We are all textualists now, or so it has been claimed. But textualism, the practice of interpreting statutes solely by reference to their words, is often associated with conservative judicial outcomes. This is especially true when a focus on statutory text is combined with the belief that the meanings of words are fixed. This combination creates a sort of textualist originalism, in which judges interpret statutes in accordance with what the words of a statute meant to the relevant linguistic community at the time of a statute’s enactment.

In reaction to this conservative interpretive method, rejecting textualism but keeping an originalist commitment to fixed meanings provides one possible progressive response. Rather than focusing primarily on the statute’s language, one might instead look to the statute’s animating logic, purpose, or potential to create certain moral or economic outcomes. But rejecting a focus on statutory text is not the only progressive response to the originalist textualist. A second approach accepts that the meanings of statutes derive from text but denies that statutory text has a meaning that is fixed and unchanging. Often, advocates of this latter approach run into difficulty specifying how and when the meanings of words used in statutes change. And some might worry that any such account will leave judges too much room to determine that the meanings of a statute’s words have changed, thus enabling judges to express their policy preferences through the act of statutory interpretation.

This Comment addresses the second approach. It engages with a particular account in philosophy of language that views meanings as resulting from inferential connections among concepts. Inferentialism suggests that we can think of these inferential connections as constitutive of meaning and thus think of meaning as in a real sense responsive to the on-the-ground effects of using a particular word. As those effects change, as new inferential connections are recognized, so too change the meanings of words. This Comment argues that this repeated process of changed inferential significance provides us with an account of dynamic meaning that judges can take notice of but not impose themselves upon. It thus provides a methodology through which judges can read statutory text to mean something new or different without thereby merely expressing a political preference.


Of all the things that judges do, central to those activities is saying what the law is, which means saying what the words in statutes mean. “This is a pure question of statutory interpretation and thus well within the judiciary’s competence.”1 So asserts Chief Judge Diane Wood in Hively v Ivy Tech Community College of Indiana,2 in which the en banc Seventh Circuit decided that discrimination on the basis of sexual orientation is sex discrimination. Courts that have considered the sex discrimination issue widely agree that the heart of the matter is what the words of Title VII3 of the Civil Rights Act of 19644 mean. While such an assessment of meaning may be “well within the judiciary’s competence,” it is also true that, as Wood notes in her very next sentence, “[m]uch ink has been spilled about the proper way to go about the task of statutory interpretation.”5

This Comment engages with that normative question. Specifically, it suggests that one place we can look for clues about what judges should do when asking what words mean is to philosophy of language, which provides generalizable theories about how
concepts become meaningful at all. Semantic theory attempts to provide an account of meaning generally by telling a story about how symbols or sounds can be thought to have content and used to communicate, what that communication entails, and how communication is achieved by human language users.

This Comment puts legal writing about language in conversation with a certain kind of philosophical thought about semantics. This Comment proceeds in four Parts: In Part I, this
Comment (briefly) surveys interpretive theories animating judicial approaches to statutory interpretation. Part II engages with the Seventh Circuit’s interpretation of Title VII’s language as an example of those approaches. In Part III, this Comment develops a semantics not explicitly considered in those debates: the inferentialism of Professor Robert Brandom. In Part IV, having sketched the basic components of an inferentialist understanding, this Comment argues that such an understanding is in tension with the originalist and dynamic “judicial interpretive” arguments Parts I and II consider and that various attempts at reconciliation are unsatisfactory.

This Comment argues that Brandomian inferentialism gives us reasons to think differently about statutory interpretation in a few ways: The framework rejects the dualism of making/finding in interpretive activity. It rejects the idea that meanings are fixed as a matter of semantic necessity. Most importantly, it suggests a certain kind of semantic externalism according to which language users’ grasp of concepts is not coextensive with the concepts’ full inferential reach. This Comment suggests that Brandomian inferentialism gives us powerful reasons to think that originalist, meaning-as-fixed theories of statutory interpretation must be incomplete and that there is a principled understanding of dynamic interpretation that does not merely reduce to a covert exercise of judicial lawmaking. Furthermore, to the extent that existing theories of statutory interpretation give us reason to think that the process of interpretation is meaningfully “dynamic”—taking account of evolving legal and social context—Brandom’s more general semantics puts some meat on those theoretical bones.6

I.  Interpreting, Generally

Here are two ways of thinking about statutory interpretation. On a first, “originalist” account, the meaning of statutory terms is fixed.7 Whether by the public understanding of the terms in the statute at the time of its passage or the intentions of its ratifiers, originalists believe that statutes have meanings that are largely static. Originalism thus emphasizes a certain sort of interpretive passivity. To talk about meaning this way makes it out to be a sort of archaeological process, an uncovering of what was actually there in the heads of the legislators or the perfectly average
English speaker at the time of the statute’s passage.8 In this way, we can understand theories emphasizing original public meaning, legislative purpose, or legislative intent as broadly originalist in the sense in which this Comment uses the term, for each understanding implies that meaning is essentially tied to some fixed, time-bound source.

On a second, “dynamic” account, statutory meaning is not static but instead changes with changes in society or the interpretive community. Rather than pegging the meaning of statutory terms to a common understanding or set of purposes at a specific point in time, a dynamic account, as this Comment understands it, allows for the possibility of meanings that change or evolve because of changing or evolving facts or beliefs. While a purposivist originalist may advocate for a certain kind of statutory evolution to best serve the original purposes of a statute in the current context, dynamic interpretation allows for statutory evolution that is not tied to any fixed point of reference.

A.    Originalist Interpretation

Originalist theories of interpretation exist in many forms, but this Comment defines originalism as reliance on static meanings. A first originalist approach sees statutory meanings as fixed by reference to the “original public meaning” of the terms in the statute as commonly understood at the time of the statute’s ratification.9 A second approach, “purposivism” (sometimes called intentionalism) also emphasizes fixedness and attempts to read statutes to be consistent with the actual or presumed intentions (or animating purposes) of the people who passed them.10

Each approach has adherents on the Supreme Court, but the unifying feature is a belief that law should be interpreted to preserve stability of legal meanings. The late archoriginalist Justice Antonin Scalia, for example, writes in Roper v Simmons:11 “In a system based upon constitutional and statutory text democratically adopted, the concept of ‘law’ ordinarily signifies that particular words have a fixed meaning. Such law does not change, and this Court’s pronouncement of it therefore remains authoritative until (confessing our prior error) we overrule.”12 In disputing the majority’s interpretation of the Eighth Amendment’s proscription on “cruel and unusual” punishment, Scalia criticizes the majority for “purport[ing] to make of the Eighth Amendment [ ] a mirror of the passing and changing sentiment.”13 Scalia would instead have the Court look to “the original meaning of the Eighth Amendment,” advocating for reasoning about the amendment that is fettered to the understanding that obtained at the time of its enactment.14

One asserted benefit of originalism, as Scalia’s opinion in Roper argues, is that it removes the individual preferences of specific judges from legal interpretation. Indeed, whenever statutory or constitutional text is read to command a new or different result, the originalist desire for stable and constant meanings often emerges. Chief Justice John Roberts, in his dissenting opinion in Obergefell v Hodges,15 mounts similar criticism of the majority’s reasoning about marriage as a fundamental right, alleging that “[f]ive lawyers have . . . enacted their own vision of marriage as a matter of constitutional law.”16 Roberts criticizes the majority opinion as “an act of will, not legal judgment . . . based not on neutral principles of constitutional law, but on its own ‘understanding of what freedom is and must become.’”17 A healthy respect for fixed meanings, by contrast, acts as a constraint on judges who are merely discovering the meaning accepted at the time of enactment rather than imposing their own preferences on the text.

Purposivism, too, proceeds by looking backward to something fixed: the explicit or underlying intentions or purposes of the legislators that enacted a statute. And while interpretation based on underlying purposes has been used to justify an updated understanding of statutory text (because those underlying purposes may be served differently in different circumstances), the interpreter must still uncover a fixed underlying purpose and interpret by reference to that stable point.18

This approach also has defenders among the Supreme Court’s current justices, perhaps most notably Justice Stephen Breyer. In National Labor Relations Board v Noel Canning,19 in which the Court clarified the extent of the president’s authority under the Recess Appointment Clause, Breyer can be read as
advancing something like a purposivist approach in his majority opinion. Breyer rejects reasoning based on the actual intents of the framers. Noting that “some argue that the Founders would likely have intended the Clause to apply only to inter-session
recesses,”20 Breyer argues that this is beside the point for statutory interpretation purposes, as “[t]he question is not: Did the Founders at the time think about intra-session recesses?”21 Rather than focus on what the Founders actually contemplated or understood, Breyer stresses that the question is whether the Founders intended that the clause may apply more broadly “to somewhat changed circumstances.”22 He presents it as dispositive that “the Framers likely did intend the Clause to apply to a new circumstance that so clearly falls within its essential purposes.”23 Thus, the issue of statutory interpretation is still being resolved by reference to something fixed and historical (the supposed unexpressed intentions of the Framers), but the meaning of the clause need not derive from any actually contemplated intentions or beliefs of the Framers. Nevertheless, while our understanding of the meaning of the clause may evolve as new circumstances cause us to reconsider those preferences, the task of the interpreter is still largely passive.

