In Search of Ordinary Meaning: What Can Be Learned from the Textualist Opinions of Bostock v. Clayton County?

Introduction

What is the meaning of the phrase “discriminate because of sex”? This was the key question the Supreme Court faced in Bostock v. Clayton County.1 The case involved the firing of two gay individuals and one transgender individual, and the Court’s answer to the question solidified Title VII’s protections to gay and transgender individuals against employment discrimination. No longer can an employer fire an individual for being gay or transgender.2

To reach this determination, Justice Neil Gorsuch’s majority opinion utilized the principles of textualism. Like any good textualist, Justice Gorsuch relied on the ordinary public meaning of the statute and refused to consider extratextual sources—such as legislative history—when the express terms of the statute gave the Court “one answer.”3 To reach his one answer, Justice Gorsuch started with dictionary definitions of individual words in a phrase, supported these definitions with the help of precedent, and then combined the meanings of the individual words to find the ordinary meaning of the phrase. The result is that Title VII’s use of the phrase “discriminate because of sex”4 means that “[a]n employer violates Title VII when it intentionally fires an individual based in part on sex.”5 Using this definition of the phrase, Justice Gorsuch concluded that Title VII protects individuals from discrimination based on their sexual orientation and gender identity.6

But Justice Gorsuch is not the only textualist on the highest court in the land. Justice Brett Kavanaugh, in dissent, agreed that the Court needed to look to the ordinary meaning of “discriminate because of sex” but disagreed with how the majority reached its one answer. While it may be literally true that the phrase could encompass sexual-orientation or gender-identity discrimination, Justice Kavanaugh contended that the majority’s reading of the statute did not comport with its ordinary meaning.7 Ordinary public meaning is the “conventional meaning of the utterance [ ] understood by ordinary but linguistically proficient speakers at the time of the utterance.”8 Justice Kavanaugh argued that when the majority relied on the “strung-together definitions of the individual words in the phrase,” it committed itself to the literal meaning of the phrase.9 The majority’s process of stringing together dictionary definitions to reach the ordinary meaning of a phrase is similar to what linguists call semantics. In semantics, “meanings of parts compositionally determine the meanings of wholes.”10 This is not how mainstream textualists determine ordinary meaning. It “misses the forest for the trees.”11 Instead, a judge should assess the “ordinary meaning of the phrase as a whole.”12 This process is similar to what linguists call pragmatics. Pragmatics is “sometimes characterized as dealing with the effects of context.”13

Textualism has become a common denominator among Supreme Court Justices, but there is disagreement over what it means to be a textualist.14 Textualism started as a methodology that only conservative judges employed. Since Justice Antonin Scalia joined the Court, the textualist movement has grown, now permeating the circuit courts. Textualism has escaped ideological boundaries and become routine in judicial arguments—regardless of political bent. For example, Justice Elena Kagan has remarked, “[W]e’re all textualists now.”15 However, Bostock indicates that textualism is not a clearly defined philosophy. Justices Gorsuch and Kavanaugh both subscribe to textualism, both claim to adhere to the text of Title VII, and both argue that the other is attempting to subvert the role of the legislature. Although Justice Gorsuch’s reliance on solely the words of the statute was the first step in being a good textualist, his reliance on semantics to find the ordinary meaning of a phrase diverged from mainstream textualism. Prominent textualists have denounced the stringing together of definitions as dealing in “a sterile literalism.”16 Despite such an expressed preference for pragmatics textualism, case law and empirical evidence may provide some support for Justice Gorsuch’s approach. As textualism has become more popular, the use of dictionaries and the stringing together of definitions has increased, suggesting that mainstream textualist judges often stray from pragmatics textualism. These judges do not practice what they preach.

The problem with straying from a pragmatics textualism in favor of semantics textualism is that a “phrase may have a more precise or confined meaning than” the semantic meaning.17 This more precise meaning comes from context. Context can be derived from knowing when a statement is made or who made the statement.18 Important to this Comment, a specific combination of words can also provide relevant societal context. For example, Chief Justice John Roberts has written that “two words together may assume a more particular meaning than those words in isolation. We understand a golden cup to be a cup made of or resembling gold. A golden boy, on the other hand, is one who is charming, lucky, and talented.”19

Bostock also shows that textualism’s lack of clearly defined requirements has consequences. Because Justice Gorsuch’s approach to textualism won the day, gay and transgender people enjoy the benefits of Title VII protection—benefits that Justice Kavanaugh argues should have come from the legislature. Although Justice Gorsuch won the battle, will Justice Kavanaugh win the war? It is possible that this case will represent the genesis of two branches of textualism. Instead of a unified theory of textualism, judges will either adhere to the semantics or pragmatics school of textualism. It is equally possible that one of these two theories will fall by the wayside.

Bearing in mind the importance of defining what it means to be a textualist, this Comment offers a sliding-scale approach that reconciles Justice Gorsuch’s semantics approach with Justice Kavanaugh’s pragmatics approach. The size of the pool of precedent is the primary variable that dictates the extent to which a judge uses a semantics or pragmatics approach. This sliding-scale approach assesses the pool of available precedent to determine whether a judge can rely solely on semantics textualism to develop the ordinary meaning of a phrase. A judge should use semantics to construct a phrase with definitions derived from precedent. I call this a precedent-based-semantics approach. When there is variance in how precedent has defined the individual words of the phrase (i.e., the pool is large), it becomes necessary to support the precedent-based-semantics meaning with pragmatics-based arguments. Relying on a judge’s determination of the variability of the precedent creates a new concern. Judges who like the precedent-based-semantics meaning will say the pool of precedent is small, and judges who do not like the precedent-based-semantics meaning will say the pool of precedent is large. To assuage this concern, I construct tiers that approximate the pool of precedent. The two factors that go into this approximation for the pool of precedent are (1) the court that created the precedent and (2) the statute that the precedent interpreted—similar to how the U.S. court system already approaches precedent.

In Bostock, Justice Gorsuch correctly relied on an entirely precedent-based-semantics approach because the pool of Supreme Court precedent interpreting the words of Title VII is adequately small. In contrast, when the pool of precedent is large, the need to support a precedent-based-semantics meaning with pragmatics comes into play. For instance, had Justice Gorsuch looked to how the Supreme Court defined words in other statutes to determine the ordinary meaning of the words of Title VII, he would have needed to address whether this meaning makes sense in the context of Title VII. The Supreme Court has likely considered the term “because of” many times and produced a variety of potential definitions. Therefore, Justice Gorsuch would have had to support his selection of one of the potential definitions with an examination of context. This context could be an examination of the rest of the statute: Does Bostock’s interpretation of the word “sex” also make sense in Title IX? Or this context could be an examination of historical or societal context: Does Bostock’s use of the word “sex” make sense with how normal people used it in 1964? Had Justice Gorsuch looked to how lower courts or state courts defined the words to determine the ordinary meaning of the words of Title VII, the sliding-scale approach would call for both forms of context.

My sliding-scale approach advances a new middle ground between absolute semantics and absolute pragmatics. I argue this middle ground is preferable to the absolutes because of its ability to eliminate value-based judgments. Eliminating value-based judgments is the stated goal of textualists because it increases the predictability of statutory interpretation, which in turn promotes the rule of law and democratic accountability.

The sliding-scale approach’s ability to limit value-based judgments is premised on two considerations: the benefits to predictability from statutory stare decisis and the relatively limited ability to cherry-pick terms defined in precedent. Statutory stare decisis is the accepted, strong presumption that a court’s precedent related to statutory interpretation is correct. Unlike constitutional stare decisis, there is general acceptance of statutory stare decisis even among textualists.20 Creating predictable law and promoting judicial restraint are two rationales supporting statutory stare decisis21 —similar rationales for textualism.22 Cherry-picking occurs when there is an abundance of options to make an argument, allowing a judge to choose the option that best comports with her value preferences. If there is an abundance of precedent defining the same words in different ways, then a precedent-based-semantics approach allows for cherry-picking. This problem also arises with the use of dictionaries. There are enough dictionaries—and dictionary definitions—that, by picking among the plethora of options, a judge could find support for almost any possible argument. My sliding-scale approach seeks to restrain judges by limiting the ability to cherry-pick. To avoid cherry-picking, pragmatics comes in to check the precedent-based-semantics meaning, limiting precedents’ ability to support value-based judgments. But to fully utilize precedent, semantics must play a role. In developing my argument for the sliding-scale approach, I offer an analysis of how textualists have approached semantics and whether the Court has tacitly recognized a distinction between dictionary-defined terms and precedent-defined terms.

This Comment proceeds in three parts. Part I provides an overview of textualism, including its history and how it relates to the Bostock opinion. Part I.A surveys “New Textualist” thought. These are the tenets about which there is little disagreement among textualists. Part I.B then looks outside of Bostock and briefly describes other methods that textualists have used to find ordinary meaning. Part II examines the major disagreement within the textualist methodology relevant to Bostock: the proper role of semantics. To that effect, Part II.A and Part II.B provide an in-depth analysis of the Bostock case, focusing on Justice Gorsuch’s majority opinion and Justice Kavanaugh’s dissent. After framing the debate between Justices Gorsuch and Kavanaugh as one of semantics versus pragmatics, Part II.C explores the validity of a precedent-based-semantics approach. Left without a clear answer between Justice Gorsuch and Justice Kavanaugh, Part III uses the guidepost of restraining judges to examine the method that textualists ought to embrace. Part III.A presents a novel sliding-scale approach that incorporates a precedent-based-semantics approach similar to Justice Gorsuch’s approach in Bostock, but it also gives credence to Justice Kavanaugh’s arguments in support of a pragmatics-based approach. Part III.B provides further justifications in support of the sliding-scale approach. Part III.C addresses the counterarguments that arise from relying on alternative methods when there is a truly novel statute. Finally, Part III.D uses the concept of judicial restraint to critique the alternative methods introduced in Part I.B.

I. Laying the Groundwork: What Does It Mean to Be a Textualist?

There are parts of the textualist theory that share universal agreement among textualist judges, and there are parts that leave room for debate. To establish a baseline for what it means to be a textualist, Part I.A lays out the concepts that unite all practitioners of “New Textualism”—a form of textualism that refuses to look to legislative history. Part I.B introduces the methods that have been suggested by commentators on textualism for finding ordinary meaning.

