Reasoning, 58 Vand. L. Rev. 1, 4 (2005) (“Judges, especially textualist-oriented judges, praise the canons for their relative clarity and commonsense virtues.”); John F. Manning, Constitutional Structure and Statutory Formalism, 66 U. Chi. L. Rev. 685, 688 (1999) (“[M]odern formalists . . . routinely derive statutory meanings from extratextual sources, including unenacted materials such as cases or treatises that define terms of art or prescribe canons of construction.”); Caleb Nelson, What Is Textualism?, 91 Va. L. Rev. 347, 383–84 (2005) (describing types of canons favored by textualists).

  12. Robert A. Katzmann, Judging Statutes 51, 55 (2014) (“[I]ndeed, I have found that universally accepted canons can be useful.”).

  13. See William N. Eskridge, Jr. et al., Cases and Materials on Legislation and Regulation: Statutes and the Creation of Public Policy 582, 617, 648, 699, 700 (6th ed. 2020) (describing pragmatic defense of canons).

  14. Frank B. Cross, The Theory and Practice of Statutory Interpretation 85 (2009) (“[T]extualism, legislative history and pragmatism . . . could theoretically rely on canons of construction or reject them.”); John F. Manning, Legal Realism & the Canons’ Revival, 5 Green Bag 283, 284, 288 (2002) (“[A] large and growing number of academics (and academics-turned-judges) now believe in the utility of canons of construction . . . [and] the newly faithful cover a broad philosophical spectrum.”).

  15. Compare William N. Eskridge, Jr., Interpreting Law: A Primer on How to Read Statutes and the Constitution 12 (2016), with Scalia & Garner, supra note 11, at xi–xvi, 9 (showing that there is a great deal of overlap among different collections of canons).

  16. Scalia, supra note 11, at 26–27 (asserting that the Latin canons are too “commonsensical” to question, and a canon should be discarded because Scalia has “[n]ever heard of it”).

  17. See Eben Moglen & Richard J. Pierce, Jr., Sunstein’s New Canons: Choosing the Fictions of Statutory Interpretation, 57 U. Chi. L. Rev. 1203, 1225–26 (1990) (noting that some of Sunstein’s proposed canons are inconsistent with current law); Anita S. Krishnakumar & Victoria F. Nourse, The Canon Wars, 97 Tex. L. Rev. 163, 174–77 (2018) (noting that some of Eskridge’s canons are “aspirational,” that is, normatively attractive but not currently embraced by the Supreme Court). Krishnakumar and Nourse themselves blur the descriptive and normative by asserting that “the basic thread connecting the canons is (or should be) established convention.” Krishnakumar & Norse, supra, at 188 (emphasis added).

  18. See Reed Dickerson, The Interpretation and Application of Statutes 228 (1975).

  19. William N. Eskridge, Jr., Dynamic Statutory Interpretation, 135 U. Pa. L. Rev. 1479, 1490 n.41 (1987) (“[F]or it rests upon the highly unreliable maxim of statutory construction ‘inclusio unius est exclusio alterius.’”).

  20. William Baude & Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079, 1121 (2017); Scalia & Garner, supra note 11, at xvii, 9, 343.

  21. Scalia, supra note 11, at 27.

  22. Frank H. Easterbrook, Do Liberals and Conservatives Differ in Judicial Activism?, 73 U. Colo. L. Rev. 1401, 1405 (2002) (arguing that the canon of constitutional avoidance is “wholly illegitimate”).

  23. Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1747 (2020).

  24. Id. at 1753.

  25. Id. at 1746.

  26. Id. at 1746–47.

  27. Id. at 1747.

  28. In other high-profile cases, too, judges rejected interpretations after concluding that they rested on a principle that is not a canon. E.g., Yates v. United States, 574 U.S. 528, 565 (2015) (Kagan, J., dissenting) (rejecting the application of an interpretive principle because it was not listed in a particular treatise of canons); FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 190 (2000) (Breyer, J., dissenting) (doubting the existence of the major questions doctrine).

  29. Gluck, supra note 8, at 182.

  30. Kevin M. Stack, The Enacted Purposes Canon, 105 Iowa L. Rev. 283, 285 (2019) (“This Article’s core argument is that this principle—which I call the enacted purposes canon—is and should be viewed as a bedrock principle of statutory interpretation.”).

  31. Heinzerling, supra note 8, at 1937.

  32. Krishnakumar, Holy Trinity, supra note 8, at 1057.

  33. Ethan J. Leib & James J. Brudney, The Belt-and-Suspenders Canon, 105 Iowa L. Rev. 735, 736 (2020) (“We aim here to christen a canon into the doctrines of statutory interpretation. We propose that a ‘belt-and-suspenders’ canon, invoked with some frequency during the legislative drafting process, should be recognized by judges and scholars as presumptively probative and at times controlling.”).

  34. E.g., Mendelson, supra note 8, at 110–15 (describing several new canons); Jonathon S. Byington, The Fresh Start Canon, 69 Fla. L. Rev. 115, 133 (2017) (introducing the fresh start canon); see also Krishnakumar, Reconsidering Substantive Canons, supra note 8, at 855–56; Rosenkranz, supra note 8, at 2143 (“The Supreme Court may . . . foreshadow[] a new canon with insinuating dicta.”).

  35. E.g., Eskridge, supra note 15, at 407–45; Barrett, supra note 11, at 120–21.

  36. Dickerson, supra note 18.

  37. Eskridge, supra note 19, at 1490 n.41(“[I]t rests upon the highly unreliable maxim of statutory construction ‘inclusio unius est exclusio alterius.’”).

  38. Baude & Sachs, supra note 20, at 1121; Scalia & Garner, supra note 11, at xvii, 9, 343.

  39. Scalia, supra note 11, at 27.

  40. Easterbrook, supra note 22 (arguing that the canon of constitutional avoidance is “wholly illegitimate”).

  41. Gluck & Bressman, supra note 6, at 926–27 (reporting results of interviews with congressional staff about their canon use); Hart, supra note 4, at 455–57 (describing legislative drafting manuals’ use of canons).

  42. Mashaw, supra note 7, at 527–28 (theorizing agency statutory interpretation); Walker, supra note 7, at 1030 (reporting results of surveys of agency statutory interpretation).

  43. Krishnakumar, Reconsidering Substantive Canons, supra note 8, at 855–56 (noting frequency with which the Court recognizes new canons); Bruhl, supra note 2, at 522 (describing that courts sometimes abandon canons); see Aaron-Andrew P. Bruhl, Statutory Interpretation and the Rest of the Iceberg: Divergences Between the Lower Federal Courts and the Supreme Court, 68 Duke L.J. 1, 30 (2018) (estimating that, in the past decade, district courts have issued more than 100,000 unpublished decisions that involve statutory interpretation); Mendelson, supra note 8, at 78 (noting that some “traditionally recognized canons . . . are used rarely and may soon evaporate altogether”).

  44. John F. Manning & Matthew C. Stephenson, Legislation and Regulation 202 (2d ed. 2013) (explaining that “semantic” or “linguistic” canons are “generalizations about how the English language is conventionally used and understood”); e.g., Eskridge, supra note 15, at 407–45 (categorizing the canons by type).