Originalism implies that meanings are found and not made. Meanings of statutes are determined by some aspect of the understanding of a fixed group of people at a fixed moment in time. Indeed, Professor Lawrence B. Solum has called this constraint on interpretive activity the “Fixation Thesis,” which posits in the context of constitutional interpretation that “the original meaning (‘communicative content’) of the constitutional text is fixed at the time each provision is framed and ratified.”24 Solum emphasizes that fixation is a semantic fact about the communicative content of words: it affects what the words of a statute mean rather than affecting only how those words are used (their pragmatic effect). He explains that the constraint is consistent with a general “fixation of conventional semantic meaning by linguistic facts at the time a communication occurs.”25

But even Solum’s fixation thesis admits of “linguistic drift,” the idea that “[w]ords and phrases acquire new meanings over time.”26 This change, however, does not affect the meaning of prior uses of a particular word or phrase—the communicative content of which “is a function of the meaning at the time the communication was produced.”27 So while the words of a particular statute might mean something different if enacted at time one as opposed to time two, the meaning of any particular enacted statute remains fixed at time one.

There is perhaps one caveat to this general rule. Judge Frank Easterbrook, himself an unapologetic originalist, has contemplated circumstances in which society has drifted so far from the original meaning of statutory text that it is no longer useful in judicial reasoning. In this narrow situation, even the originalist may recognize a need for an active judicial role in shaping statutory meanings.

Words don’t have intrinsic meanings; the significance of an expression depends on how the interpretive community alive at the time of the text’s adoption understood those words. The older the text, the more distant that interpretive community from our own. At some point the difference [of the linguistic drift] becomes so great that the meaning is no longer recoverable reliably.28

Still, these sorts of extreme examples are the exception rather than the rule and presuppose a degree of cultural and linguistic change not typically seen in the law.29

B.    Dynamic Interpretation

In contrast to originalist theories of interpretation, dynamic accounts of interpretation allow for the possibility of change in statutory meaning. This change applies not just to subsequent uses of words or phrases (as in originalist linguistic drift) but also to a given use. That is, while the originalist countenances linguistic drift such that the words in statute X may mean something different if enacted at time one as opposed to time two (with the meaning of any particular enacted statute remaining fixed), dynamic interpretation allows for the possibility that the semantic content of statute X can be different at time two. In other words, it is not merely that statute X would mean something different if enacted today but that statute X actually does mean something different today than it did when ratified.

Dynamic accounts of interpretation, too, exist in many forms. In Professor William Eskridge’s dynamic interpretation theory,

[S]tatutory interpretation involves the present-day interpreter’s understanding and reconciliation of three different perspectives, no one of which will always control. These three perspectives relate to (1) the statutory text, which is the formal focus of interpretation and a constraint on the range of interpretive options available (textual perspective); (2) the original legislative expectations surrounding the statute’s creation, including compromises reached (historical perspective); and (3) the subsequent evolution of the statute and its present context, especially the ways in which the societal and legal environment of the statute has materially changed over time (evolutive perspective).30

While Eskridge’s approach emphasizes a balance among these three perspectives, he does suggest that “the more striking the changes in circumstances (changes in public values count more than factual changes in society), the greater weight the interpreter will give to evolutive considerations.”31

Eskridge emphasizes the importance of the perspective of the particular interpreter, who must weigh the various sources of statutory meaning, to statutory interpretation. The task “involves policy choices and discretion by the interpreter over time as she applies the statute to specific problems and is responsive to the current, as well as the historical, political culture.”32

But Eskridge’s account focuses on the practical effects, rather than the semantic content, of dynamic meanings. He writes:

Because statutes have an indefinite life, they apply to fact situations well into the future. When successive applications of the statute occur in contexts not anticipated by its authors, the statute’s meaning evolves beyond original expectations. Indeed, sometimes subsequent applications reveal that factual or legal assumptions of the original statute have become (or were originally) erroneous; then the statute’s meaning often evolves against its original expectations.33

While Eskridge may accurately describe how legal actors’ use of statutes in legal reasoning or interpreters’ beliefs about statutory text change over time, Eskridge does not supply a well worked-out account of dynamic meaning at the semantic level. This
Comment seeks to tie that account of pragmatic change to a broader semantic story about how those changes in use imply semantic changes in the communicative content of the statute itself.

Dynamic statutory interpretation could also be thought of as proceeding incrementally—building on the interpretations and authoritative decisions that came before, but also shaping the entire body of doctrine so that it best conforms to some moral or rational ideal.34 This view allows meaning to evolve dynamically as new authoritative decisions are made and our understanding of closely related concepts gains a new or sharper dimension.

If originalist statutory interpretation understands meanings as found, this sort of dynamic approach understands meanings as unapologetically made—though perhaps by reference to some guiding standards. Dynamic interpretation involves not “extract[ing]” meanings but “giving a fresh meaning to a statement.”35

But there is a “specter of skepticism” that haunts the notion of a dynamic statutory interpretation.36 For we might worry that any purported change in the meaning of a statute (for example, for asserted morality- or efficiency-promoting reasons) is really nothing more than a dressed-up assertion of judicial power. The debate over the meaning of Title VII prompts the more general question of whether dynamic approaches to statutory interpretation must reduce to “overt or covert” judicial interpretive updating—an “entirely judge-made” meaning37 that is unresponsive to legal reasons characteristic of proper judicial activity.

II.  An Interpretive Case Study: “Sex Discrimination”

There are two ways of thinking about the meaning of the phrase “sex discrimination.” On a first, narrow account, “sex discrimination” refers only to discrimination on the basis of biological sex—that is, on the basis of reproductive functions. According to the narrow account, sex discrimination occurs when an individual is discriminated against solely because of the biological sex of the individual. If the discrimination occurs “against women because they are women [or] against men because they are men,”38 rather than at some other level of descriptive specificity, then we have a case of sex discrimination on the narrow understanding.

On a second, broader account, “sex discrimination” refers to a general class of discriminatory behavior that includes discrimination on the basis of reproductive functions and also certain associated concepts and behaviors. According to the broader account, sex discrimination is best thought of as a genus of discriminatory activity that comprises several species. Even if discrimination “against women because they are women [or] against men because they are men”39 came first in the evolutionary lineage of the concept, there are now related forms of discrimination that count as sex discrimination, including discrimination on the basis of sexual orientation.

The goal of this Part is to understand how courts have analyzed and understood “sex discrimination” as it relates to Title VII. To that end, this Part begins by briefly summarizing some of the Title VII cases in order to reject a broader understanding of “sex discrimination.” This Part then examines Hively as an example of various approaches to statutory interpretation in action.

A.    “Sex Discrimination” and Title VII, Historically

From the 1970s to the present, circuit courts have considered whether and how discrimination on the basis of sexual orientation fits within Title VII. The Supreme Court first recognized an enlarged understanding of “sex discrimination” in Price
Waterhouse v Hopkins.40 There, a woman brought a retaliation claim, alleging she had been denied partnership because she failed to conform to stereotypical expectations of female behavior.41 The Court found the claim cognizable under Title VII, recognizing that traditional sex discrimination includes discriminating against an individual who fails to adhere to gender stereotypes.42 The Court reasoned that Title VII clearly made gender an “impermissible motive” in adverse treatment and that, “[i]n the specific context of sex stereotyping, an employer who acts on the basis of a belief that a woman cannot be aggressive, or that she must not be, has acted on the basis of gender.”43

After the Price Waterhouse decision, circuit courts attempted to walk a fine line: the only way those with gender variance or homosexual orientation can have an actionable claim is when they are stereotyped on the basis of sex. Thus, gay and lesbian employees “may nonetheless bring suit when discriminated against on the basis of [stereotypes of] his or her sex.”44 The near-universal adoption of this approach among circuits sets the stage for Hively.

B.    Interpretive Approaches in Hively

In April 2017, the Seventh Circuit ruled that Title VII protects employees from discrimination on the basis of sexual orientation.45 In so doing, the Seventh Circuit interpreted discrimination on the basis of “sex” in the statute’s language to include sexual orientation discrimination: “[W]e conclude today that discrimination on the basis of sexual orientation is a form of sex discrimination.”46

Kimberly Hively, a part-time adjunct faculty member at a community college in Indiana, claimed that the college did not renew her contract because she identified as a lesbian.47 Hively sued her former employer, claiming that discrimination on the basis of sexual orientation is prohibited under Title VII as discrimination “on the basis of . . . sex.”48 Consistent with Seventh Circuit precedent, that claim was initially dismissed49 before the court agreed to rehear the case en banc to reconsider the status of sexual orientation discrimination claims under Title VII.50

1.   The majority opinion.

Chief Judge Wood’s majority opinion for the en banc court presents the issue as one of statutory interpretation. The opinion considers two lines of argument: first, using “the tried-and-true comparative method in which we attempt to isolate the significance of the plaintiff’s sex to the employer’s decision”; and second, an argument concerning a right to intimate association relying on Loving v Virginia51 and its progeny.52

Wood’s discussion of the comparative method analysis begins with the assertion that, if Hively had been a man married to a woman rather than a woman married to a woman, her employer’s actions would have been different.53 Thus, because the employer would treat a man married to a woman differently than Hively (a woman married to a woman), the employer has engaged in discrimination because of sex. But the dissent criticized this approach on the grounds that it actually conflates two variables in the comparison: sex and sexual orientation.54 The majority responds that such a criticism “begs th[e] question” because it assumes that consideration of sexual orientation is separable from pure sex discrimination, which was precisely the question before the court.55 But because consideration of sexual orientation discrimination requires the court to know the sex of the plaintiff, it is not clear that sexual orientation discrimination can be neatly separated from pure sex discrimination in this way.56 The majority also responds by noting that gender stereotyping was ruled impermissible under Title VII in Price Waterhouse and argues that “Hively represents the ultimate case of failure to conform to the female stereotype.”57

The majority also pursues a line of argument based on the right to intimate association and Loving. The logic is that, just as discrimination against employees for associating with a person of a different race is illegal under Title VII, so too is discrimination against employees for associating intimately with a person of a specific sex.58 The majority notes that the Supreme Court’s ruling in Loving helps us understand that miscegenation laws “are (and always were) inherently racist.”59 The majority claims that this insight applies to the sexual orientation claim at issue in Hively.60

In both lines of argument, the majority relies on methods of statutory interpretation that trace out the implications of other authoritative, precedential reasoning and that attempt to harmonize the inferential consequences of that reasoning with the present case. This tracing of a controlling, rational, precedential story through the case law is meant to undergird the court’s application of similar principles in the instant case.