A. New Textualism

Despite their differing opinions in Bostock, Justice Gorsuch, Justice Kavanaugh, and Justice Samuel Alito all purport to subscribe to a form of textualism known as New Textualism.23 Justice Scalia and Judge Frank Easterbrook popularized this interpretive methodology, and both made efforts to clearly articulate its foundational elements.24 Central to New Textualism is the goal for a judge to be a “faithful agent” of Congress—“when a statutory text is clear, that is the end of the matter.”25 This method contravenes the beliefs of purposivists who “maintain that judges are partners in governance and ought to consider that role when they apply statutes to new circumstances.”26 The goal of New Textualism explains why the majority in Bostock stressed that it was simply following the law that the legislature established in the text and why the dissents claimed that the majority was improperly legislating. To take part in any lawmaking is anathema to a textualist’s goal of being a faithful agent of Congress. Instead, a textualist operating as a faithful agent wants to remain constrained to the text, limiting a judge’s ability to read her own values into a statute.27 Keeping the judge’s values out of a statute “provide[s] greater certainty in the law, and hence greater predictability and greater respect for the rule of law.”28

In pursuit of being faithful agents, textualists have examined government behavior to further the argument that a judge should not move past a clear text. Textualists “maintain . . . that variance between a clear text and its apparent purpose does not show that Congress . . . poorly communicated its intent.”29 Instead, textualists abstain from using purpose to impute intent, offering three arguments that aim to show that the text is the best reflection of the legislature’s desires.30 First, the effect of compromising multiple interests to form the text of a statute means that legislators may not “pursue a statute’s background purpose to its logical end.”31 Promoting the ultimate purpose over the text of a statute ignores that the text represents what Congress could get passed through the legislative process.32 Second, textualists contend that it is nearly impossible to construct a collective intent from the aggregate of legislators’ individual preferences because “legislative outcomes frequently turn on non-substantive factors, such as the sequence of alternatives presented (agenda manipulation) or the practice of strategic voting (logrolling).”33 Similar to the first point, pursuing the purpose of a law at the expense of the text ignores political realities, and because the text of a law is not dependent only on purpose but also on strategic behavior, making this tradeoff calls into question a judge’s ability to impute legislative intent when interpreting a statute.34 Third, enforcing purpose and not text may undermine Congress’s desire to use a rule rather than a standard to achieve its goals.35 As Justice Scalia put it, judges “are bound, not only by the ultimate purposes Congress has selected, but by the means it has deemed appropriate, and prescribed, for the pursuit of those purposes.”36 Therefore, if a judge pursues purpose over text, then that judge may be undermining the legislature’s desire to restrict the means that a judge can use to accomplish the goals of a statute. These arguments show that purpose-based arguments impute ‘“intent’ that ultimately can be found only in the mind of the judge.”37

Textualists are left with the words in the statute and its surrounding context to determine the dictates of a particular law. Judge Easterbrook favorably quotes Justice Oliver Wendell Holmes, Jr., proclaiming that “Holmes could say in 1899 that ‘We do not inquire what the legislature meant; we ask only what the statute means.’ He was denying that original intent, as opposed to the original meaning, mattered.”38 Accordingly, textualists do not rely on external forms of evidence that serve to illuminate the purpose of the statute—namely, legislative history.39

A final point of agreement is that textualists do not rely on the literal meaning of the text but rather the ordinary meaning. This point is seemingly the center of Justice Kavanaugh’s main argument. Justice Gorsuch’s “one answer” for the text may be literally correct, but it is substituting out ordinary meaning in favor of literal meaning.40 This is a problem if Justice Gorsuch is truly practicing textualism. Textualists have been clear that literalism is not the way to determine the meaning of the statute.41 The proper way for a textualist to interpret statutes is to ask “how a reasonable person, conversant with the relevant social and linguistic conventions, would read the text in context.”42 Successfully engaging in this exercise allows one to determine the ordinary meaning of a phrase. The ordinary meaning of the phrase is important because it is (by definition) the “most accessible to the citizenry desirous of following the law.”43 “This approach recognizes that the literal or dictionary definitions of words will often fail to account for settled nuances or background conventions that qualify the literal meaning of language.”44 When a judge deviates from the ordinary meaning in favor of the literal meaning, she abandons what most people think the law means, “depriv[ing] the citizenry of fair notice of what the law is.”45

New Textualists agree that judges should rely on the ordinary meaning of statutory test, but they disagree on the appropriate methods to find it. “Textualists believe that legislation supposes that legislators and judges are part of a common social and linguistic community, with shared conventions for communication. Accordingly, they argue that a faithful agent’s job is to decode legislative instructions according to the common social and linguistic conventions shared by the relevant community.”46 However, there is no consensus on how to discern these social and linguistic conventions. Justice Scalia suggested using “valid canons” of statutory interpretation.47 In Bostock, Justice Gorsuch focused on precedent and the dictionary.48 This approach contrasted with Justices Kavanaugh and Alito’s consideration of the history surrounding Title VII, common parlance, and how government entities have used the words elsewhere.49 Judge Easterbrook has a similar approach to Justices Kavanaugh and Alito, adhering to common parlance. Judge Easterbrook argues that the meaning should be derived from how an objectively reasonable person would read the words.50 In practice, judges have not declared a clear winner.51

B. Alternative Methods for Finding Ordinary Meaning

Among textualist judges, some have explicitly offered methods to determine the ordinary meaning of phrases. The purpose of this Section is to briefly introduce these methods. The shortcomings of these methods will then be discussed in Part III.D.

1. Justice Kavanaugh’s “best reading” approach.

Justice Kavanaugh has offered what he calls the “best reading” approach.52 The first step in the best-reading approach is to abandon the threshold determination of ambiguity.53 Other interpretative methods—and some practitioners of textualism—require judges to find the text ambiguous before using extratextual sources or canons of interpretation. Justice Kavanaugh asserts that these ambiguity thresholds create an opportunity for value judgments to creep into a decision.54 Judge Easterbrook has expressed a similar reluctance to rely on ambiguity thresholds. The problem is that an ambiguity determination is unlikely to restrain because the court is the one that gets to “choose when to declare the language of the statute ‘ambiguous,’” as “[t]here is no metric for clarity.”55 This lack of an objective standard allows a political or subjective determination as to whether to move past the text.56 This apparent problem can be seen between Bostock’s majority opinion and Justice Alito’s dissent. Justice Gorsuch states that the “express terms of [the] statute give [the Court] one answer.”57 Justice Alito finds this argument arrogant and considers unambiguity to be a high standard.58 These statements do little to move the ball. Which side a reader comes down on is less likely to be tied to how ambiguous the text is than the values or purposes promoted in each opinion.

Justice Kavanaugh argues that a court should instead rely on “(1) the words themselves, (2) the context of the whole statute, and (3) any other applicable semantic canons.”59 Justice Kavanaugh defines semantic canons as “the general rules by which we understand the English language.”60 Ultimately, these steps amount to the equivalent of finding the literal meaning of the statute.61 Then, a judge should “apply—openly and honestly—any substantive canons (such as plain statement rules or the absurdity doctrine) that may justify departure from the text.”62 Justice Kavanaugh focuses on the absurdity doctrine,63 which will be discussed further in Part III.D.

2. Justice Scalia’s valid-canons approach.

Justice Scalia has spoken in vague terms of using context but has presented a list of interpretative canons to maintain predictability. Textualists have consistently said context is important to determine the meaning of a statute, and to determine the appropriate context, a judge “should look at the statutory structure and hear the words as they would sound in the mind of a skilled, objectively reasonable user of words.”64 Justice Scalia has promoted the same idea of utilizing a reasonableness inquiry: “[T]he acid test of whether a word [or phrase] can reasonably bear a particular meaning is whether you could use the word [or phrase] in that sense at a cocktail party without having people look at you funny.”65 However, an abstract determination of what is reasonable is not a limiting guideline. Fortunately, Justice Scalia has provided some greater insight as to what can be used to determine when cocktail-party attendees will look at you funny.66 Unfortunately, this insight has come in the form of fifty-seven interpretive canons, undermining the restraining power that textualism seeks to promote.67

3. Corpus linguistics.

Corpus linguistics represents a modern approach to finding ordinary meaning. Utah Supreme Court Associate Chief Justice Thomas Lee and Professor Stephen Mouritsen write that when finding ordinary meaning “we are asking an empirical question—about the sense of a word or phrase that is most likely implicated in a given linguistic context.”68 Considering the empirical nature of the question, they propose that it is best determined through a data-driven approach called corpus linguistics.69 This process looks for patterns of meaning or usage in “corpora” (i.e., large databases of naturally occurring language).70 Through their analysis, Chief Justice Lee and Mouritsen seek to define ordinary meaning and often consider it with respect to the “relative frequency of competing senses of a given term.”71 The level of frequency that equates to ordinary meaning is not answered.72 Still, a judge that relies on corpus linguistics has the tools to find the frequency of competing senses, accounting for “relevant semantic, pragmatic, temporal, and speech-community considerations.”73 Seemingly, corpus linguistics serves as an effective method for a textualist to consider all the potential factors that would allow her to find ordinary meaning.

II. Bostock v. Clayton County: Picking Between Semantics and Pragmatics

In Bostock, the Supreme Court held that Title VII protects gay and transgender people from employment discrimination.74 To reach this result, the Court engaged in a two-step process: (1) Find the ordinary meaning of the operative phrase and (2) apply the ordinary meaning to the facts of the case. At step one, the majority found the “express terms of [the] statute” provided “one answer.”75 It was therefore inappropriate to apply “extratextual considerations.”76 If the Court did rely on extratextual considerations, it would be going beyond a court’s requirement to apply the language of the law and would abandon the practice of judicial humility.77 Accordingly, the rule that emerged from the ordinary meaning of the phrase “discriminate because of sex” was that “[a]n employer violates Title VII when it intentionally fires an individual employee based in part on sex.”78 At the second step, the Court stated that “[a]n employer who fires an individual for being homosexual or transgender fires that person for traits or actions it would not have questioned in members of a different sex.”79 Consequently, discrimination because of sexual orientation or gender identity falls under Title VII’s prohibitions against discrimination because of sex.80 However, Justice Kavanaugh’s dissenting opinion suggested that the majority’s purposeful ignorance of some extratextual considerations—particularly the societal context of Title VII—means that the majority opinion is not aligned with the ordinary public meaning that it claims to represent.81 By ignoring the societal context, Justice Gorsuch committed himself to the literal meaning of the phrase. The problem with using the literal meaning is twofold: First, it is well established that textualists do not rely on the literal meaning of a statute.82 Second—and more fundamental—the majority’s decision to rely on the literal meaning of the statute allows the Court to legislate from the bench.83

But the distinction between literal and ordinary meaning is not the issue in this case.84 Justice Gorsuch was explicit that his goal was to find the “ordinary public meaning of [the statute’s] terms at the time of its enactment.”85 In pursuit of ordinary meaning, Justice Gorsuch referred to dictionary definitions of the words “sex,” “discriminate,” and “individual.”86 We can assume the dictionary definitions for these terms are aligned with their ordinary meaning because the Court sided with the plaintiffs but used a definition of “sex” the defendants provided, and there was no dispute over the definitions for “individual” or “discriminate.” Moreover, Justice Gorsuch did not rely solely on dictionaries. He also pieced together these definitions with precedent, such as the Court saying that “the ordinary meaning of ‘because of’ is ‘by reason of’ or ‘on account of.’”87 The real debate was whether it was appropriate to rely on semantics when attempting to find the ordinary meaning of the statute.88

This Part aims to weigh in on the semantics versus pragmatics debate. To do so, I first provide more depth to the debate between Justice Gorsuch and Justice Kavanaugh. In that vein, Part II.A summarizes Justice Gorsuch’s opinion. Part II.B summarizes Justice Kavanaugh’s opinion, highlighting how it is in fact an argument against the use of semantics to define the ordinary meaning of a phrase. Part II.C.1 adds to the textualist literature by focusing on the extent to which semantics is out of step with textualism. Finally, Part II.C.2 accepts that textualists express a general reluctance to rely on semantic reasoning but makes a novel contribution to textualist literature through a descriptive analysis of how textualists on the Supreme Court have relied on a precedent-based-semantics approach.