  45. Amy Coney Barrett, Congressional Insiders and Outsiders, 84 U. Chi. L. Rev. 2193, 2203–04 (2017) (arguing that linguistic canons are legitimated by the fact that they “are designed to capture the speech patterns of ordinary English speakers and, in some cases, of the subclass of lawyers”); see Scalia, supra note 11, at 25 (articulating a strong support for judicial reliance on the canons). See generally Scalia & Garner, supra note 11 (explaining fifty-seven canons and other principles of interpretation); Stack, supra note 30, at 285 (“This Article’s core argument is that this principle—which I call the enacted purposes canon—is and should be viewed as a bedrock principle of statutory interpretation.”); Leib & Brudney, supra note 33, at 736 (“We propose that a ‘belt‑and‑suspenders’ canon . . . should be recognized . . . .”).

  46. See Moglen & Pierce, supra note 17, at 1225–26 (noting that some of Sunstein’s proposed canons are inconsistent with current law); Krishnakumar & Nourse, supra note 17, at 188–89 (“[T]he ultimate test of a canon . . . reflects the agreement of Supreme Court Justices appointed by different parties and across ideological divides.” (emphasis omitted)); Baude & Sachs, supra note 20, at 1105, 1122 (“[T]he canons stand on their own authority as a form of common law.”).

  47. Stack, supra note 30, at 316–17, 333; Leib & Brudney, supra note 33, at 753–54, 769; Byington, supra note 34, at 133–34.

  48. Tara Leigh Grove, Which Textualism?, 134 Harv. L. Rev. 265, 279 (2020) (identifying competing versions of textualism); see also Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1755–56 (2020) (Alito, J., dissenting) (revealing some different ways that judges claiming the mantle of textualism engage in statutory interpretation).

  49. Victoria Nourse, Misreading Law, Misreading Democracy 104 (2016) (noting that the resurgence of canon use is largely driven by the fact that textualists are often strong “canon enthusiasts”); Barrett, supra note 11, at 121 (“Textualists routinely bring canons to bear on the interpretation of statutes. They are probably the biggest proponents of linguistic canons. Textualists, however, also embrace substantive canons.”).

  50. E.g., Neil M. Gorsuch, A Republic, if You Can Keep It 128–29 (2019) (positing textualism as the legitimate principle of statutory interpretation); Barrett, supra note 45, at 2204 (providing justification for textualism and linguistic canons); Bostock, 140 S. Ct. at 1825 (Kavanaugh, J., dissenting) (claiming to interpret the text in accordance with Scalia’s version of textualism).

  51. Dickerson, supra note 18, at 229 (referring to “plain meaning” as a canon, rule, and doctrine, interchangeably); see, e.g., Conn. Nat’l Bank v. Germain, 503 U.S. 249, 253 (1992) (“[C]anons of construction are no more than rules of thumb . . . .”); cf. Cracker Barrel Old Country Store, Inc. v. Cincinnati Ins. Co., 499 F. App’x 559, 564–65 (6th Cir. 2012) (“[C]anons of construction, though helpful, should always be tested against the other interpretive tools at a court’s disposal.” (quoting In re Estate of Tanner, 295 S.W.3d 610, 624 n.13 (Tenn. 2009))).

  52. See Dickerson, supra note 18, at 228 (suggesting that not all rules of interpretation are the same or should be treated the same); Kent Greenawalt, Statutory and Common Law Interpretation 117 (2013) (distinguishing between “canons” and “other principles of interpretation”).

  53. Krishnakumar & Nourse, supra note 17, at 182 (a sign that an interpretive principle is a canon is its frequency of use); Scalia & Garner, supra note 11, at 9 (“Mostly, the canons exist within the thousands of law reports scattered through a law library . . . .”).

  54. Heinzerling, supra note 8, at 1944 (“It may well be that valid or at least normatively unproblematic interpretive canons are usually distinguished by their antiquity.”); Cross, supra note 14, at 86 (“Many canons are so historical that they are commonly expressed in the Latin language used by Old English common law courts, and some date to the sixteenth century.”).

  55. Scalia, supra note 11, at 26 (some canons are “so commonsensical that, were the canons not couched in Latin, you would find it hard to believe anyone could criticize them”); Greenawalt, supra note 52, at 117 (arguing that some linguistic canons are “intuitively appealing approaches to understanding language”); Cross, supra note 14, at 86 (“The most basic linguistic canons are so unexceptional that they are typically unstated.”).

  56. Mendelson, supra note 8, at 134 (“[G]rammatical and syntax canons [are often] described as required by ordinary English conventions . . . .”); Krishnakumar & Nourse, supra note 17, at 180–81 (“[L]anguage canons are thought to reflect rules of grammar, logic, sentence organization, or even congressional drafting.”); Dickerson, supra note 18, at 228 (noting that canons “represent at best only probabilities based on past experience”).

  57. Eskridge et al., supra note 13, at 582, 617, 648 (introducing different types of canons); Karl N. Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons About How Statutes Are to Be Construed, 3 Vand. L. Rev. 395, 401 (1950) (setting out the now-famous thrusts and parries); Cross, supra note 14, at 85 (describing the canons of interpretation); Conn. Nat’l Bank v. Germain, 503 U.S. 249, 254 (1992) (“When the words of a statute are unambiguous, then, this first canon is also the last: ‘judicial inquiry is complete.’” (quoting Rubin v. United States, 449 U.S. 424, 430 (1981))).

  58. Cass R. Sunstein, Interpreting Statutes in the Regulatory State, 103 Harv. L. Rev. 405, 453 (1989) (describing the canons as a subset of all interpretive principles); Greenawalt, supra note 52, at 117 (distinguishing between “canons” and “other principles of interpretation”). The Supreme Court, too, has distinguished the canons from run‑of‑the‑mill principles of interpretation. Ali v. Fed. Bureau of Prisons, 552 U.S. 214, 231 (2008) (Kennedy, J., dissenting) (distinguishing between canons and “other common principles of interpretation”).

  59. TRW Inc. v. Andrews, 534 U.S. 19, 31 (2001) (“It is ‘a cardinal principle of statutory construction’ that ‘a statute ought, upon the whole, to be so construed that, if it can be prevented, no clause, sentence, or word shall be superfluous, void, or insignificant.’” (quoting Duncan v. Walker, 533 U.S. 167, 174 (2001))); INS v. Cardoza-Fonseca, 480 U.S. 421, 442–43 (1987) (“Few principles of statutory construction are more compelling than the proposition that Congress does not intend sub silentio to enact statutory language that it has earlier discarded in favor of other language.” (citing Nachman Corp. v. Pension Benefit Guar. Corp., 446 U.S. 359, 392–93 (1980))).

  60. Circuit City Stores, Inc. v. Adams, 532 U.S. 105, 114–15 (2001) (referring to the maxim ejusdem generis); Syed v. M-I, LLC, 853 F.3d 492, 501 (9th Cir. 2017) (referring to the “familiar judicial maxim expressio unius est exclusio alterius”); Warren v. Bristol Warren Sch. Dist., 159 A.3d 1029, 1039 (R.I. 2017) (referring to maxims of statutory construction); State v. Gaines, 206 P.3d 1042, 1046 (Or. 2009) (“[T]he court may resort to general maxims of statutory construction . . . .” (quoting PGE v. Bureau of Lab. & Indus., 859 P.2d 1143, 1146 (Or. 1993))).

  61. Katzmann, supra note 12, at 55 (“In approaching the interpretive task, a judge can use several tools, including: text, statutory structure, history, word usage in other relevant statutes . . . .”); Michael Sinclair, Traditional Tools of Statutory Interpretation 2 (2013); Llewellyn, supra note 57, at 401 (“Hence there are two opposing canons on almost every point. An arranged selection is appended. Every lawyer must be familiar with them all: they are still needed tools of argument.”).