2.   The Posner concurrence.

In a notable concurrence, former Judge Richard Posner proposes a different interpretive tack, which he suggests “may be more straightforward.”61 Posner asserts that statutory interpretation “comes in three flavors”: the “extraction of the original meaning,” “interpretation by unexpressed intent,” and “giving a fresh meaning to a statement.”62 Posner claims that the first of these options, “extraction,” “corresponds to interpretation in ordinary discourse.”63 The second mode, “interpretation by unexpressed intent” is discussed by reference to cases in which the meaning of a statement is informed by knowledge of the consequences desired by the speaker.64 To use William Blackstone’s example, “whoever drew blood in the streets should be punished with the utmost severity” is understood not to apply to surgeons aiding the sick.65 And finally, the third of Posner’s flavors, “giving a fresh meaning to a statement,” is discussed in relation to Hively.66 Posner explains, “Statutes and constitutional provisions frequently are interpreted on the basis of present need and present understanding rather than original meaning.”67 Posner advocates for deciding Hively by appealing to this interpretive technique of “judicial interpretive updating.”68 Posner questions the majority’s argument that sex discrimination can encompass
discrimination on the basis of sexual orientation absent such updating given that “sex” as understood at the adoption of Title VII meant biological sex, not sexual orientation.69

Posner, too, mentions some changed circumstances that warrant updating the court’s understanding of the meaning of
Title VII. He cites to Obergefell v Hodges,70 in which the Supreme Court held that there is a fundamental right to marry and referred to changing biological and social understandings of homosexuality. Toward the end of his concurrence, he writes:

The most tenable and straightforward ground for deciding in favor of Hively is that while in 1964 sex discrimination meant discrimination against men or women as such and not against subsets of men or women such as effeminate men or mannish women, the concept of sex discrimination has since broadened.71

3.   The Sykes dissent.

Finally, Hively features a dissent written by Judge Diane Sykes in the style of original public meaning originalism. Sykes agrees that the issue before the court is “one of statutory interpretation,” and she begins her opinion by stressing the logic grounding her originalist approach:

When we assume the power to alter the original public meaning of a statute through the process of interpretation, we assume a power that is not ours. The Constitution assigns the power to make and amend statutory law to the elected representatives of the people. However welcome today’s decision might be as a policy matter, it comes at a great cost to representative self-government.72

Sykes’s originalism also stresses the fixation thesis: Sykes is interested in what “a reasonable person competent in the English language would have understood” Title VII’s language to mean and cites to public, general use dictionaries’ definitions of “sex” to make her interpretive case.73 The meaning of Title VII is inseparable, on Sykes’s view, from this archaeological question of the meaning of the words at the time of enactment. Sykes further emphasizes that the meaning of Title VII’s statutory terms are fixed by “public” meanings.74 There are two ways in which this is true: meaning is fixed by reference to a member of the public’s understanding, and the statutory concept itself (“sex”) is public rather than narrowly specialized.

* * *

The opinions in Hively exemplify the legal world’s grappling with various approaches to interpreting the meaning of “sex discrimination.” Proponents of the broader meaning of sex discrimination tend to advance versions of the arguments made by Wood, emphasizing cultural and legal shifts since the adoption of Title VII in 1964 that inform the reading of the statute’s text and suggest that refusing to make sexual orientation discrimination cognizable under Title VII is inconsistent with Price Waterhouse, Loving, and Obergefell.75 Those who favor a narrow reading tend to coalesce around an originalist understanding that focuses on statutory meaning as fixed at the time of adoption and emphasizes the common public understanding of “sex.”76 Further, advocates of the originalist position may, as Sykes did, point to Posner’s concurrence in Hively to argue that, if judges endorsing a broader understanding of sex discrimination are intellectually honest, they must admit to impermissibly rewriting Title VII as they see fit.77

III.  Brandomian Inferentialism

Here are two ways of thinking about meaning.78 On a first, representationalist account, a sentence’s meaning is explained by the conditions under which it is true. This account takes truth, along with a notion of reference, as semantic primitives. What words do, according to the representationalist, is represent. That is, they stand in for parts of the world. Words can be combined such that they represent the world truly. That is, they get the parts of the world that they represent correct in some way.79 For the representationalist, reference and truth are primitive in that they name the basic or fundamental aspects of the linguistic enterprise—elements not themselves analyzable in yet more fundamental parts.80

On a second, inferentialist account, a sentence’s meaning is explained by the inferential relationships that obtain between that sentence and other sentences in the language. One way to motivate such an account, pursued by Professor Brandom, is to adopt a pragmatic approach—taking as basic an account of the act of asserting in order to explain what has thereby been asserted.81 Explanatory priority is given to pragmatics over semantics. The act of making a claim is rational—that is, takes place within the space of reasons82—if it includes appreciation of at least some of the inferential consequences associated with the act of claiming. The functional role of a sentence is thus normatively defined—asserting the sentence implies a certain responsibility on the part of the asserting subject according to norms of material inference. What is primitive, on this account, is not reference and truth but inference—the basic, necessary ingredient of any rational asserting subject within a language game.83 Reference and truth emerge from this more basic inferential unit.

The aim of this Part is to understand Brandom as providing answers to some basic philosophical questions: What is it to be rational? How do our concepts become contentful? What is the role of community and society in our discursive activity? These are, needless to say, extremely broad and contentious questions with a long history of philosophical and legal thought devoted to supplying them answers. This Comment’s goal is not to do justice to the depth of that scholarly tradition. Nor is it to attempt a novel summary of Brandomian inferentialism. Rather, it attempts to develop some general themes of Brandom’s semantics in order to see how those themes inform our thought about how words such as “sex” in Title VII should be understood.

A.    Rationality

To begin, it is helpful to understand what Brandom means by “rationality” and how he sees us, concept-wielding language
users, as distinctively rational. Brandom identifies Immanuel Kant’s thought as a tipping point in the history of philosophy because of Kant’s reconceptualization of the category of mental activity and judgment.84 Whereas prior philosophy had been concerned with judgment-as-predication—rational activity as defined by predication (categorizing objects correctly, understood as a relational activity)—Brandom reads Kant as understanding judgment as defined by responsibility (understood as a normative activity).85 That is, to be rational is not properly understood as ontological possession of a thing (a mind) that allows one to classify objects but as being responsible to a realm of deontological assessment.

A rational agent, on this view, is one that is bound by norms. Rationality involves “committing ourselves . . . [and] making ourselves subject to assessment according to rules that articulate the contents of those commitments.”86 To be rational, and to wield concepts, critically involves not just knowing what (classifying objects) but also knowing how to do something. To clarify this idea, Brandom uses the example of a parrot that is trained to respond differentially to red things by squawking “[t]hat’s red.”87 On the judgment-as-predication view, being able to reliably differentially respond to red objects might be thought to be applying the concept “red.” Not so on a normatively conceived understanding of judgment: the parrot’s behavior does not rise to the level of rational judgment precisely because it lacks a dimension of normative assessment; the parrot could not offer reasons for its classification. That is, in order to count as applying the concept “red” under the Brandomian view, a language user must not only reliably classify red things but also be able to make inferences. To make a judgment, a language user must understand the inferential relationships between the concept “red” and other concepts (being “colored,” not compatible with being “green,” etc.).88

Brandom describes this as the difference between “sentience” and “sapience.”89 Sentience is a biological trait, whereas sapience involves normative assessment.90 Sapience requires the concept user to supply reasons for what she does and to submit herself to assessment in terms of those reasons. Applying concepts, which is to say “giving and asking for reasons,” characterizes truly sapient discursive activity.91 As Brandom writes,

The space of reasons is the space of concepts. What discursive beings do is apply concepts. . . . Such discursive activity is the exercise of a distinctive kind of consciousness. . . . For it depends on the sort of conceptual understanding that consists in practically knowing one’s way about in the inferentially articulated space of reasons and concepts, rather than the sort of organic feeling we share with animals that are not rational animals.92

Brandom characterizes this understanding of judgment as Kant’s “next big idea” in the history of philosophy: viewing the normative commitment undertaken through judging as a “task responsibility” to do something.93 Specifically, Brandom thinks that we can characterize discursive activity as the undertaking of commitments—taking on a rational responsibility to provide reasons for one’s actions (justification), to acknowledge inferential consequences of those reasons (amplification), and to root out incompatibility among commitments undertaken (critical activity).94 A concept user, “a rational self,” can be seen as a web of interactions among commitments undertaken, which in turn entail and preempt the endorsement of other commitments.95
Brandom, following Kant, refers to this integrative, reweaving, rational agent as an “original synthetic unity of apperception” and distinguishes the discursive, deontic activity of apperception from mere labeling, perceptive activity.96

For Brandom then, the paradigm of rational activity is concept application, which involves a responsibility to integrate the concept into a “constellation” of related concepts connected to each other inferentially.97

B.    Social Linguistic Activity

This integrative activity extends not just to an individual’s conceptual commitments but also serves as the foundation for a general account of linguistic activity. Justification, amplification, and critical activity all structure the apperceiving individual’s rational activity. Taking this constellation into the interpersonal realm, shared among language users, provides the basic structure of Brandom’s linguistic account.