A. Justice Gorsuch’s Opinion

Justice Gorsuch started the majority opinion by stating that the Court’s process for statutory interpretation is to “interpret[ ] a statute in accord with the ordinary public meaning of its terms at the time of its enactment.”89 The Court, therefore, must determine the ordinary public meaning of Title VII’s command that it is “unlawful . . . for an employer to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex, or national origin.”90 The focus of the case was whether sexual orientation and gender identity fall under the statutorily protected characteristic of sex.91 However, both parties agreed, and the Court assumed, that “sex” refers “only to biological distinctions between male and female.”92 Justice Gorsuch was still left with the important task of defining and relating three other parts of the statute: (1) the phrase “because of,” (2) the word “discriminate,” and (3) the word “individual.”93

Justice Gorsuch determined that the definitions of these words and phrases through different means but ultimately concludes that these parts interact to make a straightforward rule. For the phrase “because of,” Justice Gorsuch relied on precedent providing that Title VII’s “because of” test is the traditional standard of but-for causation.94 According to Justice Gorsuch, a “but-for test directs [the Court] to change one thing at a time and see if the outcome changes. If it does, [the Court has] found a but-for cause.”95 This test is considered “sweeping” and allows for multiple but-for causes.96

However, the word “discriminate” serves to limit the application of the but-for test. Dictionaries and precedent related to how the Court has decided “disparate treatment” cases—the same type of case as Bostock—were considered authoritative. In these cases, the Court has held that Title VII’s use of the word “discriminate” means intentionally “treating [an] individual worse than others who are similarly situated.”97 Applying this definition to the case at hand, Justice Gorsuch concluded that “an employer who intentionally treats a person worse because of sex—such as by firing the person for actions or attributes it would tolerate in an individual of another sex—discriminates against that person in violation of Title VII.”98 Justice Gorsuch briefly entertained an alternative definition for “discriminate” offered in Justice Alito’s dissent that would have the statute “require [the Court] to consider the employer’s treatment of groups rather than individuals,” creating a law that “concerns itself simply with ensuring that employers don’t treat women generally less favorably than they do men.”99 This alternative was quickly dismissed because of the word “individual.” The statute “tells [the Court] three times . . . that [the Court’s] focus should be on individuals, not groups.”100 Having completed the final step of interpretation, Justice Gorsuch determined that a “straightforward rule emerges: An employer violates Title VII when it intentionally fires an individual employee based in part on sex.”101

The ordinary meaning of “discriminate because of sex” and the emergent rule need to be applied to the case; there still needs to be a determination whether discrimination because of sexual orientation or gender identity constitutes a case of but-for discrimination based on sex. Because “it is impossible to discriminate against a person for being homosexual or transgender without discriminating against that individual based on sex,” discrimination because of sexual orientation or gender identity is, in fact, a form of sex discrimination.102 In other words, the reason that discrimination because of sexual orientation or gender identity is a form of sex discrimination is because both sexual orientation and gender identity are “inextricably bound up with sex.”103 For example, firing a male employee for being gay is sex discrimination because the employer is firing the employee for being attracted to men, something that would not lead to a female employee being fired.104 This difference in treatment based on sex shows that sex is a but-for cause of the employer firing the employee and supports the conclusion that the employer violated Title VII.

B. Justice Kavanaugh’s Dissenting Opinion

Justice Kavanaugh accepted the majority’s premise that “firing someone because of their sexual orientation may, as a very literal matter, entail making a distinction based on sex,” but rejected that the Court must adhere to this literal meaning.105 Justice Kavanaugh stated that there is “no serious debate about the foundational interpretive principle that courts adhere to ordinary meaning, not literal meaning, when interpreting statutes.”106 Two purposes motivate adhering to the ordinary rather than literal meaning of statutory texts: the promotion of the rule of law and democratic accountability.107 The ordinary meaning is “most accessible to the citizenry desirous of following the law,”108 and to abandon the ordinary meaning for the literal meaning “deprives the citizenry of fair notice of what the law is.”109 This undermines the rule of law because “[a] society governed by the rule of law must have laws that are known and understandable to the citizenry.”110 Ordinary meaning is also most accessible to legislators, and departing from the ordinary meaning makes it more difficult for legislators to understand the meaning of the laws they enact.111 And in order to have democratic accountability, “[c]itizens and legislators must be able to ascertain the law by reading the words of the statute.”112 Therefore, relying on the “hidden or obscure [literal] interpretation of the law, and not its ordinary meaning,” causes the rule of law and democratic accountability to suffer.113

The problem with Justice Kavanaugh’s emphasis on distinguishing between literal and ordinary meaning is that Justice Gorsuch agreed with the premise that ordinary meaning trumps literal meaning. The majority opinion explicitly stated that the Court should normally interpret a statute “in accord with the ordinary public meaning of its terms at the time of its enactment.”114 The majority agreed with Justice Kavanaugh that departing from the ordinary meaning of a statute undermines the legislative process and people’s reliance on a meaning of the law that “they have counted on to settle their rights and obligations.”115 If Justice Gorsuch and Justice Kavanaugh both agreed that statutory text should be interpreted according to its ordinary meaning rather than its literal meaning, what is the dispute really about?

The real disagreement is about Justice Gorsuch’s methodology for finding the ordinary meaning of a phrase, and it is this that allegedly commits him to a literalist approach. It may be true that there is no difference between the ordinary and literal meaning of the word “sex.” But there may be a difference between the ordinary and literal meaning of the phrase “discriminate because of sex.” Justice Kavanaugh’s claim was that Justice Gorsuch’s stringing together of the separate meanings of “discriminate,” “because of,” “individual,” and “sex” is not the proper approach to determine the ordinary meaning of the phrase.116 To borrow from linguistics, Justice Gorsuch was engaging in something similar to semantics. Semantics is a logic-based approach to determining the meaning of a phrase. One must know only the meaning of the individual words of a phrase and the rules of syntax to determine the meaning of a phrase.117 After selecting what he considered to be the appropriate meaning for each word in the statutory phrase, Justice Gorsuch added them all together to define the meaning of the phrase.118 Justice Kavanaugh argued this is the incorrect approach. Instead, the unit of analysis to determine the ordinary meaning of a phrase should be the entire phrase and should rely on the context specific to that phrase.119 Linguists might treat this method of interpretation as closer to pragmatics. The problem with straying from pragmatics in favor of semantics is that a “phrase may have a more precise or confined meaning than the” semantic meaning.120 When courts ignore this fact, they “miss[ ] the forest for the trees.”121

After rejecting Justice Gorsuch’s approach to discerning the ordinary meaning of the phrase “discriminate because of sex,” Justice Kavanaugh derived ordinary meaning of the phrase from common parlance and common legal usage of the phrase as a whole. Justice Kavanaugh asserted that common parlance does not support the majority because the plaintiffs would “probably [ ] not tell their friends that they were fired because of their sex.”122 In other words, the plaintiffs would not use the phrase “discriminate because of sex” to describe what occurred to them. Instead, the plaintiffs may have used the phrase “discriminate because of sexual orientation (or gender identity).” Additionally, historical context is important to determine meaning because it provides insight into who is doing the speaking. The women’s rights movement that prompted Title VII’s enactment “was not (and is not) the gay rights movement.”123 Therefore, when Congress prohibited discrimination because of sex, it was doing so to remedy discrimination against women. This reality should inform how we read the operative phrase. Justice Kavanaugh then turned to the wording of federal laws, executive orders, federal regulations, state law, and Supreme Court precedent.124 In every context, sex is treated as a distinct concept from sexual orientation.125 This is particularly important in the context of federal laws because “the Court has often said, we ‘usually presume differences in language’ convey ‘differences in meaning.’”126 We are once again provided insight into who is doing the speaking—a Congress that knows the difference between discrimination because of sexual orientation and discrimination because of sex—and this should inform the Court’s reading of the phrase. Justice Kavanaugh concluded that “all of the usual indicators of ordinary meaning . . . overwhelmingly establish that sexual orientation discrimination is distinct from, and not a form of, sex discrimination.”127

C. Semantics Versus Pragmatics

Justice Kavanaugh’s problem with the majority opinion was that it strung together defined words to derive the meaning of an entire phrase. Justice Kavanaugh argued that a determination of ordinary meaning for a phrase must be done as a single unit, not an aggregation of individual parts.128 If the individual parts cannot be used to define the entire phrase, then there must be a reference to the broader context of the statute. This is a battle between semantics and pragmatics. These terms have started to enter the lexicon of the textualist community.129 But descriptive and prescriptive analysis of where textualists come down on semantics or pragmatics is limited. This Comment serves to fill this gap in textualist literature by examining textualist scholarship and case law, reaching an inference in support of pragmatics.130 This inference certainly calls into question Justice Gorsuch’s opinion. But this Comment continues to contribute to textualist literature by recognizing that Bostock’s reliance on precedent is an unmentioned but important distinction from the typical discussions regarding semantics. Ultimately, case law where textualists string together precedent-defined terms highlights a desire to unify text and precedent, which suggests that textualism has not abandoned semantics.

1. Semantics and the use of dictionaries.

Disparaging comments against the use of dictionaries call into question the legitimacy of a semantics approach. When a commentator or judge argues against a semantics approach, they describe the practice as stringing together dictionary definitions.131 Justice Scalia captures the concern associated with stringing together dictionary definitions when he says:

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. In the words of Learned Hand: “a sterile literalism . . . loses sight of the forest for the trees.” The full body of a text contains implications that can alter the literal meaning of individual words.132

Justice Scalia is not alone. Justice Kavanaugh’s dissent in Bostock argued that a semantics approach may miss nuances that qualify the application of a phrase.133 And Chief Justice Roberts has made a similar argument for reading words in context.134 This still leaves two questions: First, how does a textualist justify the increased use of dictionaries under a textualist judicial regime? Second, what happens when a court faces a situation like in Bostock? Supreme Court precedent has defined the words of the phrase “discriminate because of sex” in the context of Title VII. When considering a case brought under the same statute, would a textualist find it reasonable to be able to prescribe an entirely different rule with every permutation of words?

The Court’s practice suggests the role of dictionaries in determining ordinary meaning is unclear but potentially limited. Although Justice Scalia had been explicitly against too strongly adhering to dictionaries,135 the Court’s two main textualists of the twenty-first century—Justice Scalia himself and Justice Clarence Thomas—have used dictionaries the most.136 How does one justify this high usage with Justice Scalia’s warning that dictionaries and the literal meaning of words tend to lose the forest for the trees? Is there a pattern in Justice Scalia’s use of dictionaries that can provide a guide for when to deviate from the dictionary? If there is a pattern, it is not limited to using dictionaries to derive a semantic meaning of a phrase; Justice Scalia has, on more than one occasion, strung together dictionary definitions.137 However, Justice Scalia rarely stopped at the meaning constructed after stringing together dictionary definitions but instead proceeds to review historical sources, relationships with other laws, or precedent.138 Although never stated, Justice Scalia’s practice suggests that dictionary usage is a starting point to narrow potential definitions but there still needs to be more analysis to determine the appropriate definition.139 This observation suggests that Justice Kavanaugh was correct when he said, “[i]f the usual evidence indicates that a statutory phrase bears an ordinary meaning different from the literal strung-together definitions of the individual words in the phrase, we may not ignore or gloss over that discrepancy.”140 But this relegation of dictionary definitions to a class below “usual evidence” does not tell us if a semantics approach is off-limits for a textualist when the units that are strung together have more evidence to support that they represent the individual unit’s ordinary meaning.

2. A hint of acceptance for semantics.

Although dictionary definitions have limited application, Justice Gorsuch’s opinion complicates the legitimacy of semantics because of its reliance on definitions derived from precedent. Other commentators have supported or opposed Justice Gorsuch’s opinion on the basis of its linguistic techniques.141 This Comment makes the novel argument that missing from this analysis is a discussion of the interaction of precedent and semantics. Judges have long struggled to develop a unified theory of text and precedent.142 Justice Gorsuch’s stringing together of precedent-defined terms—what I call a precedent-based-semantics approach—can serve as a unifying methodology. This Comment shows that the underdiscussed role of precedent may suggest that a semantics argument reliant on the stringing together of precedent-defined terms should be treated differently than a semantic argument that is entirely reliant on dictionary definitions. Ultimately, Part III.A suggests that the restraining power of precedent is powerful enough to warrant a methodology that operationalizes the precedent-based-semantics approach.

Bostock’s majority certainly starts with dictionary definitions but ultimately relies on Supreme Court precedent that has previously defined the words “because of,” “discriminate,” and “individual” in Title VII. Indeed, precedent is one of the potential sources of ordinary meaning that has served as a way to either buttress or disregard the semantic meaning reached through the stringing together of dictionary definitions.143 However, this use of precedent is fundamentally different than in Bostock. Remember, the Bostock majority engaged in a two-step process: finding ordinary meaning and then applying ordinary meaning.144 Because semantics is a method to determine meaning and not implicated in step two, when using a precedent-based-semantics approach we should expect to see the use of precedent that engages with defining or articulating what the words of the statute mean.