  62. Dickerson, supra note 18, at 229–30, 233 (referring to the “plain meaning rule,” “golden rule,” and “whole statute rule”); Eskridge, supra note 15, at 407, 411 (referring to the “ordinary meaning rule” and “whole act rule”); United States v. Pacheco, 225 F.3d 148, 154 (2d Cir. 2000) (“The ‘whole act’ rule of statutory construction exhorts us to read a section of a statute not ‘in isolation from the context of the whole Act’ but to ‘look to the provisions of the whole law, and to its object and policy.’” (quoting Richards v. United States, 369 U.S. 1, 11 (1962))).

  63. Dickerson, supra note 18, at 224, 228 (“One group consists of rebuttable presumptions . . . .”); Medtronic, Inc. v. Lohr, 518 U.S. 470, 485 (1996) (“[W]e used a ‘presumption against the pre-emption of state police power regulations’ . . . .” (quoting Cipollone v. Liggett Grp., Inc., 505 U.S. 504, 518, 523 (1992))); see also Landgraf v. USI Film Prods., 511 U.S. 244, 265 (1994) (“[T]he presumption against retroactive legislation is deeply rooted in our jurisprudence . . . .”); Abbe R. Gluck, Justice Scalia’s Unfinished Business in Statutory Interpretation: Where Textualism’s Formalism Gave Up, 92 Notre Dame L. Rev. 2053, 2061 (2017) (“There are certainly more than seventy interpretive presumptions and, really, there are probably closer to one hundred.”).

  64. Cross, supra note 14, at 109 (“Significantly, advocates of virtually all theories of statutory interpretation accept the validity of the absurdity doctrine.”); Nat’l Cable & Telecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 985 (2005) (“There is genuine confusion in the lower courts over the interaction between the Chevron doctrine and stare decisis principles . . . .”).

  65. Dickerson, supra note 18, at 228–29 (referring to “plain meaning” as a canon, rule, and doctrine, interchangeably).

  66. Krishnakumar, Holy Trinity, supra note 8, at 1095–96 (“The Court routinely, for example, relies on the ‘dog that didn’t bark’ canon, the expressio unius canon, and the whole act rule without identifying the canon on which it is relying or even necessarily indicating that the argument it is making is based on a canon.”); see DHS v. Regents of Univ. of Cal., 140 S. Ct. 1891, 1923 (2020) (Thomas, J., dissenting) (arguing that Congress knows how to enact a law when it wants to, invoking, but not naming, the dog that didn’t bark canon).

  67. This conclusion, that identifying canons serves important practical purposes, is not universally held. See John Paul Stevens, Foreward to Eskridge, Interpreting Law, supra note 15, at v–vii, (in which Justice Stevens doubts whether “any special purpose is served by describing . . . as 'canons” the interpretive principles Eskridge articulates).

  68. George A. Kennedy, The Origin and Concept of a Canon and Its Application to the Greek and Latin Classics, in Canon Versus Culture: Reflections on the Current Debate 105, 105 (Jan Gorak ed., 2001) (“The authority of these oral canons is regularly supported by a claim of divine inspiration and by the usage of archaic or formal language that seems to certify an ancient origin.”).

  69. Id. at 106 (“[L]ibrarians and scholars in Alexandria [made] selective lists of the poets they judged most deserving to be edited and studied.”).

  70. Jan Gorak, The Making of the Modern Canon: Genesis and Crisis of Literary Idea 16 (2013).

  71. Stewart v. Kahn, 78 U.S. 493, 507 (1870).

  72. Kennedy, supra note 68.

  73. Id.

  74. Id. (“The language of canonical writings is also often archaic or a formal language distinct from what is commonly spoken at the time of composition, and after a few generations it may not be easily understood without special study.”).

  75. Scalia, supra note 11, at 26 (noting that he “find[s] it hard to believe anyone could criticize” the Latin canons).

  76. Kennedy, supra note 68, at 107 (explaining that the “[u]se of ‘canon’ to mean the books of the Bible accepted by Christians as genuine and inspired” dates to the fourth century).

  77. Gorak, supra note 70, at 19 (noting that in contemporary usage “canon . . . refers to the set of sacred texts a particular religious group accepts as permanently recording truths revealed to it by God”); Kennedy, supra note 68 (“The authority of these oral canons is regularly supported by a claim of divine inspiration . . . .”).

  78. Scalia, supra note 11, at 26.

  79. Smith v. Me. Bureau of Revenue Servs., 590 B.R. 1, 11 (D. Me.), aff’d sub nom. In re Smith, 910 F.3d 576 (1st Cir. 2018).

  80. Conn. Nat’l Bank v. Germain, 503 U.S. 249, 253 (1992).

  81. Blake A. Watson, Liberal Construction of CERCLA Under the Remedial Purpose Canon: Have the Lower Courts Taken a Good Thing Too Far?, 20 Harv. Env’t L. Rev. 199, 209–11 (1996) (arguing that during the nineteenth and early twentieth centuries, interpreters placed “almost unquestioning reliance” on the canons).

  82. E.g., Gluck, supra note 8, at 182 (identifying the “CBO Canon”); Heinzerling, supra note 8, at 1944 (identifying the “power” canons); Krishnakumar, Holy Trinity, supra note 8, at 1064–66 (identifying the “unique national institution” canon); Mendelson, supra note 8, at 110–15 (discussing the newly identified “no elephants in mouseholes,” “location of codification,” “veteran’s benefits,” and “jurisdictional rules” canons); see also Krishnakumar, Reconsidering Substantive Canons, supra note 8, at 855–56 (noting frequency with which the Court recognizes new canons); Rosenkranz, supra note 8, at 2143 (“The Supreme Court may . . . foreshadow[] a new canon with insinuating dicta.”).

  83. Sometimes, only a few Supreme Court cases are enough to launch a new canon. See Michael Coenen & Seth Davis, Minor Courts, Major Questions, 70 Vand. L. Rev. 777, 828 (2017) (arguing that lower courts should not apply the “major questions” exception to the Chevron doctrine without guidance from the Supreme Court).

  84. Specifically, Gorsuch rejected the following argument: Congress prohibited discrimination based on “sex”; Congress did not prohibit discrimination based on “sexual orientation”; therefore, even though “sexual orientation” falls within the general scope of “sex,” Congress’s silence excepted sexual orientation from the scope of the generally applicable term, sex. Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1747 (2020).

  85. Id. (defining the “canon of donut holes” as when “Congress’s failure to speak directly to a specific case that falls within a more general statutory rule creates a tacit exception”).

  86. Bond v. United States, 572 U.S. 844, 857 (2014) (“[W]e presume, absent a clear statement from Congress, that federal statutes do not apply outside the United States.”). For a description of the Court’s changing doctrine on the presumption against extraterritoriality, see William S. Dodge, The New Presumption Against Extraterritoriality, 133 Harv. L. Rev. 1582, 1639 (2020).

  87. E.g., Vt. Agency of Nat. Res. v. United States ex rel. Stevens, 529 U.S. 765, 779–80 (2000) (“We must apply to this text our longstanding interpretive presumption that ‘person’ does not include the sovereign.”).

  88. Llewellyn, supra note 57, at 400.

  89. Scalia, supra note 11, at 27. Of course, there are many cases that affirm that courts should interpret statutory language in light of its purpose. Bob Jones Univ. v. United States, 461 U.S. 574, 586 (1983) (“It is a well-established canon of statutory construction that a court should go beyond the literal language of a statute if reliance on that language would defeat the plain purpose of the statute.”).