To illustrate, Brandom describes linguistic practice as a certain kind of game.98 Imagine a game with a basic structure that involves a gameplayer and a scorekeeper. The game involves a large number of different token types available to be played by the gameplayer. The gameplayer makes a move by taking one of the token types and placing it in front of her. Think of this as undertaking a commitment to that token. This activity is noted by the scorekeeper, who records the commitments undertaken by the gameplayer. These commitments can also be disavowed: the gameplayer does this by removing the relevant token from in front of her.

Suppose further that various inferential relationships hold among the tokens. This complicates the job of the scorekeeper, who notes not only those tokens that the gameplayer has played directly (her commitments), but also keeps track of those tokens that are permissible and incompatible with the commitments of the gameplayer. That is, a gameplayer’s playing a token of a given type will rule out the possibility of legitimately playing some other tokens—one cannot be committed to the former while also being entitled to play the latter. Further, the scorekeeper can note those tokens that the gameplayer could play while remaining consistent with her current commitments. We could think of those tokens as ones that the gameplayer is entitled to. Entitlement is not a basic move that the gameplayer can make herself. Instead, entitlements flow from the more basic commitments that a gameplayer undertakes. Entitlements can be the direct result of a basic commitment or flow permissively from yet other entitlements. Furthermore, a gameplayer with commitments judged to be incompatible will correspondingly be judged entitled to neither. A scorekeeper can challenge these commitments, calling on the gameplayer to resolve the incompatibility or face normative sanction.99

These inferential connections will start to map out the “space of reasons”100 for an individual gameplayer. But the game as it is described above is also interpersonal. It involves individual gameplayers keeping score on one another. This perspectival nature101 means that the act of scorekeeping may vary subtly based on the perspective of the scorekeeper in question and what commitments, entitlements, and incompatibilities she associates with a given token. Further, gameplayer and scorekeeper are roles each participant in the game plays at the same time by giving and asking for reasons.

Because gameplayers are also scorekeepers (not only of others, but of themselves, keeping track of their own commitments and entitlements), the interpersonal act of scorekeeping may have consequences for the intrapersonal moves one makes as a gameplayer. If one keeps score on a gameplayer, and the player makes a move to which she is both committed and entitled, and there are no incompatible commitments held by the scorekeeper, the scorekeeper gains an entitlement to the claim.102 This is because ascriptions of entitlements always occur from the scorekeeper’s perspective. That is, in tracking a gameplayer’s entitlements and incompatibilities, the scorekeeper has to rely on her judgment, on what she takes the inferential consequences of the actual commitments of the gameplayer to be. Crucially, the shape of those acknowledged relationships will vary from scorekeeper to scorekeeper.103 Indeed, because gameplayers also keep score on themselves, a scorekeeper will likely recognize some consequences of the tokens she plays but also fail to recognize others.104 The exact shape of the entitlement and incompatibility consequences that result from a gameplayer’s tokening will vary depending on the identity of the specific scorekeeper in question.

C.    Determinateness and Changing Meaning over Time

The story that Brandom tells also seeks to explain what he calls the “determinateness” of concepts.105 Because the specific shape of a gameplayer’s commitments (her entitlements and incompatibilities) depends on the judgment of the scorekeeper, we might start to worry about the move from the actual practices of scorekeepers to a determinate content that can be associated with a concept. For one, inferential role is intelligible only when a specific scorekeeper updates the score. For another, regardless of how many instances of scorekeeping activity there are, they will underdetermine the number of material inferential consequences that are appropriately associated with a given tokening.106

The activity of updating the score associated with a given tokening by a gameplayer might differ from scorekeeper to
scorekeeper based on the auxiliary sentences that the scorekeeper considers when assigning commitments, entitlements, and incompatibilities. To the extent that two different scorekeepers perform their activities differently in response to a tokening by a given gameplayer (that is, a concept deployed by a gameplayer), does the tokening have two different meanings? If it does, we might worry about the implications for the possibility of communication among participants in the game: if the meaning of a sentence differs based on all of the associated connections between the sentence and other sentences, then it seems unlikely that any two people will ever mean the same thing when they use a given tokening.107

This is one big takeaway from how Brandom says we should think about meaning: the challenge above seems to want to describe something concrete—a conceptual content—that each participant in the linguistic practice has access to individually.108 But Brandom says that meanings are not little packets, all alike, to be copied and distributed to each of the members of a linguistic practice.109 Instead, communication is a social enterprise that succeeds because each participant in the enterprise does something different. Like partners in a dance, the communication emerges from the coordinated—but different—activity of the participants.110 While the partners in communication may not “share” some fixed and determinate thing (a meaning), their interaction counts as successful communication insofar as they can coordinate their activity together. This type of coordination need not require an exact match between the inferential consequences of concept application recognized by each scorekeeper/gameplayer.111

Brandom also endorses semantic externalism. What counts as a correct “move” in the communicative dance depends on standards that are external to the judgment of any individual participant. The gameplayer, when she undertakes a commitment, is thus meaningfully accountable to normative conceptual standards. Brandom writes:

The norms I am binding myself to by using the term “molybdenum”—what actually follows from or is incompatible with the applicability of the concept—need not change as my views about molybdenum and its inferential surround change. And you and I may be bound by just the same public linguistic and conceptual norms in the vicinity in spite of the fact that we are disposed to make different claims and inferential moves. It is up to me whether I play a token of the “molybdenum” type in the game of giving and asking for reasons. But it is not then up to me what the significance of that move is.112

Through this notion of external, public standards of correct or incorrect application of conceptual norms, Brandom is trying to carve out a sense in which inferentialism can still be held compatible with a sort of determinate stability. With “molybdenum,” this sort of objectivity is achieved by making the standards for correct or incorrect application independent of the views of any single gameplayer/scorekeeper and resting them instead on an independent expert consensus. In this way, a language user who employs a given concept can still do so correctly or incorrectly even if she does not recognize an incorrect application when keeping score on herself.

While the norms structuring a given account of conceptual content provide a sort of objectivity, those norms can clearly change and evolve over time. Current participants in the language game can look back on past applications of concepts and past judgments surrounding inferential consequences and trace through that history a narrative about the norms that structure a concept. By acknowledging some applications as correct while sanctioning others, current participants both make the concepts in question determinate and find them to be so through their interpretation of past use, knowing that they stand by this history of concept use just as future generations will stand by them.
Using “copper” as his example, Brandom explains:

A recollective reconstruction of the tradition culminating in the current set of conceptual commitments-and-contents shows, from the point of view of that set of commitments-and-concepts, taken as correct, how we gradually, step by step, came to acknowledge (in our attitudes) the norms (normative statuses such as commitments) that all along implicitly governed our practices—for instance, what we were really, whether we knew it or not, committed to about the melting point of a piece of metal when we applied the concept copper to it. From this point of view, the contents of our concepts have always been perfectly determinate . . . though we didn’t always know what they were.113

Brandom also makes this point by appealing to a specifically legal analogy. For, on his view, this social-integrative activity is exactly the sort of activity undertaken in common-law judging. Brandom argues that, in the common law, “[a]ll there is to give [concepts] content is the actual applications that have been made of them over the years. They are case law all the way down.”114 In judging a case at common law, the judge has only the precedential decisions of past common law cases to consult. Those precedents have shaped concepts relationally, defining the contours of the applicability of legal concepts by the inferential logical relationships that hold among cases. Brandom describes the process by saying,

The task in each case is to decide the applicability of some distinguished legal vocabulary. . . . The judge in each new case makes a decision, to apply or not to apply the legal concept in question, given the facts of the case. . . . In this process, each new decision, with its accompanying rationale, including a selection of precedents, relevant considerations, and rules of inference and incompatibility, helps to determine further the conceptual content of the legal term whose application is up for adjudication.115

The judge in these cases operates in the context of the current institutional web of permitted legal inferences. The judge’s application of a concept in new circumstances also serves to force the web to reweave and accommodate the new inferences that follow from a given application of a concept. It thus helps to shape the future practices of the legal community that will, in adjudicating future cases, incorporate the current judgment as precedential.

In this way Brandom sees the rational integration of concepts as not only socially defined by linguistic communities but historically defined by those communities over time.116 Just as the current members of a community recognize each other as fellow discursive language users, so too they recognize past instantiations of the community as members having something like precedential authority. To justify the application of a concept is to offer a “rational reconstruction of the tradition that makes it visible as authoritative insofar as, so presented, the tradition at once determines the conceptual content one is adjudicating the application of and reveals what that content is, and so how the current question of applicability ought to be decided.”117 The entire process of integrating rational commitments into a “constellation of prior commitments”118 is based in this sociohistorical process of normative attitudes.