Just because a case considered the application of a statutory phrase does not mean that it explicitly defined the meaning of said phrase. For example, the Court held in Oncale v. Sundowner Offshore Services, Inc.145 that the ordinary meaning of the phrase “discriminate because of sex” protected individuals from same-sex sexual harassment. Similarly, in Bostock, the Court held that the same phrase protected gay and transgender individuals from employment discrimination. These holdings do give a court some sense of the ordinary meaning of the phrase “discriminate because of sex.” However, Bostock engaged in the preliminary step of defining the ordinary public meaning of the phrase “discriminate because of sex,” establishing an understanding of the phrase that can be used going forward and without reference to the facts of Bostock. To determine the ordinary meaning of this phrase, the Bostock Court committed to using precedent that explicitly defined the words of the statute. This commitment to precedent-defined words suggests that no precedent existed defining the entire phrase. Otherwise, we would expect the Court to have cited that precedent and moved immediately to step two. Therefore, assuming a commitment to precedent-defined words, stringing together precedent-defined words is the best the Court can do.

Textualists on the Court have supported stringing together precedent-defined words. A recent example predating Justice Kavanaugh’s tenure on the Court is Justice Gorsuch’s opinion in Wisconsin Central Ltd. v. United States.146 Three railroads challenged a tax levied against employee stock options.147 The language of the Railroad Retirement Tax Act148 was central to the case—specifically, the phrase “money remuneration.”149 If the phrase money remuneration was broad enough to encompass stock options, then the three railroads would lose.150 To make this determination, Justice Gorsuch first split up the phrase and started the textual analysis with the dictionary definitions of “money” and “remuneration.” As in Bostock, Justice Gorsuch supported the dictionary-defined term with precedent that clarified the dictionary definition of the word “money.”151 There was no debate over the word “remuneration,”152 so Justice Gorsuch combined dictionary and precedent-defined meaning of “money” with the accepted meaning for “remuneration” to create the ultimate meaning of the phrase “money remuneration.”153 In the process of arguing for the semantic meaning, Justice Gorsuch explicitly rejected the dissent’s attempt to use a literal meaning for the word “money” that could encompass stock.154 While it might have been true that there was a literal, dictionary definition for money that could encompass stocks, precedent dictated that money had a narrower meaning.155 When this narrower meaning of money was used to modify remuneration, stock options did not fall under the ordinary meaning of money remuneration.156

Beyond the attempt at using literal meaning, the dissent offered a surplusage argument, stating that the majority’s understanding of the phrase would render other parts of the statute useless.157 Justice Kavanaugh offered a similar argument in his Bostock dissent,158 but Justice Gorsuch rejected it in Bostock and rejected it in Wisconsin Central too.159 Notably, Justice Thomas and Justice Alito signed onto Justice Gorsuch’s Wisconsin Central opinion.160

Another example can be found in Carcieri v. Salazar.161 This case concerned the Court’s interpretation of the Indian Reorganization Act (IRA).162 Under the IRA, the Secretary of the Interior was authorized “to acquire land and hold it in trust ‘for the purpose of providing land for Indians.’”163 Acting under the authority of this statute, the Secretary of the Interior accepted into trust a thirty-one-acre parcel for the benefit of the Narragansett Tribe.164 But the Court had to determine if the IRA’s definition of “Indian” extended to the Narragansett Tribe. The IRA defined “Indian” to “include all persons of Indian descent who are members of any recognized Indian tribe now under Federal jurisdiction.”165 Whether the Secretary of the Interior was authorized to hold the thirty-one-acre parcel in trust depended on the Court’s interpretation of the phrase “now under Federal jurisdiction.”166 The Narragansett Tribe was not under federal jurisdiction when the statute was enacted.167 Therefore, the question became whether the Secretary of the Interior’s authorization extended to Indian tribes that were under federal jurisdiction when the trust was accepted.168 Justice Thomas wrote the majority opinion and held that authorization was in fact restricted to Indian tribes under federal jurisdiction at the time of the statute’s enactment.169 The Court’s other main textualist, Justice Scalia, joined the majority.170 The Court focused its analysis on the ordinary meaning of the word “now.”171 To do so, Justice Thomas first relied on a dictionary definition of the word “now” and then Supreme Court precedent of the usage of “now” in different statutes.172 The resulting meaning was then tacked onto “under Federal jurisdiction” to determine the ordinary meaning of the phrase.

To be sure, neither opinion provides a definitive answer to the import of stringing together precedent-defined terms. They offer support for the use of a semantic meaning but both further their argument with an examination of the “broader statutory context”—a form of argument aligned with pragmatics.173 For example, in Wisconsin Central, Justice Gorsuch looked to a statute under the same title that “expressly treated ‘money’ and ‘stock’ as different things.”174 Congress’s explicit treatment of these two words as different concepts created a presumption that these two words have different meanings.175 Therefore, Justice Gorsuch provided additional support to the precedent-based-semantics meaning that indicated money remuneration did not extend to stock options. (It is worth pointing out that Justice Kavanaugh made a similar argument in his Bostock dissent.176 ) The precedent-based-semantics approach can serve as a starting point. But are there scenarios where a precedent-based-semantics meaning could and should stand alone?177

The ultimate conclusion of this Section is that, although not all textualists have uniformly endorsed either semantics or pragmatics, textualists view semantics—at the very least—as a valid approach up until some undefined point. While textualists have expressed concern with the semantics approach, when textualists interpret statutes, they tend to engage in semantics as well as pragmatics. This “do as I say, not as I do” approach is wanting and suggests that the history of textualism does not determine who is right between Justice Gorsuch and Justice Kavanaugh. Similar to how dictionaries work with the meaning of individual words, whether semantics should be abandoned for a pragmatics approach may depend on a case-by-case basis.178

III. Determining the Ordinary Meaning of Phrases: A Sliding Scale Between Semantics and Pragmatics

It should be clear by this point that a textualist’s goal is to find the ordinary meaning of a statute’s phrase. The problem is that Justices Gorsuch and Kavanaugh offer two opposing ways to find ordinary meaning—a semantics approach and a pragmatics approach. This Part proposes a solution that uses a sliding-scale approach to incorporate both semantics and pragmatics when finding ordinary meaning. My approach considers the restraining power of precedent and uses the potential pool of precedent as the metric to determine whether a semantics or pragmatics approach is needed. The pool of precedent is said to increase when there is variation in how precedent is defining the words of a phrase. As the pool increases, the ability for a precedent-based-semantics approach to restrain a judge is weakened, creating a need to supplement it with pragmatics-based arguments. The sliding-scale approach considers textualism’s two purposes of protecting the rule of law and democratic accountability.179 But the approach leans more on the principle that supports these two purposes—judicial restraint.180

This Part proceeds as follows: With restraint as the lodestar, Section A presents a novel solution of using a sliding-scale approach with the primary variable being the size of the pool of precedent. Section B presents additional justifications for why it is important to have an approach that systematically incorporates precedent, trading on the goals of textualism and its preference for rules. Section C addresses the counterarguments that a truly novel statute presents. Finally, Section D examines why other means of determining ordinary meaning are suboptimal.

A. A Sliding-Scale Approach to Using Precedent, Semantics, and Pragmatics

This Section seeks to present a novel solution that reconciles the debate between semantics and pragmatics by creating a sliding-scale approach that focuses on the restraining power of precedent. A semantics approach maximizes the restraining power of precedent. The alternative to semantics would be using precedent to reason backward to a sense of the ordinary meaning of a phrase. For example, in Oncale, the Court interpreted “discriminate because of sex” to protect individuals from same-sex sexual harassment.181 The result in Oncale gives other courts some sense of the ordinary meaning of the phrase, although there are many iterations of the ordinary meaning of the phrase that could lead to the protection against same-sex sexual harassment. A court tasked with finding the ordinary meaning of the phrase “discriminate because of sex” would have significant discretion to construct a meaning aligned with precedent and also aligned with the judge’s value-based preferences. A semantics approach to precedent, however, constrains by focusing on precedent’s treatment of words and not facts. “Because of” means but-for causation; that is how precedent has defined the terms and a judge would have no discretion on this matter. If the other words of the phrase also have established precedent, then a judge using a precedent-based-semantics approach can find the ordinary meaning of the phrase through a mechanical process devoid of discretion.

But when the pool of precedent is large, the precedent’s restraining power is diminished, and the need for a semantics approach goes with it. I suggest that a judge should still be willing to utilize a precedent-based-semantics approach, but, as the restraining power of precedent wanes, a judge should support the precedent-based-semantics approach with pragmatics-based arguments. The move to pragmatics is premised on a criticism of dictionary definitions and other methods for deriving ordinary meaning—too many options allow a judge to cherry-pick.182 The argument is that if the pool of potential arguments is large (i.e., there are many plausible dictionary definitions or interpretative canons), the judge is not restrained because she can choose a definition or canon that is aligned with her political beliefs. The restraint that the text provides is an illusion; instead, the text has been bent to the judge’s will.183 Precedent-defined meaning, on the other hand, is more difficult to cherry-pick. Precedent is similar to dictionaries in that it defines words, but precedent is better at clearly identifying the correct interpretation of a word in a statute. It is the goal of a judge to clearly articulate what a word means in a statute. Also, there are few cases defining the original meaning of a word in a statute, and they often reaffirm a meaning that has been previously stated.184 Therefore, when precedent exists, it is like each word in a phrase has a single dictionary definition, decreasing the ability to cherry-pick. To be sure, if there are many precedents providing many different definitions, we would be back to square one. When we see this variation in how precedent has defined the meaning of words, the size of the pool of precedent can be considered large. When the pool is large, pragmatics plays a role. Recall Justice Scalia’s use of dictionaries as a starting point.185 After establishing the meaning of a phrase by combining words from the dictionary, Justice Scalia would supplement his argument with an interrogation of the text and its context. In effect, Justice Scalia would start with semantics and then use pragmatics to support his semantics-based meaning. My sliding-scale approach requires a judge to do the same thing when the pool of precedent is large enough that cherry-picking is not limited. When the restraining ability of a precedent-based-semantics method is limited, its purpose then becomes to serve as a better starting point than dictionaries.

A problem with the sliding-scale approach is that a judge may have discretion to determine when the pool of precedent is sufficiently large to move past the precedent-based-semantics meaning. As an attempt to limit this unwanted discretion, I propose four discrete tiers to determine when a judge should support her precedent-based-semantics meaning with pragmatics. To develop these tiers, I use two considerations as proxies for the size of the pool of precedent: (1) the court that created the precedent and (2) the statute the precedent was interpreting. These proxies represent a commonsense approach to determine the size of the pool of precedent. The Supreme Court hears a relatively small number of cases a year, and even fewer Title VII cases. Accordingly, the pool of precedent interpreting Title VII is small enough to alleviate concerns of cherry-picking. It is not necessary to consider pragmatics-based arguments. However, there are ninety-four district courts hearing many thousands of cases a year.186 If a judge must look to precedent from district courts to create a precedent-based-semantics meaning for a phrase, then her ability to cherry-pick is quite significant. Therefore, it is necessary for a judge relying on district court precedent to support her argument with pragmatics.

I explain and offer cases that fall into the four major tiers, exemplifying how all courts might utilize my sliding-scale approach. I also make a more general suggestion about how the lower courts could further apply my sliding scale. The four tiers that apply to all courts are as followed: (1) Supreme Court precedent interpreting the words of the statute in question, (2) Supreme Court precedent defining similar words in other statutes, (3) courts other than the Supreme Court defining similar words, and (4) truly novel statutes.