  90. This confusion is reminiscent of the confusion that often surrounds debates over legal positivism versus natural law. See generally H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593, 625 (1958) (arguing that natural law criticisms of legal positivism are sometimes opaque).

  91. Dickerson, supra note 18, at 228.

  92. Scalia & Garner, supra note 11, at 9; Baude & Sachs, supra note 20, at 1126; William N. Eskridge, Jr., The New Textualism and Normative Canons, 113 Colum. L. Rev. 531, 544 (2013); Heinzerling, supra note 8, at 1944 (“It may well be that valid or at least normatively unproblematic interpretive canons are usually distinguished by their antiquity.”); see also J.E.M. Ag. Supply v. Pioneer Hi-Bred Int’l, Inc., 534 U.S. 124, 146 (2001) (Scalia, J., concurring) (arguing that the case pits “two perfectly valid canons of interpretation” against each other).

  93. Asplundh Tree Expert Co. v. NLRB, 365 F.3d 168, 173 (3d Cir. 2004) (“This canon of construction is a valid approach whereby unexpressed congressional intent may be ascertained.”); Owens v. Balt. City State’s Att’ys’ Off., 767 F.3d 379, 411 (4th Cir. 2014) (referring to a “valid canon of construction”).

  94. Baude & Sachs, supra note 20, at 1121.

  95. Scalia, supra note 11, at 27 (“There are a number of other faux canons in Llewellyn’s list.”).

  96. Scalia & Garner, supra note 11, at 341–411. Nourse characterized Scalia and Garner’s list of “falsities” as “anti-canons.” Nourse, supra note 49, at 103.

  97. Easterbrook, supra note 22, at 1405 (arguing that the canon of constitutional avoidance is “wholly illegitimate”).

  98. Duquesne Light Holdings, Inc. & Subsidiaries v. Comm’r, 861 F.3d 396, 408 (3d Cir. 2017) (rejecting reliance on a particular canon, which “cannot be regarded as a legitimate canon of estate tax interpretation”); cf. Cia. Petrolera Caribe, Inc. v. Arco Caribbean, Inc., 754 F.2d 404, 428–29 (1st Cir. 1985) (explaining that a particular canon “is therefore an especially reliable and legitimate canon of construction”).

  99. Keen v. Helson, 930 F.3d 799, 805 (6th Cir. 2019).

  100. Barrett, supra note 11, at 169.

  101. Cia. Petrolera Caribe, Inc., 754 F.2d at 429 (“[I]t is therefore an especially reliable and legitimate canon of construction.”).

  102. Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2130 (2016) (Thomas, J., dissenting) (“But the distributive canon is less helpful in cases such as this . . . .”); United States v. Johnson, 961 F.3d 181, 188 (2d Cir. 2020) (“[T]he ‘nearest reasonable referent’ canon is less helpful than . . . the common-sense linguistic principles the canon is meant to embody.”).

  103. State Farm Mut. Auto. Ins. Co. v. Comm’r, 698 F.3d 357, 370 (7th Cir. 2012) (“One of the more reliable canons of statutory construction—the normal practice—is that a term or phrase is ordinarily given the same meaning throughout a statute.”).

  104. United States v. Councilman, 418 F.3d 67, 74 (1st Cir. 2005) (“[T]he canon may be a less reliable guide to Congressional intent.”).

  105. Eskridge, supra note 19, at 1490 n.41 (“[F]or it rests upon the highly unreliable maxim of statutory construction ‘inclusio unius est exclusio alterius.’”).

  106. Ohio v. U.S. Dep’t of the Interior, 880 F.2d 432, 458 (D.C. Cir. 1989) (“[I]ndeed, a canon of equal worth with the acquiescence-by-reenactment rule is the one disfavoring repeal by implication.”).

  107. Einer Elhauge, Statutory Default Rules: How to Interpret Unclear Legislation 193 (2008) (noting that some canons “are often used as makeweight, to support conclusions otherwise merited on independent grounds”); Dickerson, supra note 18, at 234 (arguing that the Latin canons “masquerade as rules of interpretation while doing nothing more than describing results reached by other means”); Llewellyn, supra note 57, at 401 (setting out the now-famous thrusts and parries); Greenawalt, supra note 52, at 118 (“[J]udges who are actually deciding in a particular way because they favor the policy it promotes may choose to put forward a canon that supports their conclusion, portraying it as much more decisive than it really is for them.”); William N. Eskridge, Jr., Dynamic Statutory Interpretation 275 (Harvard Univ. Press ed. 1994) (noting that the legal realists considered the canons “window dressing”).

  108. Rosenkranz, supra note 8, at 2154 (rejecting the “hyperrealist account” that the Court reaches results irrespective of interpretive principles as “a caricature” and arguing that an interpretive rule can make a difference in “countless cases”); Sunstein, supra note 58, at 452 (“[S]ome of the canons actually influenced judicial behavior insofar as they reflected background norms that helped to give meaning to statutory words or to resolve hard cases.”); Krishnakumar, Reconsidering Substantive Canons, supra note 8, at 894–95 (tentatively staking out an intermediate position between hard formalism and legal realism).

  109. Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1747 (2020).

  110. Indeed, as noted above, interpretive principles generally considered canons are quite similar to the principle rejected by the Court. E.g., Vt. Agency of Nat. Res. v. United States ex rel. Stevens, 529 U.S. 765, 780–81 (2000) (“We must apply to this text our longstanding interpretive presumption that ‘person’ does not include the sovereign.”).

  111. Yates v. United States, 574 U.S. 528, 543–44 (2015).

  112. Id. at 531–32.

  113. Id. at 539–40.

  114. Id. at 541.

  115. Id. at 539, 539 n.3, 541.

  116. Id. at 559–60 (noting Scalia and Garner’s list of canons).

  117. Id.

  118. Ransom v. FIA Card Servs., 562 U.S. 61, 81 (2011) (Scalia, J., dissenting) (opining that there is no “canon against verbosity”).

  119. In re Pierce, 115 B.R. 523, 525 n.3 (Bankr. N.D. Tex. 1990) (“Administrative convenience, however, is not a canon of statutory construction.”).

  120. FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 190 (2000) (Breyer, J., dissenting) (doubting the existence of the major question canon); Aposhian v. Wilkinson, 989 F.3d 890, 897 n.4 (10th Cir. 2021) (Tymkovich, C.J., dissenting) (“But Chevron is neither a standard of review nor a canon of construction.”); In re Cuozzo Speed Techs., 778 F.3d 1271 (Fed. Cir. 2015) (Newman, J., dissenting) (opining that a court-created doctrine is not a canon of construction and so does not need to be followed like one); Motion Picture Ass’n of Am., Inc. v. Oman, 969 F.2d 1154, 1157 (D.C. Cir. 1992) (“A discretionary power emanating from the federal courts is not a canon of construction.”); People v. Sumner, 69 Cal. Rptr. 15, 19 (Ct. App. 1968) (noting that a rule that governs the exercise of discretion is considered “a ‘rule of convenience’” rather than a canon-limiting power (first quoting Hibernia Sav. & Loan Soc’y v. Doran, 118 P. 526, 526 (Cal. 1911); and then quoting Shank v. Blackburn, 215 P. 559, 560 (Cal. Dist. Ct. App. 1923))); Cortesy v. Territory, 30 P. 947, 951 (N.M. 1892) (rejecting an argument that would allow alcohol sales on Sunday because there “is not a canon of statutory construction in the books which ever contemplated such an outrage upon common sense”).