This process implicates “determinateness” in two different ways: scorekeepers look retrospectively to past uses of concepts to inform their present judgments but are also aware that present judgments prospectively affect the norms that will inform future judgments.119 Brandom emphasizes:

From this point of view, conceptual norms are never fully determinate . . . since there is always room for further determination. The conceptual norms are not completely indeterminate either, since a lot of actual applications have been endorsed as correct by potentially precedent-setting judgments. All the determinateness the content has is the product of that activity.120

While this may seem to give a sort of privileged authority to the scorekeeper, this privilege is also administered by future participants in the game. Those participants will also, when keeping score, have to decide whether to take the judgments of past scorekeepers as authoritative. This is analogous to the fact that the common-law judge has the authority to determine the current inferential shape of the legal precedent she inherits (though that determination remains authoritative only if the next judge, when faced with the same issue, chooses to treat the decision as correct or precedential). Future scorekeepers have the ability to recast the judgments made by current scorekeepers as errors or to accept them as correct.121

IV.  Toward a Brandomian Statutory Interpretation

This Comment argues that dynamic interpretive approaches map well onto Professor Brandom’s semantics. These dynamic approaches to interpretation can countenance broader conceptions of sex discrimination by acknowledging the inferential pull of cultural attitudes, intervening Supreme Court decisions, and better understanding of gay and lesbian people’s experiences. This sort of inferential analysis fits well with a general semantic understanding like Brandom’s that emphasizes meaning as emerging from dynamic apperceptive activity (deontic scorekeeping). Further, Brandom’s semantics suggests that originalist interpretations undergirding narrow understandings of Title VII are misguided in two ways: First, such understandings improperly focus on statutory meanings that are found rather than made. Second, and relatedly, the originalist arguments presented in the debate surrounding Title VII suggest a rejection of the sort of semantic externalism typical of Brandom’s thinking. Brandom describes the process of semantic development as one of discovering those norms that all along structured our conceptual contents, whether or not we realized it at the time.122 This is the view that Judge Posner in Hively derides as “imply[ing] that the statute forbade discrimination against homosexuals but the framers and ratifiers of the statute were not smart enough to realize that.”123

To clarify, this Comment seeks to explore the ways in which a Brandomian understanding of semantics can be used to support dynamic interpretive methods and to keep at bay concerns about such methods being unprincipled or mere power plays. That is, it argues that Brandom’s account gives the dynamic interpreter one story that she might tell about how the meaning of statutory terms can change over time. It is beyond the scope of this
Comment to attempt any sort of evaluative or comparative semantics—that is, to argue that Brandom’s semantics is in fact the best account of language that we currently have and that our interpretive activity should be brought into conformity with
Brandom’s account as a result. Of course, to the extent that is the case, that provides us with an independent reason to prefer dynamic interpretation to forms of originalism. But even if we are agnostic as to which semantic account provides us with the best way of talking about language, Brandomian inferentialism can give us tools to help build up dynamic statutory interpretation into an internally coherent account with sufficient conceptual resources to respond to various criticisms from originalists and to give force to some criticisms of its own.

A.    Brandom contra Originalism

At first glance, there appears to be a clear conflict between the originalist and Brandomian approaches. Originalism emphasizes semantic fixation by appealing to a stable source of conceptual meaning. Brandomian inferentialism, by contrast, emphasizes the pragmatic fluidity of scorekeeping activity—tracking inferential connections among sentences. What semantics is for, in other words, is making explicit the normatively enforceable commitments of a certain group of gameplayers rather than primarily for speaking truly or representing things in a certain way. Commitment to one sentence by a speaker/gameplayer becomes meaningful when it is recognized by a scorekeeper who updates the score accordingly: noting the commitment, permissive entitlements, and incompatibilities that result from the move. Scorekeepers are thereby not conceived of as merely uncovering and observing those meanings imposed on terms by their speakers (as in the archaeological originalist account); rather, the activity of scorekeeping necessarily involves some judgment about inferential consequences that puts the scorekeeper actively at the heart of semantic content.124

But perhaps a reconciliation is possible. Scorekeeping activity necessarily depends on those auxiliary hypotheses accepted by the scorekeeper. It is those hypotheses that determine, based on the acknowledged commitments of a given gameplayer, what the further committive consequences, entitlements, and incompatibilities associated with a given score will be. Prohibiting commitments that would not have been accepted by an average member of the linguistic community at the time of a given statute’s adoption may be thought to constrain an originalist judge’s scorekeeping. This is not a necessary feature of judicial interpretive activity—the shape of the judge’s auxiliary commitments could be otherwise so as to remove this constraint if the judge was not committed to a conservative scorekeeping heuristic. But perhaps we ought to have this constraint in place for reasons of democratic government: the constraint improves the publicity of laws, makes law more objective, and ensures that laws’ contents are determined by elected legislators.125

This Comment does not argue that it can never be appropriate to employ these kinds of conservative constraints on scorekeeping activity. Indeed, such a claim would be implausible on its face. Someone seeking to understand Paul’s epistles, for example by engaging with concepts like “justification” or “the law” as they are used in the letters, will surely need to adopt auxiliary scorekeeping hypotheses that recognize only those consequences that would have been accepted by a Pauline Christian.126 (Keeping score on Paul’s epistles based on the sort of understanding of “the law” that includes Title VII, for example, is likely inappropriate.) There, the biblical scholar really is doing a sort of finding and adopts some hermeneutics that reflects the desire to uncover the original meaning of the letter as intended by Paul.

But importantly, note that the scorekeeping constraints adopted in this sort of conservative interpretation are optional constraints self-imposed by scorekeepers with a particular goal (for example, understanding what Paul intended in the epistles). Such constraints do not run the other way—there is nothing about the text itself or something external to the scorekeeper that would force the adoption of a conservative scorekeeping heuristic. This Comment does not argue that a conservative heuristic might not also make sense for a certain kind of legal reader; it argues only that the question of whether it does or not must be answered at the contested level of utilitarian calculus.

Judge Sykes seems to acknowledge this in her Hively opinion. She frames originalism as a consequence of a certain kind of rule-of-law concern. She argues that originalism is based in the structure of our democracy and Constitution, which is consistent with originalism being a contingent, optional approach to legal interpretation.127 But Sykes, despite her initially more utilitarian framing of the originalist argument, overplays her hand when she writes of the relevance of the “robust debate” surrounding treatment of gay and lesbian people in society:

This striking cultural change informs a case for legislative change and might eventually persuade the people’s representatives to amend the statute to implement a new public policy. But it does not bear on the sole inquiry properly before the en banc court: Is the prevailing interpretation of
Title VII—that discrimination on the basis of sexual orientation is different in kind and not a form of sex discrimination—wrong as an original matter?128

For the inferentialist, there is no reason to think that considerations such as original understanding necessarily do not bear on the question of statutory interpretation before the court. And that’s because the court’s interpretation is a choice of interpretive frames, of recognition or nonrecognition of certain auxiliary hypotheses that give the concept “sex discrimination” its inferential shape. Present cultural attitudes and original public meanings are both, equally, examples of the sorts of auxiliary hypotheses that can give our understanding of sex discrimination a particular inferential shape. There is no way to suppose that one is properly before the court while the other is off limits without begging the interpretive question at issue.

Moreover, an originalist heuristic runs counter to the semantic externalism urged by Brandomian inferentialist semantics. Sykes’s originalism points to democratic legitimacy as an important normative constraint on the scorekeeping activities of judges. One effect of the adoption of an originalist scorekeeping heuristic is that the concepts being interpreted are isolated, meaningfully sealed off from the ordinary public realm of concept use due to their specialized attendant scorekeeping considerations. I take this to be what happens with something like the exegesis of Paul’s epistles considered earlier: the notions of justification or law employed there are local to the epistles and to a certain biblical frame of reference. Concepts employed in that context, then, need not be broadly public but instead are specialized. We can use subscript to denote this sort of local conceptual concern such that someone reading the epistles is interested in
lawPaul rather than “law” as a public concept. Originalist statutory interpretation, however, is explicitly not confined to the local realm of legal concepts and judicial interpretation. Originalist judges take themselves to be applying public concepts and so cannot confine their concern to the local inferential realm of specialized legal use of the concepts in question.129

The recollective process that Brandom describes with regard to “molybdenum”130 can equally be applied to Title VII and “sex.” Just as we historically may not have realized what we were committed to when we applied the concept “molybdenum” (for example, that it entails melting at 4,752 degrees Fahrenheit),131 legislators in the 1960s may not have realized what they were committing themselves to by outlawing discrimination “on the basis of sex.” But as we gradually came to acknowledge, through attitudes that treat discrimination on the basis of sexual orientation as a form of the sort of sex stereotyping that is at the heart of sex discrimination, the consequences of a commitment not to discriminate on the basis of sex became clear. In other words, we can be thought to have made progress in understanding what application of the concept “sex” in Title VII entails, in just the same way that metallurgical experts’ progress in understanding the properties of copper entails an evolving set of inferential consequences of application of that concept.

A gameplayer can also count as having made use of a concept even if she does not grasp or endorse all of the inferential consequences of the concept in question. As Professor Jeremy Wanderer describes, one can analogize this to playing a game of soccer with a child.132 A child who merely runs around a soccer field but who shows no interest in the ball and seems to have absolutely no grasp of any rules of soccer will not be taken as making any soccer-
playing moves. But at some point, even a rudimentary grasp of the concept of soccer will be judged as soccer playing by observers.133 That is, if the child knows to kick the ball, not to handle it with her hands, and that the touch lines mark the boundaries of the field, etc., then the child’s activity can genuinely count as playing soccer. Furthermore, this is true even if the child does not subjectively endorse all sorts of inferential consequences of soccer playing: that the ball cannot be passed to a teammate in an offside position, for example. Observers (scorekeepers) who take the child to be playing soccer but who themselves have a better knowledge of the rules and requirements of the game will recognize inferential consequences of the child’s soccer-playing activity that extend beyond the child’s own grasp of the concept.134 Nevertheless, we can still think of the child soccer player’s activity as involved in a determinate, objective enterprise of soccer playing.