1. The highest tier: Supreme Court precedent defining the words of the statute in question.

When there is Supreme Court precedent interpreting the statute in question, the courts should rely fully on a precedent-based-semantics approach. The force of a meaning derived from a precedent-based-semantics approach is strongest in these situations because the potential pool of precedent is likely small and the arguments supporting statutory stare decisis are easily applied.187 The Supreme Court has determined the ordinary meaning of the words of the phrase. These definitions would be the law of the land for the statute in question, and all courts will follow suit. The result is little to no variation in how the individual words of the statutory phrase are defined. This lack of variation leaves little room to inject value judgments into the legal analysis, promoting the major goals of textualism. Therefore, a judge adhering to a sliding-scale approach would follow Justice Gorsuch’s path in Bostock and stop after using a precedent-based-semantics approach.188 It is possible to present convincing alternative determinations of ordinary meaning based on hypothetical conversations with citizens from 1964 or inferences drawn from canons of interpretation.189 But abandoning the precedent-based-semantics approach in favor of a meaning that feels right provides greater opportunity for value judgments to seep in, undermining the ultimate pursuits of textualism. However, Justice Gorsuch’s approach is not a perfect analog for this tier because of its initial reliance on dictionary definitions. Using dictionary definitions is not necessarily bad for the argument, but it does little to limit the ability to cherry-pick.190

2. The middle tier: Supreme Court precedent defining similar words in other statutes.

When a court must rely on Supreme Court precedent that interprets similar language in other statutes, it becomes necessary for the court to support the precedent-based-semantics approach with minimal pragmatics-based arguments. Looking to Supreme Court precedent interpreting the words of other statutes will increase variation in how the words are defined. The Supreme Court’s definition of the word “quickly” in a financial statute will differ from its definition of the word “quickly” in an environmental statute. Therefore, when a judge has the choice to rely on either definition to construct her semantic meaning of a phrase comprising “quickly,” the risk of cherry-picking is heightened. That is why some pragmatics-based support is required.

Wisconsin Central and Carcieri exemplify the application of this middle tier. In both cases, the Court used a precedent-based-semantics approach where the precedent interpreted other statutes.191 The Court supplemented the precedent-based-semantics meaning with an examination of the broader statutory context—a limited application of pragmatics-based arguments.192 The continued reliance on precedent preserves predictability and uniformity, but there are more precedents to choose from when the Court must turn to other statutes. The consequence is an increased ability to cherry-pick. The ability to cherry-pick decreases restraint and means that semantics’ inability to capture all of the nuances of the text becomes more difficult to justify. But both cases show how the Court can mitigate this concern through a reliance on the broader statutory context. This increased use of the broader statutory context is the “slide” in the sliding-scale approach. Focusing on what the rest of the text can teach us starts to move into the realm of pragmatics.

At this tier, courts should supplement precedent with additional pragmatic tools. Carcieri serves as an example of how a judge should use the broader statutory context to support the precedent-based-semantics meaning. To support the precedent-defined meaning of “now,” Justice Thomas looked to the use of “now” in other parts of the text.193 Specifically, Justice Thomas rejected the argument that “now” means “now and hereafter” because the statute has expressly used the phrase “now or hereafter.”194 It would be difficult to explain Congress’s use of “now” alone to mean “now or hereafter” if other parts of the same statute explicitly use “now or hereafter.” Looking at other parts of the statute to determine the meaning of “now” is within the field of pragmatics, but it is a fairly limited use. There is no reliance on contemporaneous legislation. These sorts of techniques can enter the discussion only when the pool of precedent is large enough that cherry-picking is easy.

3. The lowest tier: courts other than the Supreme Court defining similar words in other statutes.

When courts resort to lower court precedent to use a precedent-based-semantics approach, pragmatics-based arguments take on a greater role in the analysis. There will be situations where there is insufficient relevant Supreme Court precedent to construct a semantic understanding of a phrase, but I argue that a court should continue to rely on precedent. The court should look to how other courts have interpreted the statute or similar language as a starting point. Certainly, this is where judicial restraint is at its weakest. The sheer number of courts defining the terms of statutes will create significant variation in how the terms are defined, increasing the pool of precedent. The larger pool presents an opportunity to cherry-pick definitions to create an ordinary meaning—potentially leading to a value-based result. There will still be minimal value attached to the uniformity and predictability associated with the practice. To make up for the larger pool of precedent, the judge will need to provide more pragmatics-based evidence to buttress the precedent-plus-semantic meaning.

Justice Kagan’s dissent in Yates v. United States195 exemplifies the lowest tier and illustrates how to use precedent and semantics as a starting point for finding the ordinary meaning of a phrase. (Justice Kagan’s dissent received support from textualists Justice Scalia and Justice Thomas,196 and Justice Kavanaugh has called the opinion “brilliant.”197 ) Yates was a commercial fisherman who caught undersized fish. “To prevent federal authorities from confirming that he had harvested undersized fish, Yates ordered a crew member to toss the suspect catch into the sea.”198 Because of this action, Yates was charged and convicted under § 1519 of the Sarbanes-Oxley Act of 2002.199 Section 1519 proscribes “knowingly alter[ing], destroy[ing], mutilate[ing], conceal[ing], cover[ing] up, falsif[ying], or mak[ing] a false entry in any record, document, or tangible object with the intent to impede, obstruct, or influence the investigation.”200 The case turned on whether the interpretation of the phrase “tangible object” encompassed the abandoned fish in a statute that prohibits tampering with evidence.201 The plurality and the concurrence narrowed the meaning of the phrase through the use of purpose-based arguments and interpretative canons.202 Justice Kagan granted to the other opinions that “sometimes . . . the dictionary definition of a disputed term cannot control. But this is not such an occasion, for here the text and its context point the same way.”203 The method that Justice Kagan used to determine the context of the phrase is similar to Justice Gorsuch’s approach in Bostock, diverging with respect to the nature of the precedent—Justice Kagan used state court precedent interpreting other statutes while Justice Gorsuch used Supreme Court precedent interpreting the statute in question. The first step was finding the ordinary meaning of “tangible object,” which is done through a dictionary definition and an acknowledgment of state court precedent.204 The next step was to look to the surrounding word “any.” Once again, Justice Kagan started with the dictionary definition and then the precedent.205 The word “any” expands the application of the phrase “tangible object,” and ignoring this would ignore the ordinary meaning of the text.206 This process resembles how Justice Gorsuch tacks on the ordinary meaning of the word “individual” to the phrase “discriminate because of sex” but with the opposite effect. The word “any” expands the meaning of “tangible object” and the word “individual” eliminates the possibility that the phrase “discriminate because of sex” calls for a group-based determination—an important factor for the result of Bostock’s majority.207 To be sure, it is not exactly parallel because the definition of “individual” was not from precedent but solely from the dictionary.

Justice Kagan also advanced arguments that look outside the surrounding text to determine the broader statutory context. Some of these arguments cut against Justice Gorsuch’s reasoning. Justice Kagan considered a hypothetical conversation with a neighbor and asks whether a neighbor would consider a fish to be a tangible object.208 Both Bostock dissents offered a similar argument when they presented the hypothetical that an ordinary citizen would disagree with the idea that “discriminate because of sex” encompasses discrimination on the basis of sexual orientation or gender identity.209 These hypothetical conversations tried to consider context beyond what the words of the statute provide and are seen as valid arguments in pragmatics. But this style of argument should be resorted to only when the pool of precedent is large. Therefore, under my sliding-scale approach, these arguments would not be effective against Justice Gorsuch’s reading of the statute in Bostock.

4. A truly novel statute: pragmatics is the only option.

When there is no relevant precedent for a statute, pragmatics should be a court’s only interpretative method. A precedent-based-semantics approach gets its restraint from the precedent. Therefore, when we are left with only the semantics approach, we are stringing together dictionary definitions. For the reasons expressed in Part II.B.1, this is a relatively poor option when compared to pragmatics. Therefore, semantics should be abandoned. There still remains a question of what pragmatics-based arguments should be preferred. It may be necessary for a judge to carefully consider which of the alternative methods best promotes judicial restraint.210

5. How lower courts can apply the sliding scale.

A final consideration for my sliding-scale approach is to determine how the structure of our court system can create additional tiers that approximate the variation of precedent. In lower courts, the focus remains on the balance between the benefits associated with a limited pool of precedent and the need for broader statutory context. How the sliding-scale approach may apply in circuit courts will serve as an example. For these courts, between the highest tier (situations where there is Supreme Court precedent interpreting the words of a statute in question) and the middle tier (where there is Supreme Court precedent interpreting other statutes) there is a tier where a circuit’s precedent interprets the statute in question. This situation arises when the Supreme Court has not yet defined a word within a statute but a circuit’s precedent has. Consider the “quickly” example with the modification that it is the Seventh Circuit, rather than the Supreme Court, interpreting a phrase in an immigration statute that includes the word “quickly.” Also consider if there were Seventh Circuit precedent that had defined “quickly” in the context of the same statute. Relying on the Seventh Circuit’s definition would require less pragmatics-based support than relying on Supreme Court precedent defining other statutes. There would be less variation in how the Seventh Circuit had defined the words of the statute in question when compared to how the Supreme Court had defined similar words in financial and environmental statutes. The decreased variation would mean a lesser concern for cherry-picking and a decreased need for broader statutory context.

For a similar reason, between the middle tier of Supreme Court precedent interpreting other statutes and the lowest tier of lower courts interpreting other statutes should be a tier of the circuit’s precedent interpreting other statutes. Imagine if the Seventh Circuit were defining the ordinary meaning of a phrase in an immigration statute that included the word “quickly” and looking to its own precedent defining the word “quickly” in other statutes. The variation in how the Seventh Circuit has defined the word “quickly” in all other statutes would be less than the variation of how all other courts have defined the word “quickly” in all other statutes. One could continue to break up the last tier into multiple tiers: other circuits interpreting other statutes, district courts interpreting other statutes, and state courts interpreting other statutes. If doing so, one should focus, for each tier, on how the size of the pool of precedent creates a need to consider pragmatics-based arguments.

B. Further Justifications for a Sliding-Scale Approach

Beyond precedent’s commonsense ability to limit cherry-picking, there are further justifications for the sliding-scale approach and its attachment to precedent. I will start with drawing a comparison to the arguments supporting statutory stare decisis—a “super-strong” presumption of correctness for statutory precedent211 —and textualism’s goals. Then, I will explain how the combination of this background and a semantics approach of constructing the meaning of phrases is aligned with textualism’s preference for creating and adhering to rules.

1. The natural connection between precedent and textualism.

Although textualism’s underlying principle is promoting judicial restraint, other goals associated with the doctrine are explicit justifications for statutory stare decisis. “[S]tatutory precedents are treated to a ‘super-strong’ presumption of correctness.”212 In defense of this presumption, Justice Louis Brandeis famously commented that “in most matters it is more important that the applicable rule of law be settled than that it be settled right.”213 Professor Frank Horack similarly said that if the Court decides to reverse its ruling on a statutory meaning, then “it is affirmatively changing an established rule of law under which society has been operating.”214 This characterization highlights the importance of predictability and its effect on the rule of law. Horack also states that the changing of the law—which now includes the Court’s decision—is “explicitly and unquestionably the exercise of a legislative function.”215

Despite the need to examine extratextual sources (i.e., precedent), “many textualists . . . still embrace statutory stare decisis.”216 Justice Hugo Black, a textualist, expressed similar separation-of-powers concerns with respect to overturning precedent and was in favor of an absolutist approach.217 Although there is debate on the extent to which statutory stare decisis should be absolute, Justice Black’s basic theory has been influential.218 Despite a discomfort with judges serving as policymakers, Justice Black recognized that “the resolution of statutory ambiguity inevitably requires some degree of policymaking.”219 However, he argued that to deviate from the statutory precedent was to usurp the legislature.220 As seen from all of Bostock’s opinions lodging attacks against legislating from the bench, this fear is aligned with textualism and has a similar motivation to textualism’s faithful-agent goal.221 More recently, and providing a clear connection to textualism, then-Professor Amy Coney Barrett invoked this separation-of-powers idea as a form of judicial restraint.222 All of this is to show that an approach that attempts to recharacterize the unit of linguistic analysis in order to ignore statutory precedent should be met with suspicion. To get around statutory stare decisis is to miss the opportunity to internalize many of the goals of textualism.