  121. This would have been the case in Bostock. There, the majority’s case would have been stronger had it demonstrated that the putative canon of donut holes is not a canon. And a test for spotting canons would have enabled the majority to make this argument expressly. Conversely, a test for spotting canons would have given the dissenters a basis for disputing the claim that it does not exist. See Bostock v. Clayton, Cnty., 140 S. Ct. 1731, 1747 (2020) (“Congress’s failure to speak directly to a specific case that falls within a more general statutory rule creates a tacit exception.”).

  122. Richard J. Lazarus, Advocacy Matters Before and Within the Supreme Court: Transforming the Court by Transforming the Bar, 96 Geo. L.J. 1487, 1563 (2008) (“Better, more effective advocates influence the development of the law . . . .”); see also Brian Sheppard & Andrew Moshirnia, For the Sake of Argument: A Behavioral Analysis of Whether and How Legal Argument Matters in Decisionmaking, 40 Fla. St. U. L. Rev. 537, 591–93 (2013) (“Argument, therefore, seems to be of greater importance when a case involves a rule rather than a standard.”).

  123. E.g., Gluck & Bressman, supra note 6, at 906 (finding that some statutory drafters “were not only aware of some of the interpretive rules that courts employ . . . but told us that these legal rules affect how they draft”).

  124. Id. at 952 (describing canons with a “dialogic or teaching function”); Scalia & Garner, supra note 11, at 52 (“The canons fulfill another purpose—a self-fulfilling purpose, so to speak: They promote clearer drafting.”).

  125. Cross, supra note 14, at 85 (noting that textualism, legislative history, and pragmatism “could theoretically rely on canons of construction or reject them”); Manning, supra note 14, at 284 (“[A] large and growing number of academics (and academics‑turned‑judges) now believe in the utility of canons of construction . . . [and] the newly faithful cover a broad philosophical spectrum.”).

  126. Gluck, supra note 63, at 2056 (“Justice Scalia’s aim was to bring rules, objectivity, and a disciplined approach to statutory cases.”); Ofer Raban, Between Formalism and Conservatism: The Resurgent Legal Formalism of the Roberts Court, 8 N.Y.U. J.L. & Liberty 343, 345 (2014) (“[T]extualism . . . seeks to reduce legal interpretation, respectively, to an exercise in following semantics and syntax . . . .”).

  127. Grove, supra note 48, at 279 (identifying different competing versions of textualism); Evan C. Zoldan, The Conversation Canon, 110 Ky. L.J. (forthcoming 2022) (manuscript at 2–3) (identifying competing visions of textualism and ordinary meaning).

  128. Jonathan T. Molot, Ambivalence About Formalism, 93 Va. L. Rev. 1, 8–9 (2007) (noting the argument that “purposivist judges may exercise interpretive leeway in a manner that reflects their own political biases”); Raban, supra note 126, at 368 (noting that textualism is the result of the formalist claim that judges must “adopt judicial methodologies that prevent them from injecting their value judgments into their judicial decision-making”); Gluck, supra note 63, at 2056 (“Justice Scalia’s aim was to bring rules, objectivity, and a disciplined approach to statutory cases.”).

  129. E.g., Nourse, supra note 49, at 106 (arguing that “‘canon textualism,’ a form of textualism seeking to supplant legislative evidence with canons of construction . . . encourages judicial activism”).

  130. Id. at 106–07 (describing “petty textualism” as picking and choosing text to create a “false plainness” or “false ambiguity”).

  131. Id. at 106.

  132. Id. at 107.

  133. The ordinary meaning of statutory language is far from obvious: it is more akin to a legal conclusion than to a fact that can be discovered. See Eskridge, supra note 106, at 290 (“[W]hat is or is not a plain meaning is to a large extent constructed, not found, by the Court.”); see also Arthur L. Corbin, The Interpretation of Words and the Parole Evidence Rule, 50 Cornell L.Q. 161, 164 (1965) (“[W]hen a judge refuses to consider relevant extrinsic evidence on the ground that the meaning of written words is to him plain and clear, his decision is formed by and wholly based upon the completely extrinsic evidence of his own personal education and experience.”).

  134. The Supreme Court and lower courts routinely rely on dictionaries to define statutory terms. See, e.g., Taniguchi v. Kan Pac. Saipan, Ltd., 566 U.S. 560, 566–67 (2012) (relying on approximately ten different dictionaries to define a single statutory term); see also Barrett, supra note 45, at 2203 (“Dictionaries are useful to the textualist not because the textualist assumes that legislators use them but because they offer some evidence of the meaning attributed to words by ordinary English speakers.”). Dictionary use has been persuasively critiqued. See, e.g., James J. Brudney & Lawrence Baum, Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras, 55 Wm. & Mary L. Rev. 483, 490–91, 566 (2013); Brian Slocum, Ordinary Meaning 217–21 (2018) (suggesting that the use of dictionaries by the Supreme Court reflects a desire to appear neutral); Pamela Hobbs, Defining the Law: (Mis)using the Dictionary to Decide Cases, 13 Discourse Stud. 327, 332 (2011); see also United States v. Costello, 666 F.3d 1040, 1043–44 (7th Cir. 2012).

  135. Whitfield v. United States, 574 U.S. 265, 268 (2015) (relying on the language used in novels to interpret statutory language). In other work, I have critiqued the use of the language in novels to interpret statutory language. Zoldan, supra note 127 (manuscript at 48, 51).

  136. Muscarello v. United States, 524 U.S. 125, 128–30 (1998) (relying on the language used in newspaper articles to interpret statutes).

  137. For a description of corpus linguistics, see Graeme Kennedy, An Introduction to Corpus Linguistics 1 (1998) and Krzyztof Kredens & Malcolm Coulthard, Corpus Linguistics in Authorship Identification, in The Oxford Handbook of Language and Law 504, 504–505 (Peter M. Tiersma & Lawrence M. Solan eds., 2012), noting various definitions of “corpus” and “corpus linguistics.” For recent judicial uses of corpus linguistics techniques, see, for example, Am. Bankers Ass’n v. Nat’l Credit Union Admin., 306 F. Supp. 3d 44, 68 (D.D.C. 2018); People v. Harris, 885 N.W.2d 832, 833–34 (Mich. 2016); In re Adoption of Baby E.Z., 266 P.3d 702, 726 (Utah 2011) (Lee, J., concurring); Caesars Ent. Corp. v. Int’l Union of Operating Eng’rs, 932 F.3d 91, 95–96 (3d Cir. 2019); and Wilson v. Safelite Grp., Inc., 930 F.3d 429, 440–44 (6th Cir. 2019) (Thapar, J., concurring). For a critique of the use of corpus linguistics techniques for statutory interpretation, see generally Evan C. Zoldan, Corpus Linguistics and the Dream of Objectivity, 50 Seton Hall L. Rev. 401, 446 (2019).