As Wanderer describes, the child partakes of a concept that is external to her, with implications that exceed her grasp.135 We might think that future metallurgical experts, or future legal scholars, stand in much the same relation to us that the more soccer-knowledgeable observers stand in to the child soccer player. From that as-yet-unspecified future perspective, we likely make mistakes about how we apply the concepts “molybdenum” or “sex discrimination.” We likely endorse things that we will come to recognize as wrong and fail to see inferential implications of those concepts that appear obvious to future observers. But we can still count as applying, making use of, the same public concepts.136 We need not say that the child soccer player is actually playing soccerchild while mistakenly being evaluated by reference to the inferential standards of regulation soccer. But that means that the inferential consequences of a commitment not to discriminate on the basis of “sex” are similarly not up to us. They evolve as our perspective changes and as we move from our childlike grasp of the concepts in question to the perspective of more mature observers.

There is also good reason to think that “sex” as discussed in originalist judicial opinions on Title VII cannot be a particular sexlegal, a concept different in kind from our ordinary term “sex.” Sykes appeals explicitly to “original public meaning” and “common, ordinary usage” and cites to public, nonspecialized dictionaries.137 Posner discusses the “three flavors” of statutory interpretation in connection with “ordinary discourse,” and he concedes that “even today if asked what is the sex of plaintiff Hively one would answer that she is female or that she is a woman, not that she is a lesbian.”138 Both Sykes and Posner thus connect their reasoning with ordinary, public understandings of the concept in question. This sort of appeal to public understandings implies that our understanding of correct or incorrect application of “sex discrimination” in Hively ought to follow our general understandings of how that concept becomes contentful. Brandomian inferentialism suggests that “sex discrimination” has content by virtue of its inferential consequences. If Sykes and Posner meant to apply a more narrowly localized understanding of “sex discrimination” (perhaps one held fixed by originalist scorekeeping heuristic), then their emphasis on ordinary concept use is misplaced. On a Brandomian understanding, ordinary concepts are ones that (may) have inferential consequences that exceed our current grasp of the concept, and thus there is nothing in principle incompatible with our appreciation of the inferential consequences of a given concept changing over time. Indeed, perhaps the only way to be sure that a concept’s meaning did not change over time would be to define it in some narrowly local way, but that approach contrasts with Sykes and Posner’s presentation of their analysis as ordinary or public.

It is certainly easier to animate a more localized, narrow understanding of the concept in question for a legal issue that involves something that is unambiguously a term of art—like the discussion of “recess appointments” in Noel Canning.139 There, it is pretty clear that the game being played, the concept being deployed, was created by and for lawyers engaged in legal reasoning. Because the term is so clearly localized and because it makes no real claims on broader, more public concepts, it becomes harder to tell a story about how the true inferential reach of the concept exceeds the grasp of those who invented and use it. But it is not impossible to do so. The concept is determined by the inferential connections that link it to other concepts. These connections are clarified through use—actual applications of the concept and judgments about its use in certain situations as proper or improper. So the meaning of the concept will almost certainly be clarified and modified as these inferential connections are made clear. There may be local normative constraints on scorekeeping activity that inform how those judgments are made, but by definition on an inferentialist model the actual meaning of the concept will undergo change as it is applied. Even on this localized understanding, meaning still emerges from the relationships in the apperceptive inferential web of concepts—no concept is truly an island.

Moreover, it is harder to tell this kind of story for Title VII, whose drafters chose to use concepts like “sex” with very public meanings and uses rather than to expressly define a legal term of art. For this reason, a localized legal reading loses force as applied to Hively in particular.

In sum, Brandom’s inferentialism shares a general affinity with the sort of dynamic interpretive techniques used to support broader understandings of sex discrimination. While those opinions may not draw explicitly on these semantic ideas, Brandom provides one possible way to tell the dynamic-meaning story without resorting to “judicial interpretive updating” rationales. Brandom’s inferentialism also gives us reasons to think that originalism is not a built-in feature of our activity as concept users and that, as an optional normative constraint on scorekeeping activity, it gains force only to the extent that it is a useful vocabulary for helping us get what we want—in other words, originalism is swallowed up by a broader pragmatism. This contrasts with the standard originalist view—that original public meaning or some similar source of evidence provides a semantic, as opposed to merely pragmatic, constraint on interpretive activity.140

Brandom’s philosophy provides us with reasons to think that, to the extent that law makes use of public concepts, those concepts will evolve in accord with the scorekeeping activity of ordinary concept users. New applications of terms, new judgments about the correctness or incorrectness of the inferential implications of the concept in question, will affect the meaning of those concepts that the law deploys. This is also true for local legal terms of art, but it takes on heightened significance for broader public concepts given the array of scorekeeping activities that have the potential to affect those concepts’ inferential role.

B.    Democracy and Constraint

One concern we might have about a Brandomian interpretive method is how it interacts with democratic values. As Sykes explains in her Hively dissent, we might worry that judicial recognition of a changed statutory meaning is best thought of as legislation from the bench.141 This is why Sykes decries the majority’s ruling in Hively as (borrowing Posner’s phrase) “[j]udicial statutory updating” that “assume[s] the power to alter the original public meaning of a statute through the process of interpretation, [which is] a power that is not [the court’s].”142

But Sykes’s criticism misses the mark as applied to the
Brandomian conception of judicial interpretive activity. What Sykes criticizes is an activist judiciary, one that actively “updates” or “rewrites” statutes according to its own preferences. But this one-sided conception of the Brandomian interpreter is a caricature. Rather than privileging judicial “making” of the meanings of legal concepts over Sykes’s preferred “finding,” the Brandomian picture suggests that this dualism is inadequate—judges (and all scorekeepers) are always both finding the concepts they employ to be rationally determinate through elaboration of a pattern of appropriate and inappropriate use and making them so through the process of tracing out a recollective reconstruction of the concept’s content. Judges in this picture are still beholden to standards of correct and incorrect application of concepts that are formally determinate, and those standards are informed by the patterns of licensed and sanctioned application of concepts that the judge recognizes. Judges in this way remain constrained on the Brandomian picture, but the constraint is exercised by a broader class of information (the recognized web of inferential connections that hold among legal concepts) than on Sykes’s original public meaning originalism picture of interpretive activity.

Perhaps the originalist will respond that constraint by the recognized web of inferential connections that hold among legal concepts fails to constrain much at all. Indeed, as Part IV.C discusses, the Brandomian account as this Comment understands it does instruct the judge to look to a much broader class of information than traditional theories of interpretation. But the fundamental point remains that judges on the Brandomian picture are not actively substituting their own preferred legal concepts for those to be found in the statutes enacted by Congress. Rather, judges still try to remain faithful to the concepts employed by Congress while at the same time recognizing that those concepts are themselves given semantic content by the real-world process of application by language users. The Brandomian approach here is no less democratic than the original public meaning originalist’s—neither approach gives special pride of place to the intentions or thoughts of individual democratic actors. In both approaches, the text that is actually enacted controls, regardless of its democratic popularity.143 But the Brandomian acknowledges that the concepts employed in statutes, while controlling, are themselves controlled by actual linguistic activity—the approved and sanctioned applications of the concept that provide the material from which the Brandomian scorekeeper rationally reconstructs the concept’s meaning. To return to an earlier example, the Brandomian evaluating the meaning of a child’s playing soccer is bound to recognize that playing soccer entails respecting the offsides rule regardless of the interpreter’s views on the desirability of such a rule. The public nature of the concept in question thus works both ways: just as children playing soccer are bound by offsides whether they acknowledge it or not, so too the scorekeeper evaluating soccer playing is bound to acknowledge the offsides rule as binding. The bindingness of offsides is up to neither an individual gameplayer nor an individual scorekeeper.

Recognizing the necessity of fidelity to actual concept use is part of a general shift to a world in which “semantics must answer to pragmatics”144—where semantics helps make sense of the actual actions of language users. The bedrock of Brandom’s inferentialism is actual linguistic practice: how concepts are actually deployed in language. It is from this actual use of language that we get the circumstances of correct and incorrect inferential consequences that give inferentialism its force. The web of inferential connections that is constitutive of inferential meaning is not static—it does not dictate terms to the use of concepts always and forever. Rather the web is dynamic, reweaving to accommodate the significance of new information supplied by new licensed or unlicensed deployment of concepts in particular circumstances.

And if this is our model of conceptual change, then judges, not legislators, are the institutional actors best situated to do the incremental, sensitive work of tracing out inferential commitments. Such activity, consistent with Brandom’s legal analogy, looks like the sort of common-law concept elaboration that judges engage in rather routinely. Sykes, in Hively, wants to draw a line between “common-law statutes” like the Sherman Act and ordinary judicial work that is “interpretive only.”145 But Brandom’s general account of conceptual activity suggests that it makes no sense to attempt to draw such a line: all concepts are to be seen as responsive to the real-world implications of application to new circumstances and of changes in what is taken among language users to be proper and improper use of the concept. The only difference is the speed with which the community applies shifting norms of correct or incorrect concept application, but the judiciary remains sensitive to the changing circumstances of application regardless. That is, it may be that some concepts, as deployed by language users on the ground, change more quickly than others in terms of what usages language users recognize as correct or incorrect, but all semantic content will be sensitive to and shift along with those on-the-ground changes.