2. A preference for rules and its connection to a pragmatics-plus-semantics approach.

Combining the benefits of a strict adherence to precedent with a semantics approach furthers a textualist’s goal of judicial restraint through a reliance on rules. When precedent exists, it is as if each word in a phrase has a single dictionary definition.223 As a consequence, a judge can reach an understanding of the text without considering the exact problem facing the court.224 Whether this is a positive depends on where one comes down on the rules versus standards debate. But there is a link between textualism and rules.225 Although I have conceived of predictability as a corollary of judicial restraint, promoting predictability can lead to judicial restraint too.226 Because rules provide greater predictability, they can limit value judgments.227

Applying a precedent-based-semantics approach to a novel Title VII context provides evidence of how it develops a rule. For example, in a case examining the protections that Title VII affords to religion, the path to reaching the rule is clear. A judge can examine the limited precedent interpreting “because of,” “discriminate,” and “individual” in the context of Title VII and reach the straightforward rule that an employer violates the statute when it intentionally fires an individual employee based in part on religion. The consequences of this rule on the case before the court would be the basis of a judge’s argument. In a post-Bostock world, the rule does not change. Maintaining the religion example, Justice Gorsuch’s rule resulting from the semantics approach duplicates much of the work but adds the element that concepts that are inextricably linked with religion must be encompassed within Title VII’s religious protections. How a particular judge feels about discrimination against religion, sex, or sexual orientation plays a more limited role because the judge cannot cherry-pick a meaning of a statute that aligns with her values.

C. Addressing Novel Statutes

My resolution to rely on alternative methods when there is no precedent reveals something important about my sliding-scale approach—the source of restraint is derived more from precedent than the words of the statute. Because precedent interpreting the original meaning of words severely limits the ability to cherry-pick, a judge that uses a precedent-based-semantics approach successfully promotes textualism’s goals of judicial restraint, predictability, rule of law, and democratic accountability. But is this judge being a textualist? As mentioned, “many textualists . . . still embrace statutory stare decisis.”228 That does not mean that a textualist is comfortable using precedent when she does not have to.229 In the case of a novel statutory phrase, there is in fact no on-point precedent. Instead, a judge using the precedent-based-semantics method would be going out of her way to use precedent in pursuit of promoting the goals of textualism. It would not be a stretch to say that precedent interpreting a statute is extratextual. Therefore, allowing such a precedent to be a preliminary step would violate the idea that a clear text should not require extratextual considerations.230 This returns the interpreter to ambiguity thresholds, potentially undoing all the restraint that the precedent-based-semantics approach could offer.231

When the precedent that a judge is using to determine the meaning of the individual words is from a nontextualist judge’s opinion, a similar tension exists. The job of a textualist judge is to “interpret[ ] a statute in accord with the ordinary public meaning of its terms at the time of its enactment.”232 But it is possible to imagine that a nontextualist judge got the meaning of one of the words in a phrase wrong. This can mean that the interpretation is misaligned with the ordinary meaning at the time of enactment or completely detached from the ordinary meaning of the terms. If this occurs, should a textualist ignore the benefits associated with precedent in favor of the process that defines textualism?

Dictionary usage is evidence that textualists are willing to use extratextual sources to help determine the ordinary meaning of the text, but evidence is limited on whether this willingness extends to precedent defining terms. Judge Easterbrook has recognized both that precedent and the text are an “old pair” and the importance of precedent to the legitimacy of court opinions.233 In attempting to formulate this theory—admittedly, never reaching a conclusion—Judge Easterbrook provides an unsatisfying answer that “precedent can be a destabilizing as well as a stabilizing influence.”234 Judge Easterbrook does suggest that there is no compelling reason to have weak constitutional stare decisis and argues that it makes more sense for constitutional stare decisis to have the same presumption of statutory stare decisis.235 Judge Easterbrook does not, though, state how strong the presumption for stare decisis should be and doubts the adequacy of the rationales supporting a strong statutory stare decisis.236 Professor Adrian Vermeule does argue for the super-strong presumption for statutory stare decisis, suggesting that this rule would reduce costs to the legal system.237 Although not connecting it with statutory stare decisis, Vermeule makes a similar argument for a more textualist approach to judging.238 However, Justice Thomas has taken an aggressive approach, expressing a willingness to overturn wrongly decided precedent when interpreting the Constitution.239 Justice Thomas has also recently expressed willingness to ignore statutory stare decisis.240 Professor Anita Krishnakumar says that textualists of the “post-Scalia era” do not adhere to statutory stare decisis because of a presumption that there is a correct answer to interpretative questions, creating discomfort when following a precedent that reaches the wrong answer.241 However, the choice to overrule precedent increases discretion and the ability to inject value judgments, undermining the goals of textualism.242

D. Problems with the Alternative Methods for Finding Ordinary Meaning

This Section assesses the relative restraining power of the alternative methods for finding ordinary meaning discussed in Part III.C. Arguments regarding interpretative methodology are centered around relative measures.243 As such, arguing in favor of my solution requires an examination of how effectively alternative methods restrain judges. Because my solution does fully rely on pragmatics when interpreting a truly novel statute, this Section also selects the alternative method that is best at restraining judges.

Justice Kavanaugh’s best-reading approach and its open application of the absurdity doctrine244 create the same conditions for value-based judgments that it aims to solve. The Court has said that the absurdity doctrine dictates that “a court’s obligation to the text ceas[es] when ‘the absurdity and injustice of applying the provision to the case, would be so monstrous, that all mankind would, without hesitation, unite in rejecting the application.’”245 Professor John Manning, a prominent textualist scholar, argues that the absurdity doctrine is flawed because it does not provide judges an “intelligible basis on which to set aside clear textual commands in favor of likely legislative intent.”246 Similar to the problems that Justice Kavanaugh raises against ambiguity, it is not clear what makes a statute absurd and what level of absurdity should trigger the absurdity doctrine’s application.247 This missing “intelligible basis” serves to open the door for value-based judgments. Manning does relent on throwing out the absurdity doctrine entirely, arguing that the baseline meaning to consider when using the absurdity doctrine is not the literal meaning but a meaning that relies on context.248 My goal is to determine what considerations are appropriate when considering the context and the resulting ordinary meaning, so I am back to square one.

Justice Scalia’s approach, which relies on canons of interpretation, also does little to restrain judges from making value-based judgments. The criticism lodged against dictionaries has been that “there are so many of them and each offers a variety of definitions for common terms,” so judges can cherry-pick a definition and are left unrestrained to make value-based arguments.249 When a judge is able to make a value-based judgment, predictability is lost, and the rule of law suffers. The existence of fifty-seven canons, which Justice Scalia and Professor Bryan Garner suggest might be only one-third of the total “valid canons,” presents a similar issue.250 When there are too many approaches to choose from, a judge can cherry-pick the canons that work best to support their value judgment. When this possibility exists, predictability suffers, and the purposes of textualism are not met.251

Corpus linguistics is not free from problems that should give pause for anyone committed to finding ordinary meaning. The first problem is fundamental. Corpus linguistics has a tendency to reach the “prototypical” meaning and not the ordinary meaning.252 This means that the meaning corpus linguistics provides will be too narrow.253 Supporters of corpus linguistics find this concern to be overstated.254 Corpus linguistics does more than give just the most common meaning of the word. Instead, it provides the relative frequency of multiple meanings. Whether ordinary meaning equates to the prototypical meaning or encompasses senses that are lower down the list is a question of law.255 This response, however, sheds light on the second problem with corpus linguistics—it is an effective tool to find ordinary meaning but does not necessarily constrain judges. The relative frequency that equates to ordinary meaning is up to the judge. Similarly, whether the types of considerations—semantic, pragmatic, and time period—are applicable is up to the judge’s discretion.256

Although precedent is likely the best limit on cherry-picking, corpus linguistics may be the appropriate approach when there is no precedent.257 The absurdity doctrine or an ambiguity threshold have two unknowns that leave room for value-based judgments—what makes something ambiguous or absurd and how much is too much. Corpus linguistics still requires a determination of how much is enough to equate to ordinary meaning, but a judge has the metric of relative frequency. Plus, the empirical nature of corpus linguistics restrains by “facilitat[ing] transparency and scrutiny.”258 Judges being better able to check the work of other judges incentivizes against cherry-picking considerations that allow for a particular, ideological result.

Conclusion

Regardless of a judge’s selected school of statutory interpretation, the text is always the starting place for a judge trying to apply a law. But the increasing prominence of textualism has emphasized the importance of carefully finding the original meaning of the statute’s text. This commonsense approach is an attractive means to maintain the legitimacy of the Court. Forcing a judge to be restrained by words on the page promotes predictability, further promoting the preservation of the rule of law and democratic accountability. However, the concept of ordinary meaning is itself elusive. Bostock exemplifies the consequence of this problem. Three justices claiming to be adhering to the ordinary meaning of the phrase “discriminate because of sex” go about finding the ordinary meaning of the phrase in three different ways. The end result is a majority that held that Title VII protects gay and transgender individuals and two dissents that felt that the Court had usurped the legislature. Going forward, a resolution of who is correct—or most correct—will benefit textualism and the general predictability of judges who adhere to the philosophy.

With textualism’s goal of judicial restraint in mind, this Comment offers a novel sliding-scale approach that reconciles Justice Gorsuch’s and Justice Kavanaugh’s absolutist positions. Piecing together the meaning of terms derived from precedent offers fewer opportunities to cherry-pick than alternative methods for finding ordinary meaning and comports with a textualist’s preference for rules. This is particularly true when the Supreme Court has interpreted the words in the statute in question. But this sort of precedent may not always exist. The ability of a judge to cherry-pick increases when she must rely on precedent interpreting other statutes than the one in question. To counteract this decrease in restraint, the meaning that the precedent-plus-semantic approach provided requires more support. A judge can offer this support with an examination of the broader statutory context (i.e., looking to pragmatics-based arguments). This need for support further increases when relying on precedent from other courts. When a court encounters a truly novel statute, it then becomes appropriate to consider alternative methods that incorporate pragmatics. If judges choose to follow my sliding-scale approach, then they will be engaging in a method that actively seeks to restrain judges. This will, in turn, benefit textualism’s underlying goals of predictability, rule of law, and democratic accountability.