  138. Nourse, supra note 49, at 104 (noting that the resurgence of canon use is largely driven by the fact that textualists are often the strongest “canon enthusiasts”). For more on the connection between textualists and the canons, see Scalia, supra note 11, at 25–27, articulating a strong support for judicial reliance on the canons; Scalia & Garner, supra note 11, at 7–9, explaining and advocating for the use of valid canons; Manning, supra note 11, at 705 n.140, stating, “[T]extualist judges also use canons of construction, which routinely form part of the context used to interpret statutes.”; Barrett, supra note 11, at 121, noting, “[Textualists] are probably the biggest proponents of linguistic canons. Textualists, however, also embrace substantive canons.”; Brudney & Ditslear, supra note 11, at 4, stating, “Judges, especially textualist-oriented judges, praise the canons for their relative clarity and commonsense virtues.”; and Gorsuch, supra note 50, at 132, stating, “Maybe the most prominent interpretive tools used by textualists are the so-called ‘canons of construction.’”

  139. Molot, supra note 128, at 45–46 (“The semantic context that textualists favor includes dictionary definitions, the use of identical language in other statutory provisions, and ‘textual’ or ‘linguistic’ canons of construction that have nothing to do with statutory purposes or societal effects.”).

  140. Mendelson, supra note 8, at 111 (arguing that interpretive principles should be considered canons when they apply across different statutory settings).

  141. Eskridge, supra note 15, at 18, 19, 25 (noting that some substantive canons can be justified on normative grounds; that is, they “give vitality to unenforced constitutional norms,” like “nondelegation, federalism, and even some individual rights” that ultimately make the system more responsive and democratic); Gluck & Bressman, supra note 6, at 926 (“[Some] canons make under-the-radar constitutional law or advance underenforced (judicial or societal) policy preferences.” (footnotes omitted)); cf. Nelson, supra note 11, at 383–85 (noting some canons can be used to approximate legislative intent).

  142. As others have described at length, canon use can certainly be used by interpreters to import their preferences into their interpretations. E.g., Llewellyn, supra note 57, at 401–03 (describing how canons can be deployed to counteract one another); Gluck, supra note 8, at 179 (“[Canons] are too numerous to be predictably chosen; there is no ranking among them; and they are not treated as black-letter, precedential law . . . . They thus find little justification in their potential to advance a formalist, rule‑of‑law vision.”).

  143. Notable exceptions include the interpretive principles identified by Eskridge as “extrinsic legislative sources.” Eskridge, supra note 15, at 422–25. These interpretive principles include direction to consider legislative history, including conference committee explanations, committee reports, rejected proposals, sponsor’s statements, and other legislative evidence of statutory purpose. Id.

  144. See William N. Eskridge, Jr. & Philip P. Frickey, Quasi-Constitutional Law: Clear Statement Rules as Constitutional Lawmaking, 45 Vand. L. Rev. 593, 595–96 (1992) (noting that the linguistic canons are policy-neutral); Larry M. Eig, Cong. Rsch. Serv., Statutory Interpretation: General Principles and Recent Trends 2 (2014) (“When reading statutory text, the Supreme Court uses content-neutral canons developed by the judiciary that focus on word usage, grammar, syntax and the like.”).

  145. As scholars have demonstrated, however, the canons do not appear to live up to their reputation for imposing significant constraints on judicial discretion. Brudney & Ditslear, supra note 11, at 6 (“Our findings indicate that canon usage by Justices identified as liberals tends to be linked to liberal outcomes, and canon reliance by conservative Justices to be associated with conservative outcomes.”).

  146. Krishnakumar, supra note 2, at 1277 (“[A] discernible decline in the rate at which the Court invokes legislative history, a marked increase in its use of dictionary definitions to interpret statutes, and a rise in its use of both linguistic and substantive canons of construction.” (footnotes omitted)); Raban, supra note 126, at 390 (“This formalistic resurgence has been in the making for decades, and there is little surprise in its clear upswing under the conservative Roberts Court: after all, legal formalism has been almost uniquely a conservative project.”).

  147. E.g., Jonathan R. Siegel, The Legacy of Justice Scalia and His Textualist Ideal, 85 Geo. Wash. L. Rev. 857, 860 (2017) (“[T]he Supreme Court has never accepted Justice Scalia’s textualist ideal. The Court has always looked to other methods of statutory interpretation.”).

  148. Gluck, supra note 8, at 179 (“The rules that judges employ are too numerous to be predictably chosen; there is no ranking among them; and they are not treated as black‑letter, precedential law—that is, the same interpretive rules do not apply to the same questions from case to case—as a formalist approach should logically require.”).

  149. Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 412 n.29 (1971).

  150. See Siegel, supra note 147, at 858 (“[Justice Scalia] made textualism one of his signature issues and waged a sustained, decades-long campaign to promote it.”).

  151. Nevertheless, textualist judges may rely on purpose covertly. Krishnakumar, supra note 2, at 1280 (“[A]lthough the textualist Justices do not explicitly invoke statutory purpose or intent at high rates, they often employ practical reasoning, language canons, and the whole act rule in ways that impute purpose and intent to Congress.” (footnotes omitted)).

  152. King v. Burwell, 135 S. Ct. 2480, 2496 (2015) (“[A] fair reading of legislation demands a fair understanding of the legislative plan . . . . Section 36B can fairly be read consistent with what we see as Congress’s plan, and that is the reading we adopt.”).

  153. McGirt v. Oklahoma, 140 S. Ct. 2452, 2464 (2020).

  154. Guerrero-Lasprilla v. Barr, 140 S. Ct. 1062, 1072 (2020); see also Sturgeon v. Frost, 139 S. Ct. 1066, 1085 (2019) (“The legislative history (for those who consider it) confirms, with unusual clarity, all we have said so far.”); Return Mail, Inc. v. U.S. Postal Serv., 139 S. Ct. 1853, 1865 n.8 (2019) (“Moreover, for those of us who consider legislative history . . . .”). In this same vein, I would consider Justice Kagan’s statement that the Justices “are all textualists now” rhetorical rather than descriptively accurate. Harvard Law School, The 2015 Scalia Lecture: A Dialogue with Justice Elena Kagan on the Reading of Statutes, YouTube (Nov. 25, 2015), https://www.youtube.com/watch?v=dpEtszFT0Tg&t=1s [https://perma.cc/5DSP-4EKZ]. It is clear that Justice Kagan was not suggesting that she followed the textualist ideal of excluding legislative history from the permissible tools of interpretation. Instead, her statement is better read as an acknowledgment of the great weight that judges now feel compelled to place on text when interpreting statutes.

  155. Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1825, 1827 (2020) (Kavanaugh, J., dissenting).

  156. Id. at 1827, 1834 (“Justice Scalia explained the extraordinary importance of hewing to the ordinary meaning of a phrase: ‘Adhering to the fair meaning of the text (the textualist’s touchstone) does not limit one to the hyperliteral meaning of each word in the text.’” (emphasis omitted) (quoting Scalia & Garner, supra note 11, at 356)).

  157. Gorsuch, supra note 50, at 128–29.

  158. Id. at 131–32.

  159. Barrett, supra note 11, at 121 (“Textualists routinely bring canons to bear on the interpretation of statutes. They are probably the biggest proponents of linguistic canons. Textualists, however, also embrace substantive canons . . . .” (footnote omitted)); id. at 177 (“When employed, however, to stretch plain language . . . [substantive canons] conflict with the obligation of faithful agency.”).

  160. Van Buren v. United States, 141 S. Ct. 1648, 1654 (2021) (relying on several dictionaries to interpret statutory language); see also Barrett, supra note 45, at 2203 (arguing that linguistic canons are legitimated by the fact that they “are designed to capture the speech patterns of ordinary English speakers and, in some cases, of the subclass of lawyers”).

  161. It has been criticized, however. Eskridge, supra note 19, at 1490 n.41 (discussing “the highly unreliable maxim of statutory construction ‘inclusio unius est exclusio alterius’”).