Sykes ends her dissent by asserting that “[t]he court’s new liability rule is entirely judge-made; it does not derive from the text of Title VII in any meaningful sense.”146 Applying Brandom, dynamic statutory interpretation can have it both ways in just the sense that Sykes denies here. Brandom shows us how to think of judges as in some sense “making”—actively tracing out a rational reconstruction that reveals the inferential connections of a concept in question—but also how to think of the meaning that results as faithful to the statutory text. And that occurs because the statutory text itself, its semantic content, must further be faithful to the pragmatic doings of the relevant linguistic community. The applications of the concept in question, the real-world judgments about proper or improper application, are themselves the fodder that constitutes the meaning of a concept. And a judge’s careful attention to the changes in those applications is therefore faithful to the text in that it recognizes that the text and the concepts it employs must answer to the world.

When we learn more about “copper” (when we recognize new consequences of applying the concept “copper”), a scorekeeper’s inferential accounting of sentences involving the concept will change to accommodate that progress. So too, when we learn about a concept like “sex discrimination,” a diligent scorekeeper will update her books accordingly. Learning more about copper or sex discrimination is just recognizing new inferential consequences of deploying those concepts, recognizing new applications of those concepts as appropriate or inappropriate. Thinking of judges as faithful to text therefore need not mean cutting judges off from the changing thoughts and behaviors of the very linguistic community whose activity gives concepts their inferential shape. Legal concepts will not achieve this sort of faithfulness to the pragmatic doings of individuals all on their own: there is no self-enforcing mechanism to ensure that changes in how a linguistic community thinks about a concept will be applied in a particular case. That requires the diligent work of a scorekeeper sensitive enough to those changes to recognize them in her own scorekeeping activity.

C.    “Doing” Brandomian Interpretation

For all their flaws, one of the benefits of the most popular forms of originalism is that they give relatively clear instructions to the interpreter. Original public meaning interpretation entails a discernable task: try to figure out what an individual using a particular word at a particular point in history is likely to have meant by it. Purposivism, too, gives the interpreter relatively clear marching orders: understand the motivating purpose behind a statute and act consistently with that animating spirit.

Brandomian semantics, by contrast, asks the interpreter to do something much messier and much less amenable to clear instructions: consult and understand the web of permissible and impermissible inferential connections that obtain among concepts and apply concepts so as to be consistent with that understanding. There is, consequently, no silver bullet, no magic source to guide the task of Brandomian interpretation. There is only the attempt to reason well about concepts and to make a web of concepts cohere with the ways that concepts are actually deployed, licitly and illicitly, in the world, by language users.

So if, over the years, we as language users begin to acknowledge that there is such a thing as workplace sexual
harassment, that women can be attracted to women, that those same women have a right to marry women, that biases surrounding sexual orientation are just as engrained and pernicious as other sex or gender stereotypes, and a hundred other, more nuanced things besides, then the base of examples from which the Brandomian draws her interpretive lessons will have changed. These changes in how people deploy concepts, and which applications of concepts in which situations they recognize as appropriate, will change the meaning of those concepts themselves. Those changes give the concept a new inferential shape with which the Brandomian will try to make her interpretation cohere.

An example may help to clarify. Consider another case related to how the law affects gay people: Morrissey v United States.147 The case involves a homosexual male taxpayer who sought to deduct as medical care expenses the costs related to in vitro fertilization (IVF).148 Professor Joseph Morrissey, the taxpayer, chose to pursue the costly IVF treatments because he wanted a child but was unable to conceive one with his chosen partner.149 The case turned on a matter of statutory interpretation—the correct reading of 26 USC § 213(d)’s definition of “medical care” as “amounts paid . . . for the diagnosis, cure, mitigation, treatment, or prevention of disease, or for the purpose of affecting any structure or function of the body.”150 The question in the case was whether the IVF treatments were for the purpose of affecting a “function” of Morrissey’s body.151 The Eleventh Circuit cited dictionary definitions of key terms,152 looking to the “plain meaning” of the statutory text. It concluded that the “function” of Morrissey’s sex organs was to produce healthy sperm.153 Because the IVF treatments did not affect that function, the deduction was denied.154

For an inferentialist, the scope of this inquiry can be broadened because the question “What is our best understanding of what it means for a biological system to have a function?” admits more nuance and a greater body of evidence than do the questions “What do contemporaneous dictionaries define ‘function’ to mean?” or “What does the legislative history reveal the legislators who passed this statute to have understood biological ‘function’ to mean?” The relevant question of what constitutes a “function of the body” may perhaps be informed by dictionary definitions, but the overarching goal is to apply our best understanding of the concept “function” in the context of human biology. For that, we can look not only to dictionary definitions but also to philosophical and scientific thought that attempt to systematically describe what it is for a biological structure to have a “function” or to function properly.155 Furthermore, such robust engagement may indeed compel the same conclusion reached by the court in Morrissey. But the point here is to emphasize that interpretation on the Brandomian view is not reducible to neat and tidy reference to a single type of evidence. It requires broad engagement with the circumstances of appropriate application of a concept like “biological function” that goes beyond definition of “function” from Webster’s. That engagement is likely to bring in leading work from biology, philosophy, sociology, or any number of other disciplines that define the norms of appropriate application of the concepts used in statutory language.

Here, originalist fears about a lack of constraint on judicial interpretive activity likely emerge. If interpretation requires such a broad and varied inquiry, there would seem to be a lot of room for judges to steer that inquiry toward favored outcomes. But if we are willing to assume bad faith on the part of judges, then it seems just as possible to worry that there is room in consultation of original sources, or of legislative records, for judges to steer originalist interpretive inquiries in just the same way.156 Of course, even if judges are not operating in bad faith, they may nevertheless simply be bad at doing the work of Brandomian inferentialist scorekeeping (or at least worse at the Brandomian inferentialist task than they are at originalist tasks). But we already ask judges to do all sorts of complicated and varied things when exercising their judgment or interpreting statutes. Judges’ experience consulting a wide array of evidence ought to give us comfort that they are up to the task of carefully tracing inferential consequences among a wide array of concepts in the way that Brandom’s scorekeeping requires.

The opinions in Obergefell, for example, acknowledge changed attitudes toward gay and lesbian people in just the way that inferentialism suggests that they ought to take note of how concepts are being applied on the ground.157 Judges are also asked to review cost-benefit analyses,158 evaluate scientific evidence,159 consider psychological claims,160 and confront challenges posed by advancing technology.161 We also ask judges to evaluate the competing empirical claims of litigants and their amici, who may cite contradictory or false scholarship.162 It is not clear that asking judges to do careful historical analysis to identify “original public meaning”163 is any less demanding than asking judges to carefully evaluate other sources and the present use of concepts in question. While the task of originalist interpretation may be easier to state succinctly, it requires substantial and rigorous investigation. Judges are capable of careful and detailed thought about a given statutory term or legal concept.

This Comment suggests that, consistent with a Brandomian view, statutory analysis should be directed at sources that go beyond those contemporaneous with a statute’s passage. There may be a range of relevant evidence to consider when evaluating the present inferential shape of a given concept (such as the subsequent judicial rulings and evolving understanding of sex and gender that motivated the Hively majority’s analysis), and it is difficult in advance of any specific question to identify what sort of evidence will be relevant to an inquiry about what constitutes the best modern understanding of a legal concept. Perhaps, then, Brandomian statutory interpretation would not be easier than more familiar originalist methods, but any method of statutory interpretation will require rigor and sophistication of judicial interpreters. What’s more, Brandom’s semantics gives us an account for how a judicial finding of changed statutory meaning can be more than an unconstrained and unprincipled exercise of power.

This lack of a user-friendly “method” for Brandomian interpretation is a drawback compared to more familiar interpretive methods, but I want to close by mentioning one comparative advantage. What Brandomian interpretation lacks in simplicity it compensates for in hope. Brandom’s semantics gives us a skeleton on which we can build a meaningfully and rationally dynamic interpretive theory. It grounds dynamic interpretation in such a way as to keep at bay worries about judges merely “rewriting” statutes, merely exercising power in law’s name. It tells a story about how the meanings of statutes respond to changes in the world. And that story implies that some as-yet unrealized changes in the behavior of language users (citizens) can force legal concepts into conformity with a web of inferential consequences hitherto unseen. And that means that there is hope for that future web to be more beautiful and more just than any we have yet imagined.


Law, like all discursive (concept-wielding) activity, must reckon with semantics. The semantics developed in this Comment is normative and rational. That normativity is spelled out in terms of a social, temporal practice of linguistic gameplaying—of asserting and scorekeeping in communities and through time. That rationality is elaborated in inferential terms—fundamental to the account are the inferential consequences that constitute the space of reasons that give concepts meaning. Thinking about semantics this way yields a view of concepts as dynamic rather than fixed, containing content that exceeds our grasp of that content, and created both by the making/authoritative activity of language users and the finding/rule-governed administration of that authority by other and future scorekeepers. It gives us a way to think of Judge Posner’s three flavors not as discrete interpretive options but potentially as aspects of a broader, multifaceted discursive practice that is the Brandomian inferentialist game of giving and asking for reasons. Judges, like all concept-users, are participants in this game, and this Comment suggests that reasoning through the structure of that game gives us a new perspective on their dual role within it.