• 1140 S. Ct. 1731 (2020).
• 2Id. at 1754.
• 3Id. at 1737.
• 442 U.S.C. § 2000e-2(a)(1).
• 5Bostock, 140 S. Ct. at 1741 (emphasis added).
• 6Id. at 1743.
• 7Id. at 1824–25 (Kavanaugh, J., dissenting).
• 8Ash McMurray, Semantic Originalism, Moral Kinds, and the Meaning of the Constitution, 2018 BYU L. Rev. 695, 711 (2018).
• 9Bostock, 140 S. Ct. at 1827 (Kavanaugh, J., dissenting).
• 10Korta Kepa & John Perry, Pragmatics, in The Stanford Encyclopedia of Philosophy (Edward N. Zalta ed., 2020), https://perma.cc/BDQ9-YSRV.
• 11Bostock, 140 S. Ct. at 1827 (Kavanaugh, J., dissenting).
• 12Bostock, 140 S. Ct at 1826 (emphasis in original).
• 13Kepa & Perry, supra note 10 (emphasis in original).
• 14The 2015 Scalia Lecture: A Dialogue with Justice Elena Kagan on the Reading of Statutes, Harv. L. Today at 8:29 (Nov. 17, 2015), https://perma.cc/APM3-Y9EU.
• 15Id.
• 16Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 356 (2012) (quoting N.Y. Tr. Co. v. Comm’r, 68 F.2d 19, 20 (2d Cir. 1933)).
• 17Bostock, 140 S. Ct. at 1826 (Kavanaugh, J., dissenting); see also FCC v. AT&T Inc., 562 U.S. 397, 406 (2011).
• 18Kepa & Perry, supra note 10.
• 19AT&T, 562 U.S. at 406. Judge Learned Hand expressed a similar belief, albeit in a more fanciful manner, when he wrote that “the meaning of a sentence may be more than that of the separate words, as a melody is more than the notes.” Helvering v. Gregory, 69 F.2d 809, 810–11 (2d Cir. 1934).
• 20See Amy Coney Barrett, Statutory Stare Decisis in the Courts of Appeals, 73 Geo. Wash. L. Rev. 317, 326 (2005). But see Anita S. Krishnakumar, Textualism and Statutory Precedents, 104 Va. L. Rev. 157, 163–65 (2018) (noting that textualist judges have been willing to ignore statutory stare decisis).
• 21See Barrett, supra note 20, at 325–27.
• 22See William N. Eskridge, Jr., Interpreting Law: A Primer on How to Read Statutes and the Constitution 81 (2016) (“[Textualism] helps judges create a predictable and objective rule of law, enjoys great legitimacy because it is faithful to the premises of the legislative drafting process, and is the most reliable way to carry out the great plans to which legislators have committed our society.”).
• 23See Andrew Koppelman, Bostock, LGBT Discrimination, and the Subtractive Moves, 105 Minn. L. Rev. Headnotes 1, 6 (2020). The difference between old textualism and New Textualism is the consideration of legislative history. New Textualism rarely, if ever, consults legislative history, while old textualism sees it as an important part of confirming or finding statutory meaning. William N. Eskridge, Jr., The New Textualism and Normative Canons, 113 Colum. L. Rev. 531, 532–33 (2013) (reviewing Scalia & Garner, supra note 16).
• 24See generally Scalia & Garner, supra note 16; Frank H. Easterbrook, The Role of Original Intent in Statutory Construction, 11 Harv. J.L. & Pub. Pol’y 59 (1987).
• 25John F. Manning, Equity of the Statute, 101 Colum. L. Rev. 1, 7 (2001).
• 26Eskridge, supra note 23, at 532.
• 27Id. at 532–33.
• 28Scalia & Garner, supra note 16, at xxix.
• 29Manning, supra note 25, at 7.
• 30Purpose can be said to be the “general aim,” and intent approximates the meaning and the “particularized application which the statute was ‘intended’ to be given.” Archibald Cox, Judge Learned Hand and the Interpretation of Statutes, 60 Harv. L. Rev. 370, 370–71 (1947).
• 31Manning, supra note 25, at 7.
• 32Id. at 18.
• 33Id. at 19.
• 34Id.
• 35Id. at 20.
• 36MCI Telecomms. Corp. v. Am. Tel. & Tel. Co., 512 U.S. 218, 231 n.4 (2001).
• 37Easterbrook, supra note 24, at 66.
• 38Id. at 61 (quoting Oliver Wendell Holmes, Jr., The Theory of Legal Interpretation, 12 Harv. L. Rev. 417, 419 (1899), reprinted in Oliver Wendell Holmes, Collected Legal Papers 203, 207 (1920)).
• 39Eskridge, supra note 23, at 532.
• 40Bostock, 140 S. Ct. at 1737.
• 41See Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 24 (1997) (“[T]he good textualist is not a literalist.”); Caleb Nelson, What Is Textualism?, 91 Va. L. Rev. 347, 376 (2005) (“[N]o mainstream judge is interested solely in the literal definitions of a statute’s words.”).
• 42John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387, 2392–93 (2005).
• 43Eskridge, supra note 22, at 81.
• 44Manning, supra note 42, at 2393. For an empirical study on the truth of this statement, see Kevin P. Tobia, Testing Ordinary Meaning, 134 Harv. L. Rev. 726, 771 (2020) (finding that dictionary definitions can reflect extensive uses).
• 45Bostock, 140 S. Ct. at 1828 (Kavanaugh, J., dissenting).
• 46Manning, supra note 25, at 16.
• 47Scalia & Garner, supra note 16, at 9.
• 48See infra Part II.A.
• 49See infra Part II.B–C.
• 50See Easterbrook, supra note 24, at 65 (“The meaning of statutes is to be found not in the subjective, multiple mind of Congress but in the understanding of the objectively reasonable person.”).
• 51See Jeffrey L. Kirchmeier & Samuel A. Thumma, Scaling the Lexicon Fortress: The United States Supreme Court’s Use of Dictionaries in the Twenty-First Century, 94 Marq. L. Rev. 77, 84–92 (2010) (tracking the increasing usage of dictionaries of Supreme Court Justices); Scalia & Garner, supra note 16, at 415–24 (arguing that dictionaries may not solve every interpretative problem and providing dictionary recommendations); Manning, supra note 42, at 2458–59 (noting that dictionaries may be applicable but should be consulted only after narrowing the pool of definitions through an examination of context); Tara Leigh Grove, Which Textualism?, 134 Harv. L. Rev. 265, 279–85 (2020) (describing the divide between “formalistic textualism” and “flexible textualism”).
• 52See Brett M. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2144 (2016) (reviewing Robert A. Katzmann, Judging Statutes (2014)).
• 53See id.
• 54See id. at 2138–39 (citing Ward Farnsworth, Dustin F. Guzior & Anup Malani, Ambiguity About Ambiguity: An Empirical Inquiry into Legal Interpretation, 2 J. Legal Analysis 257, 290 (2010)).
• 55Easterbrook, supra note 24, at 62; see also Kavanaugh, supra note 52, at 2136–37.
• 56See Saul Levmore, Ambiguous Statutes, 77 U. Chi. L. Rev. 1073, 1077–78 (2010); Farnsworth et al., supra note 54, at 290.
• 57Bostock, 140 S. Ct. at 1737.
• 58Id. at 1757, 1763 (Alito, J., dissenting). In another context, Justice Scalia disagrees with unambiguity being a high standard. See Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 520–21 (1989).
• 59Kavanaugh, supra note 52, at 2145.
• 60Id.
• 61See id. at 2150 (“In a world without initial determinations of ambiguity, judges would instead decide on the best reading of the statute. In that world, legislative history would be largely limited to helping answer the question of whether the literal reading of the statute produces an absurdity.”).
• 62Id. at 2144.
• 63See id. at 2156–59.
• 64Easterbrook, supra note 24, at 65.
• 65Johnson v. United States, 529 U.S. 694, 718 (2000) (Scalia, J., dissenting).
• 66See generally Scalia & Garner, supra note 16.
• 67See Eskridge, supra note 23, at 534, 540–41. For a discussion of why this is unfortunate, see infra Part III.D.
• 68Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning, 127 Yale L.J. 788, 795 (2018).
• 69See id. at 828–29.
• 70Id.
• 71Id. at 829.
• 72See id. at 800–02 (noting the different definitions of ordinary meaning employed by courts).
• 73Lee & Mouritsen, supra note 68, at 828.
• 74Bostock, 140 S. Ct. at 1743.
• 75Id. at 1737.
• 76Id.
• 77Id. at 1753.
• 78Id. at 1741.
• 79Bostock, 140 S. Ct. at 1737.
• 80See id. at 1743.
• 81See id. at 1828 (Kavanaugh, J., dissenting); see also id. at 1755 (Alito, J., dissenting).
• 82See id. at 1825 (citing Scalia, supra note 41, at 24).
• 83See id. at 1761 (Alito, J., dissenting); id. at 1836–37 (Kavanaugh, J., dissenting).
• 84See infra notes 114–16 and accompanying text.
• 85Bostock, 140 S. Ct. at 1738.
• 86See id. at 1739–41.
• 87See id. at 1739 (quoting Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570 U.S. 338, 350 (2013)) (quotation marks omitted).
• 88See infra notes 116–20 and accompanying text.
• 89Bostock, 140 S. Ct. at 1738.
• 9042 U.S.C. § 2000e-2(a)(1).
• 91Bostock, 140 S. Ct. at 1739 (“The only statutorily protected characteristic at issue in today’s cases is ‘sex.’”).
• 92Id.
• 93See id. at 1739–41.
• 94Id. (citing Nassar, 570 U.S. at 346).
• 95Bostock, 140 S. Ct. at 1739.
• 96Id.
• 97Id. at 1740 (first citing Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 59 (2006); and then citing Watson v. Fort Worth Bank & Tr., 487 U.S. 977, 986 (1988)).
• 98Id. at 1740.
• 99Id.
• 100Bostock, 140 S. Ct. at 1740.
• 101Id. at 1741.
• 102Id.
• 103Id. at 1742.
• 104Id. at 1741.
• 105Bostock, 140 S. Ct. at 1824–25 (Kavanaugh, J., dissenting).
• 106Id. at 1825; see also Manning, supra note 42, 2392–93; Robertson v. Salomon, 130 U.S. 412, 414 (1889); Milner v. Dep’t of Navy, 562 U.S. 562, 577–78 (2011).
• 107Bostock, 140 S. Ct. at 1825 (Kavanaugh, J., dissenting); see also Eskridge, supra note 22, at 81; Scalia, supra note 41, at 17.
• 108Eskridge, supra note 22, at 81.
• 109Bostock, 140 S. Ct. at 1828 (Kavanaugh, J., dissenting).
• 110Id. at 1825.
• 111Id.
• 112Id.
• 113Id.
• 114Bostock, 140 S. Ct. at 1738 (majority opinion).
• 115Id. (citing New Prime Inc. v. Oliveira, 139 S. Ct. 532, 538–39 (2019)).
• 116Id. at 1827 (Kavanaugh, J., dissenting).
• 117Id.
• 118Id. at 1741.
• 119Bostock, 140 S. Ct. at 1826 (Kavanaugh, J., dissenting).
• 120Id.; see also FCC v. AT&T Inc., 562 U.S. 397, 406 (2011).
• 121Bostock, 140 S. Ct. at 1827 (Kavanaugh, J., dissenting).
• 122Id. at 1828.
• 123See id. at 1828–29.
• 124See id. at 1829–33.
• 125See id.
• 126Bostock, 140 S. Ct at 1829 (quoting Wis. Cent. Ltd. v. United States, 138 S. Ct. 2067, 2071 (2018)).
• 127Id. at 1833.
• 128Id. at 1826.
• 129See Lee & Mouritsen, supra note 68, 818–24 (defining and discussing the terms “semantic meaning” and “pragmatic meaning”).
• 130See, e.g., Zarda v. Altitude Express, Inc., 883 F.3d 100, 144 n.7 (2d Cir. 2018) (Lynch, J., dissenting). But see Grove, supra note 51, at 303–07 (arguing that judges should favor “formalistic textualism” because it protects the legitimacy of the judiciary).
• 131See, e.g., Zarda, 883 F.3d at 144 n.7 (Lynch, J., dissenting).
• 132Scalia & Garner, supra note 16, at 356 (quoting N.Y. Tr. Co. v. Comm’r, 68 F.2d 19, 20 (2d Cir. 1933)) (emphasis in original).
• 133Bostock, 140 S. Ct. at 1826 (Kavanaugh, J., dissenting); see also Manning, supra note 42, at 2458; Eskridge, supra note 22, at 62.