  162. Bell v. United States, 349 U.S. 81, 83 (1955) (“When Congress leaves to the Judiciary the task of imputing to Congress an undeclared will, the ambiguity should be resolved in favor of lenity.”).

  163. Eskridge, supra note 92, at 544. Indeed, in his foreword to Eskridge’s book, Interpreting Law, Justice Stevens doubts whether “any special purpose is served by describing . . . as ‘canons’” the interpretive principles articulated by Eskridge. Stevens, supra note 67, at vii.

  164. See generally Eskridge, supra note 15, app.

  165. See Scalia & Garner, supra note 11, at xi–xvi, 9 (referring to some established canons as invalid). Justice Kagan appears to prefer the Scalia and Garner treatise, as evidenced by her discussion of the treatise in Yates v. United States, 574 U.S. 528, 557, 559–60 (2015) (Kagan, J., dissenting); cf. Gluck, supra note 63, at 2062 (“But the Scalia catalogue unquestionably leaves many canons out, without explanation.”).

  166. See generally Norman J. Singer & J.D. Shambie Singer, Statutes and Statutory Construction, Part V (7th ed. 2012) and Shambie Singer, Statutes and Statutory Construction, Part V (8th ed. 2020), for descriptions of approximately 150 canons.

  167. See Gluck, supra note 63, at 2061 (“There are certainly more than seventy interpretive presumptions and, really, there are probably closer to one hundred.” (emphasis omitted)).

  168. Gluck, supra note 8, at 188.

  169. Stack, supra note 30, at 285 (“This Article’s core argument is that this principle—which I call the enacted purposes canon—is and should be viewed as a bedrock principle of statutory interpretation.”).

  170. Heinzerling, supra note 8, at 1944.

  171. Krishnakumar, Holy Trinity, supra note 8, at 1094–95.

  172. Leib & Brudney, supra note 33, at 736 (“We propose . . . a ‘belt-and-suspenders’ canon . . . .”).

  173. E.g., Mendelson, supra note 8, at 111–15 (describing several new canons); Byington, supra note 34, at 133–34 (introducing the fresh start canon); see also Krishnakumar, Reconsidering Substantive Canons, supra note 8, at 855–56; Rosenkranz, supra note 8, at 2143 (“The Supreme Court may . . . foreshadow[] a new canon with insinuating dicta.”).

  174. See Bruhl, supra note 2, at 522–24, 524 fig.3 (tracking canon usage in the Supreme Court and lower courts).

  175. Mendelson, supra note 8, at 78.

  176. Aaron-Andrew P. Bruhl, The Jurisdiction Canon, 70 Vand. L. Rev. 499, 502–03 (2017) (arguing that the jurisdiction canon “has not (yet) been forced into retirement” but at least demoted “to part-time status”).

  177. Dodge, supra note 86, at 1636 (“Despite the theoretical problems with changing canons, it is inevitable that canons of interpretation will change. The presumption against extraterritoriality is just one example of this phenomenon.”).

  178. See e.g., Mendelson, supra note 8, at 111 (“To make this list, the purported canon had to take a rule-like form—to be articulated as an interpretive principle applicable across a range of statutory settings—and had to have been applied repeatedly.”); Heinzerling, supra note 8, at 1944 (“By using the term ‘canon,’ I mean simply to refer to the way I believe the Supreme Court has treated the power canons—as rules or principles of interpretation.”); Dodge, supra note 86, at 1645 (“When the Supreme Court changes a canon of statutory interpretation, it ought to justify the new canon in normative terms. The point may be obvious, but if the new canon is not an improvement, then the change is not worth making.”).

  179. Scalia, supra note 11, at 26–27 (suggesting that the test of a canon is whether he has “heard of it”); see also Scalia & Garner, supra note 11, at 174 (justifying the surplusage canon as “sensible”); Singer & Singer, supra note 166, § 47:16 (referring to noscitur a sociis as a “commonsense canon”); Cross, supra note 14, at 86. As Gluck has noted, Justice Scalia’s list of canons “unquestionably leaves many canons out, without explanation.” Gluck, supra note 63, at 2062 (referring to the list of canons in the Scalia and Garner treatise).

  180. See Moglen & Pierce, supra note 17, at 1226–27 (noting that some of Sunstein’s proposed canons are inconsistent with current law); Krishnakumar & Nourse, supra note 17, at 174–75, 177 (noting that some of Eskridge’s canons are “aspirational,” that is, normatively attractive but not currently embraced by the Supreme Court). Krishnakumar and Nourse themselves blur the descriptive and normative by asserting that “the basic thread connecting the canons is (or should be) established convention.” Krishnakumar & Nourse, supra note 17, at 188 (emphasis added) (emphasis omitted).

  181. Krishnakumar & Nourse, supra note 17, at 189 (“[T]he ultimate test of a canon . . . reflects the agreement of Supreme Court Justices appointed by different parties and across ideological divides.” (emphasis omitted)).

  182. Id. at 182–90 (identifying and evaluating criteria for canonicity).

  183. Id. at 188 (emphasis omitted).

  184. Michael Sinclair, "Only a Sith Thinks Like That": Llewellyn’s “Dueling Canons,” One to Seven, 50 N.Y.L. Sch. L. Rev. 919, 922 (2006) (emphasis omitted).

  185. Id. at 992.

  186. Eskridge, supra note 15, at 20 (“That the Supreme Court and state high courts cite them pervasively means that lower court judges as well as private attorneys need to pay close attention to the canons.”); id. app. & nn.1–305 (citing Supreme Court usage as evidence of canonical status); see also Krishnakumar & Nourse, supra note 17, at 188 (focusing on use by Supreme Court for identifying canons).

  187. Krishnakumar & Nourse, supra note 17, at 182–83 (highlighting the Court’s special role in setting trends of statutory interpretation).

  188. Eskridge, supra note 15, at 21 (“[T]he common law of statutory interpretation entails legislative and administrative as well as judicial practice at both the state and federal levels.”).

  189. Byington, supra note 34, at 133 (relying on circuit court opinions to establish the existence of a “fresh start” canon); Heinzerling, supra note 8, at 1944 (considering the power canons canonical in part because of lower court recognition).

  190. Krishnakumar & Nourse, supra note 17, at 182 (agreeing with commentators who “have taken for granted that citation by the U.S. Supreme Court should be the yardstick by which frequency of use of a particular canon is measured”).

  191. Id. at 183 (discounting the utility of relying on lower courts because they tend to be idiosyncratic and not necessarily transferrable to other jurisdictions).

  192. Caseloads: Supreme Court of the United States, Method of Disposition, 1970-2016, Fed. Jud. Ctr., https://www.fjc.gov/history/courts/caseloads-supreme-court-united-states-method-disposition-1970-2016 [https://perma.cc/3WRM-MYMV] (last visited Oct. 5, 2021) (providing data of case disposition); John G. Roberts, Jr., 2020 Year-End Report on the Federal Judiciary app. at 5 (2020), https://www.supremecourt.gov/publicinfo/year-end/2020year-endreport.pdf [https://perma.cc/355J-SN9U]; John G. Roberts, Jr., 2019 Year‑End Report on the Federal Judiciary app. at 5 (2019) [hereinafter Roberts, 2019 Report], https://www.supremecourt.gov/publicinfo/year-end/2019year-endreport.pdf [https://perma.cc/LQK8-DJCR].

  193. See Bruhl, supra note 43, at 30 (estimating that, in the past decade, district courts have issued more than 100,000 unpublished decisions that involve statutory interpretation). To take another metric, district courts resolve 60,000 to 70,000 criminal cases per year; although some of these are resolved on constitutional grounds, the underlying claims by the United States are statutory. Caseloads: Criminal Cases, 1870‑2017, Fed. Jud. Ctr., https://www.fjc.gov/history/courts/caseloads-criminal-cases-1870-2017 [https://perma.cc/79NJ-S3JP] (last visited Oct. 5, 2021); see also Roberts, 2019 Report, supra note 192, app. at 5–6 (providing information about federal caseloads); Caseloads: U.S. Courts of Appeals, 1892-2017, Fed. Jud. Ctr., https://www.fjc.gov/history/courts/caseloads-us-courts-appeals-1892-2017 [https://perma.cc/EF3V-S87H] (last visited Oct. 5, 2021) (providing additional information about federal caseloads).

  194. Mashaw, supra note 7, at 526 (“[Agencies] publish thousands of rules and formal opinions every year.”). Because of the depth and frequency of agency interpretations of statutes, federal administrative agencies have been described as the “primary official interpreters of federal statutes.” Id. at 502–03.

  195. State Court Caseload Statistics, Bureau of Just. Stat., https://www.bjs.gov/index.cfm?ty=tp&tid=30 [https://perma.cc/SRT4-R32Z] (last visited Oct. 5, 2021) (providing caseload information).

  196. Bruhl, supra note 43, at 29 (“Especially in the lower courts—and in the district courts most of all—where caseloads are high and ideological stakes are usually low, canons and other legalistic tools probably exert significant influence on decisions.”).

  197. Id.

  198. Id.

  199. Abbe R. Gluck, The States as Laboratories of Statutory Interpretation: Methodological Consensus and the New Modified Textualism, 119 Yale L.J. 1750, 1777–78, 1838 (2010) (noting that states prefer a modified textualism that includes reference to legislative history, unlike textualists in the vein of Justice Scalia).

  200. Mashaw, supra note 7, at 531–35 (noting stark differences between agency statutory interpretation and courts’ review of those interpretations); see also Walker, supra note 7, at 1030 (noting divergence in dictionary use between agency interpreters and the Supreme Court). The differences in citation practices result from agencies’ radically different relationship to statutes; agencies are dependent on the legislature in a way that federal courts are not. Mashaw, supra note 7, at 505 (“While American courts must somehow balance their position as faithful agents of the Constitution and Congress, agencies seem to enjoy a less conflicted constitutional role: they are the executors of the President’s constitutional responsibility to ‘take Care that the Laws be faithfully executed.’”); Matthew C. Stephenson, Statutory Interpretation by Agencies, in Research Handbook on Public Choice and Public Law 285, 285 (Daniel A. Farber & Anne Joseph O’Connell eds., 2010) (“[A]gencies are not fully autonomous when they make their interpretive decisions. Other political institutions—Congress, the President, and the courts—exert a powerful influence over how agencies interpret statutes.”).

  201. Krishnakumar & Nourse, supra note 17, at 183–84 (discounting the utility of relying on lower courts because they tend to be idiosyncratic and not necessarily transferrable to other jurisdictions).

  202. Eskridge, supra note 15, at 21 (“[T]he canons are also America’s common law of statutory interpretation—the nation’s ongoing experience with applying statutes to concrete circumstances.”); Baude & Sachs, supra note 20, at 1122 (“[T]he canons stand on their own authority as a form of common law.”).

  203. Eskridge, supra note 15, at 21; cf. Abbe R. Gluck, The Federal Common Law of Statutory Interpretation: Erie for the Age of Statutes, 54 Wm. & Mary L. Rev. 753, 757 (2013) (“There is also the possibility that some of the canons might be federal common law, while others might not.”).

  204. See Dodge, supra note 86, at 1639 (noting that canons can change over time).

  205. The adoption of the doctrine of res ipsa loquitur is instructive. Introduced first in England in 1863, the doctrine spread to the United States and was rapidly adopted by state court judges throughout the country by the beginning of the twentieth century. Jeffrey H. Kahn & John E. Lopatka, Res Ipsa Loquitur: Reducing Confusion or Creating Bias?, 108 Ky. L.J. 239, 260 (2019) (noting that by 1905, the doctrine of res ipsa loquitur had spread rapidly throughout the United States).

  206. Gluck, supra note 199, at 1834–35 (describing modified textualism).

  207. For a critique of this practice, see Zoldan, supra note 137, at 445–47, arguing that corpus linguistics fails to live up to its promise of leading the interpreter to objective interpretations.

  208. For a description of corpus linguistics, see Kennedy, supra note 137, at 1 and Kredens & Coulthard, supra note 137, at 504–05, describing different definitions of “corpus” and “corpus linguistics.”

  209. E.g., In re Adoption of Baby E.Z., 266 P.3d 702, 726–28 (Utah 2011) (Lee, J., concurring) (advocating the adoption of corpus linguistics to interpret statutes).

  210. E.g., People v. Harris, 885 N.W.2d 832, 833–34, 838–39, 849–51 (Mich. 2016) (relying on corpus linguistics techniques to interpret the statutory term “information”).

  211. See Am. Bankers Ass’n v. Nat’l Credit Union Admin., 306 F. Supp. 3d 44, 68 (D.D.C. 2018) (discussing corpus linguistics techniques); Caesars Ent. Corp. v. Int’l Union of Operating Eng’rs Loc. 68 Pension Fund, 932 F.3d 91, 95–96 (3d Cir. 2019) (same); Wilson v. Safelite Grp., Inc., 930 F.3d 429, 440–42 (6th Cir. 2019) (Thapar, J., concurring) (same); Carpenter v. United States, 138 S. Ct. 2206, 2238 n.4 (2018) (Thomas, J., dissenting) (describing corpus linguistics data).

  212. Krishnakumar & Nourse, supra note 17, at 174.

  213. The exclusion of putative canons that have been introduced only by scholars does not exclude canons simply because they have not yet been named by a court. Indeed, courts often rely on interpretive principles without specifically naming them. Krishnakumar, Holy Trinity, supra note 8, at 1095–97; Krishnakumar & Nourse, supra note 17, at 190; Heinzerling, supra note 8, at 1944.

  214. Krishnakumar & Nourse, supra note 17, at 183–84.

  215. Id.; see also Eskridge, supra note 15, app. at 438 (describing an interpretive principle that applies only to the Bankruptcy Reform Act of 1978).

  216. For example, consider the major questions doctrine, which has been used only rarely by the Supreme Court. King v. Burwell, 135 S. Ct. 2480, 2488–89 (2015) (“Whether those credits are available on Federal Exchanges is thus a question of deep ‘economic and political significance’ that is central to this statutory scheme; had Congress wished to assign that question to an agency, it surely would have done so expressly.” (quoting Util. Air Regul. Grp. v. EPA, 573 U.S. 302, 324 (2014))).

  217. See, for example, the various Bostock opinions, all of which claimed to be applying the “ordinary meaning” of the statute without agreeing on a method for determining what the ordinary meaning is. Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1738–39, 1741 (2020); id. at 1767 (Alito, J., dissenting); id. at 1833 (Kavanaugh, J., dissenting).

  218. Ruggero J. Aldisert, Winning on Appeal: Better Briefs and Oral Argument 140 (2d ed. 2003) (describing the phenomenon of throwing arguments “against the wall” to “see what will stick”); see also 2