• 134See supra note 19 and accompanying text.
• 135See Eskridge, supra note 23, at 534 (citing Scalia & Garner, supra note 16, at 419–24).
• 136Kirchmeier & Thumma, supra note 51, at 86.
• 137See, e.g., Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 31–32 (2003) (stringing together the dictionary definitions of “origin” and “goods”). For a list of every time twenty-first century Justices have used a dictionary, see generally Kirchmeier & Thumma, supra note 51.
• 138See, e.g., Dastar, 539 U.S. at 33–34 (reviewing precedent and the relationship between the Lanham Act and copyright law to support the semantic definition of “origin of goods”).
• 139See Kirchmeier & Thumma, supra note 51, at 123–24.
• 140Bostock, 140 S. Ct. at 1827 (Kavanaugh, J., dissenting).
• 141Compare Grove, supra note 51, at 303–07 (defending Gorsuch’s approach because of its ability to promote judicial restraint), with id. at 283 n.108 (collecting criticisms of Gorsuch’s opinion).
• 142See generally Frank H. Easterbrook, Stability and Reliability in Judicial Decisions, 73 Cornell L. Rev. 422 (1988) (discussing the problems with developing a theory of precedent and historical attempts).
• 143See, e.g., Dastar, 539 U.S. at 33–34 (using precedent to support a meaning reached through stringing together dictionary definitions); Helvering v. Gregory, 69 F.2d 809, 811 (2d Cir. 1934) (using precedent to override a meaning reached through the stringing together of dictionary terms).
• 144See supra Part II.A.
• 145523 U.S. 75 (1998).
• 146138 S. Ct. 2067 (2018).
• 147Id. at 2070.
• 14826 U.S.C §§ 3201–41 (1937).
• 149Id.
• 150See id.
• 151Wis. Cent. Ltd., 138 S. Ct. at 2071.
• 152See id. at 2080 (Breyer, J., dissenting) (“[N]o one disputes that granting employees stock options is a form of remuneration.”).
• 153Id. at 2071 (majority opinion).
• 154See id. at 2072.
• 155Id. at 2071.
• 156Wis. Cent. Ltd., 138 S. Ct. at 2075.
• 157Id. at 2077–78 (Breyer, J., dissenting).
• 158Bostock, 140 S. Ct. at 1828 (Kavanaugh, J., dissenting).
• 159See id. at 1747 (majority opinion); Wis. Cent. Ltd., 138 S. Ct. at 2073.
• 160Wis. Cent. Ltd., 138 S. Ct. at 2069.
• 161555 U.S. 379 (2008).
• 162See id. at 381–82.
• 163Id. (quoting 25 U.S.C. § 465 (1934) (codified at 25 U.S.C. § 5108)).
• 164Id. at 385.
• 16525 U.S.C. § 479 (currently codified at 25 U.S.C. § 5129).
• 166Carcieri, 555 U.S. at 388.
• 167See id. at 383–84.
• 168See id. at 388.
• 169Id. at 390–91.
• 170Id. at 380.
• 171Carcieri, 555 at 388–91.
• 172Id. at 388–89.
• 173Wis. Cent. Ltd., 138 S. Ct. at 2071; see also Carcieri, 555 U.S. at 389–90.
• 174Wis. Cent. Ltd., 138 S. Ct. at 2071.
• 175Id. at 2071–72.
• 176See Bostock, 140 S. Ct. at 1829 (Kavanaugh, J., dissenting) (citing Wis. Cent. Ltd., 138 S. Ct. at 2071).
• 177See infra Part III.
• 178See Kirchmeier & Thumma, supra note 51, at 128 (“The Court focuses on resolving the issue presented in the case before it, using dictionaries where the individual Justices find them instructive.”).
• 179See supra Part II.B.
• 180See supra notes 25–28 and accompanying text.
• 181See Oncale, 523 U.S. at 82.
• 182See infra note 249 and accompanying text.
• 183See Eskridge, supra note 23, at 545–51 (critiquing the idea that text, dictionaries, and canons meaningfully restrain Justices).
• 184See, e.g., Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570 U.S. 338, 350 (2013) (citing Gross v. FBL Fin. Servs., Inc., 557 U.S. 167, 176 (2009)).
• 185See supra notes 138–40 and accompanying text.
• 186See Admin. Off. of the U.S. Cts., Federal Judicial Caseload Statistics 2020, U.S. Cts. (Mar. 31, 2020), https://perma.cc/6HY3-CVE5 (finding 425,945 filings in the U.S. district courts as of March 2020).
• 187See infra Part III.B.1.
• 188See Bostock, 140 S Ct. at 1739–43.
• 189See id. at 1828; infra notes 249–51 and accompanying text.
• 190See supra Part II.C.1.
• 191See Wis. Cent. Ltd., 138 S. Ct. at 2071; Carcieri, 555 U.S. at 388–89.
• 192See Wis. Cent. Ltd., 138 S. Ct. at 2071; Carcieri, 555 U.S. at 388–89.
• 193Carcieri, 555 U.S. at 389–90.
• 194Id.
• 195574 U.S. 528 (2015).
• 196Id. at 552 (Kagan, J., dissenting).
• 197Kavanaugh, supra note 52, at 2161.
• 198Yates, 574 U.S. at 531.
• 199See id.; Pub L. No. 107–204, 116 Stat. 745 (codified in scattered sections of 15 U.S.C, 18 U.S.C., 28 U.S.C, and 29 U.S.C.).
• 20018 U.S.C. § 1519.
• 201See Yates, 574 U.S. at 536.
• 202See id. at 539–42; id. at 549–51 (Alito, J., concurring).
• 203Id. at 555 (Kagan, J., dissenting) (citation omitted).
• 204See id. at 553–55.
• 205See id. at 555–56.
• 206See Yates, 574 U.S. at 555–56 (Kagan, J., dissenting).
• 207See Bostock, 140 S. Ct. at 1740–41.
• 208See Yates, 574 U.S. at 567 (Kagan, J., dissenting).
• 209See Bostock, 140 S. Ct. at 1767 (Alito, J., dissenting); id. at 1828 (Kavanaugh, J., dissenting).
• 210See infra Part III.D.
• 211See Krishnakumar, supra note 20, at 165 (quoting William N. Eskridge, Jr., Overruling Statutory Precedents, 76 Geo. L.J. 1361, 1362 (1988)).
• 212Id. at 165 (quoting Eskridge, supra note 211, at 1362).
• 213Burnet v. Colo. Oil & Gas Co., 285 U.S. 393, 406 (1932) (Brandeis, J., dissenting).
• 214Frank E. Horack, Jr., Congressional Silence: A Tool of Judicial Supremacy, 25 Tex. L. Rev. 247, 251 (1947).
• 215Id.
• 216Barrett, supra note 20, at 326.
• 217See id. at 325–26 (citing Boys Mkts., Inc. v. Retail Clerks Union, Local 770, 398 U.S. 235, 256–58 (1970) (Black, J., dissenting)).
• 218Id. at 326.
• 219Id. at 325.
• 220See id. at 326.
• 221See Bostock, 140 S. Ct. at 1738; id. at 1754 (Alito, J., dissenting); id. at 1836–37 (Kavanaugh, J., dissenting).
• 222See Barrett, supra note 20, at 347–49.
• 223See supra Part III.A.
• 224See, e.g., Bostock, 140 S. Ct. at 1741 (“[A] straightforward rule emerges: An employer violates Title VII when it intentionally fires an individual employee based in part on sex.”).
• 225See Antonin Scalia, The Rule of Law as a Law of Rules, 56 U Chi. L. Rev. 1175, 1183–85 (1989) (discussing a link between textualism and “general rules”). But see Easterbrook, supra note 24, at 61 (“Meaning comes from the ring the words would have had to a skilled user of words at the time, thinking about the same problem.”).
• 226See Thomas W. Merrill, Originalism, Stare Decisis and the Promotion of Judicial Restraint, 22 Const. Comment. 271, 275 (2005) (“[Judicial r]estraint in this sense simply requires that the judge adhere to whatever method produces the most easily-predicted results.”).
• 227Scalia, supra note 225, at 1179–80.
• 228Barrett, supra note 20, at 326.
• 229See James R. Maxeiner, Scalia & Garner’s Reading Law: A Civil Law for the Age of Statutes?, 6 J. Civ. L. Stud. 1, 11 (quoting Scalia & Garner, supra note 16, at 411) (“Stare decisis . . . is not a part of textualism. It is an exception to textualism (as it is to any theory of interpretation) born not of logic but of necessity.” (emphasis omitted)).
• 230See Bostock, 140 S. Ct. at 1737 (“When the express terms of a statute give us one answer and extratextual considerations suggest another, it’s no contest. Only the written word is the law.”).
• 231See supra Part II.C.1; supra Part III.A.1.
• 232Bostock, 140 S. Ct. at 1738.
• 233Easterbrook, supra note 142, at 422.
• 234Id. at 433.
• 235Id. at 426, 429.
• 236Id.
• 237Adrian Vermeule, Interpretive Choice, 75 N.Y.U. L. Rev. 74, 144–45 (2000) (arguing that strong statutory stare decisis decreases judicial decision costs and is more stabilizing than weak statutory stare decisis).
• 238Id. at 139 (recognizing that there are potential increases in decision costs with textualism but that the decision costs of legislative history are still theoretically greater).
• 239See, e.g., Gamble v. United States, 139 S. Ct. 1960, 1984 (2019) (Thomas, J., concurring) (“When faced with a demonstrably erroneous precedent, my rule is simple: We should not follow it.”).
• 240See Krishnakumar, supra note 20, at 211 n.203 (“Justice Thomas joined thirteen of seventeen Roberts Court opinions advocating overruling a statutory precedent (and authored ten of them).”).
• 241Id. at 204–05. Despite the broad label of “textualists of the ‘post-Scalia era,’” Professor Krishnakumar’s data ends in 2015 and does not include Justice Gorsuch’s or Justice Kavanaugh’s time on the Court. Id. at 228–33.
• 242Cf. William Baude, Precedent and Discretion, 2019 Sup. Ct. Rev. 313, 329–30 (2019) (“[D]iscretionary features render precedent worse than useless. They make it a tool for evading other requirements of the law, and a threat to certain aspects of judicial neutrality.”).
• 243See, e.g., Eskridge, supra note 23, at 551 (criticizing Scalia’s “canons-based textualism” for being “a relatively less constraining approach” than the use of “legislative materials”).
• 244See Kavanaugh, supra note 52, at 2144 (noting that the best-reading approach allows judges to use the absurdity doctrine to “justify departure from the text”).
• 245Manning, supra note 42, at 2387–88 (quoting Sturges v. Crowinshield, 17 U.S. (Wheat 8.) 122, 203 (1819)); see also Scalia & Garner, supra note 16, at 235 (“What the rule of absurdity seeks to do is what all rules of interpretation seek to do: make sense of the text.” (emphasis in original)).
• 246Manning, supra note 42, at 2454–55.
• 247See supra notes 52–54 and accompanying text. But see Scalia & Garner, supra note 16, at 388 (arguing that a judge may consult legislative history to find at least one “rational legislator” who interprets the statute to mean what the present court considers an absurd result; if this legislator exists, then the statute is not absurd).
• 248Manning, supra note 42, at 2461–64.
• 249See Eskridge, supra note 23, at 534.
• 250See Scalia & Garner, supra note 16, at 9.
• 251See Eskridge, supra note 23, at 544 (discussing problems with cherry-picking among canons).
• 252Tobia, supra note 44, at 761 (showing that corpus tracks “prototypical” definitions better than “technical[ ]” definitions).
• 253See id. at 795–97 (detailing several “fallacies” that are common in arguments from corpus linguistics but tend to produce narrow meanings of terms).
• 254See Lee & Mouritsen, supra note 68, at 874.
• 255See id. at 874–75.
• 256See id. at 866 (“[Judges] should not . . . overlook the potential for subjectivity or even strategic manipulation [when using corpus linguistics].”).
• 257See id. at 867 (“The potential for subjectivity and arbitrariness is not heightened but reduced by the use of corpus linguistics.”).
• 258Id. at 868.