  • 1. Hively v Ivy Tech Community College of Indiana, 853 F3d 339, 343 (7th Cir 2017).
  • 2. 853 F3d 339 (7th Cir 2017).
  • 3. 42 USC § 2000e-2(a). The text of Title VII prohibits employers from discriminating against employees “because of such individual’s race, color, religion, sex, or national origin.” 42 USC § 2000e-2(a)(1), 2(a)(2).
  • 4. Pub L No 88-352, 78 Stat 241, codified as amended at 42 USC § 2000e et seq.
  • 5. Hively, 853 F3d at 343.
  • 6. Brandom himself suggests as much:

    It is clear that this model is getting at something important about case law (and about common law, which is case law all the way down). . . . [But t]he model itself provides no more than a portmanteau formulation; it sketches only the form of an account. . . . Important points are being made, but what is offered is hardly a theory—it is more like a set of reminders of questions to ask.

    Robert B. Brandom, A Hegelian Model of Legal Concept Determination: The Normative Fine Structure of the Judges’ Chain Novel, in Graham Hubbs and Douglas Lind, eds, Pragmatism, Law, and Language 19, 31 (Routledge 2014) (describing Professor Ronald Dworkin’s “chain novel” model of judicial decision-making).

    While many theories of interpretation (Dworkin’s, Professor William N. Eskridge Jr’s “Dynamic Interpretation,” etc.) may reject strictly fixed textual meanings, this Comment argues that Brandomian inferentialism can provide a robust general semantics, of which dynamic legal semantics is a type. Brandom’s inferentialism also supplies a set of more general semantic considerations that, this Comment argues, give us philosophical reasons to prefer this sort of interpretive model rather than provide arguments in the realm of governmental design, economic analysis, or legal process.

  • 7. This Comment follows Eskridge in calling an approach “originalist” when it “assume[s] that the legislature fixes the meaning of a statute on the date the statute is enacted.” William N. Eskridge Jr, Dynamic Statutory Interpretation, 135 U Pa L Rev 1479, 1480 (1987).
  • 8. Professor Brandom calls this model “communication as conveyance.” Robert B. Brandom, Making It Explicit: Reasoning, Representing, and Discursive Commitment 479 (Harvard 1994):

    Communicating is naturally conceived of as conveying something. According to such a conception, before an episode of communication takes place only the communicating agent possesses what is to be conveyed; after successful communication the recipient possesses it as well. . . . Communication is a way for speaker and audience to achieve a shared idea.


    . . . What the producer of a meaningful performance has initially and what in the case of successful communication its consumers eventually acquire is something—a content or meaning determining the significance of the remark—that is understood by both parties.

    In Part IV, this model of communication as conveyance is contrasted with Brandomian inferentialism, which does not require that inferential significance be preserved in

  • 9. See Richard S. Kay, Adherence to the Original Intentions in Constitutional Adjudication: Three Objections and Responses, 82 Nw U L Rev 226, 230 (1988). See also generally, Jack M. Balkin, The Construction of Original Public Meaning, 26 Const Commen 71 (2016).
  • 10. See Earl M. Maltz, Statutory Interpretation and Legislative Power: The Case for a Modified Intentionalist Approach, 63 Tulane L Rev 1, 6–13 (1988); Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U Chi L Rev 1, 32–39 (1985) (applying a purposivist approach to the Supreme Court’s constitutional interpretation). This Comment uses “purposivism” broadly to refer to a focus on explicit intentions or a more general underlying purpose—though they are distinguishable. That distinction, however, does not result in either approach not being “originalist,” as this Comment understands the term, and thus I generally elide the distinction between the original public meaning and purposivism approaches.
  • 11. 543 US 551 (2005).
  • 12. Id at 629 (Scalia dissenting).
  • 13. Id.
  • 14. Id at 608 (Scalia dissenting).
  • 15. 135 S Ct 2584 (2015).
  • 16. Id at 2612 (Roberts dissenting).
  • 17. Id (quotation marks omitted).
  • 18. Richard H. Fallon Jr, Three Symmetries between Textualist and Purposivist Theories of Statutory Interpretation—and the Irreducible Roles of Values and Judgment within Both, 99 Cornell L Rev 685, 704–07 (2014).
  • 19. 134 S Ct 2550 (2014).
  • 20. Id at 2564.
  • 21. Id.
  • 22. Id at 2565.
  • 23. Noel Canning, 134 S Ct at 2565.
  • 24. Lawrence B. Solum, The Fixation Thesis: The Role of Historical Fact in Original Meaning, 91 Notre Dame L Rev 1, 6–7 (2015).
  • 25. Id at 23.
  • 26. Id at 17.
  • 27. Id at 17–18.
  • 28. Frank H. Easterbrook, Foreword, in Antonin Scalia and Bryan A. Garner, Reading Law: The Interpretation of Legal Texts xxv (Thomson/West 2012).
  • 29. Solum uses the example of changing meanings of “domestic violence” as a case of linguistic drift. Solum, 91 Notre Dame L Rev at 16–17 (cited in note 24). Others have discussed the meaning of “cruel and unusual punishment” as an example, arguing that “unusual” originally meant “government practices that are contrary to ‘long usage’” or otherwise innovative but now means something like “different from that which is generally done” or abnormal. See, for example, John F. Stinneford, The Original Meaning of “Unusual”: The Eighth Amendment as a Bar to Cruel Innovation, 102 Nw U L Rev 1739, 1744–45 (2008).
  • 30. Eskridge, Dynamic Statutory Interpretation 135 U Pa L Rev at 1483 (cited in note 7).
  • 31. Id at 1496. This point evokes former Judge Richard Posner’s emphasis on the “lengthy interval” required before reinterpretive updating becomes an appealing possibility. See Hively, 853 F3d at 352–53 (Posner concurring).
  • 32. William N. Eskridge Jr, Dynamic Statutory Interpretation 48 (Harvard 1994).
  • 33. Id at 49.
  • 34. This could include, for example, Dworkin’s “law as integrity” approach, which allows judges to identify principles that fit with law’s institutional history and to select the principle(s) that provide the best moral justification for that institutional history. See Ronald Dworkin, Law’s Empire 190–92 (Harvard 1986). One could similarly characterize efficiency maximizing versions of law and economics, but there the judge’s task is to minimize cost rather than maximize moral appeal. See, for example, Richard A. Posner, The Law and Economics of Contract Interpretation, 83 Tex L Rev 1581, 1589–92 (discussing the efficiency implications of different strategies for interpreting contracts).
  • 35. Hively, 853 F3d at 352 (Posner concurring).
  • 36. Brandom, discussing the importance of semantics to legal reasoning, explains that semantically indeterminate legal concepts undercut the ability of law to serve as a source of rational justification. Brandom, A Hegelian Model of Legal Concept Determination at 20 (cited in note 6):

    It is essential to the normative bindingness of applications of legal concepts to particular cases that those applications can be rationally licensed by laws articulated by those concepts. Insofar as legal concepts are (whether for global, systematic reasons or local, contingent ones) semantically indeterminate in a way that precludes their functioning appropriately in justifications of legal decisions, one would be obliged to adopt a form of legal realism about those decisions that is indistinguishable from legal nihilism. For the idea that there is a difference between exercising normative authority by appeal to law and simply exercising power in its name depends on the possibility of distinguishing applications of the law that are rationally justifiable in virtue of the meanings of the concepts that articulate the law and those that are not.

    We might worry that some forms of dynamic interpretation provide just such a local, contingent source of semantic indeterminacy that leads merely to “exercising power in [law’s] name” in the way that Brandom identifies. Id.

  • 37. Hively, 853 F3d at 360, 373 (Sykes dissenting).
  • 38. Ulane v Eastern Airlines, Inc, 742 F2d 1081, 1085 (7th Cir 1984). The court uses “women” to refer to females and “men” to refer to males.
  • 39. Id.
  • 40. 490 US 228 (1989).
  • 41. Id at 235–37.
  • 42. Id at 250.
  • 43. Id.
  • 44. Doe v City of Belleville, 119 F3d 563, 593 (7th Cir 1997).
  • 45. Hively, 853 F3d at 340–41.
  • 46. Id at 341.
  • 47. Id.
  • 48. Id at 340, citing 42 USC § 2000e-2(a).
  • 49. Hively v Ivy Tech Community College of Indiana, 2015 WL 926015, *3 (ND Ind). See also Hively v Ivy Tech Community College of Indiana, 830 F3d 698, 699 (7th Cir 2016) (affirming the district court’s dismissal of Hively’s claim).
  • 50. Hively, 853 F3d at 343.
  • 51. 388 US 1 (1967).
  • 52. Hively, 853 F3d at 345.
  • 53. See id at 346–47.
  • 54. Id at 365–67 (Sykes dissenting).
  • 55. Id at 347.
  • 56. Hively, 853 F3d at 350 (“It would require considerable calisthenics to remove ‘sex’ from ‘sexual orientation.’”).
  • 57. Id at 346.
  • 58. Id at 347–49.
  • 59. Id at 348. Wood’s assertion that miscegenation laws “always were” racist previews the semantic externalism arguments discussed below. See note 113 and accompanying text.
  • 60. Hively, 853 F3d at 349.
  • 61. Id at 352 (Posner concurring).
  • 62. Id. The first two of these flavors, “extraction of original meaning” and “interpretation by unexpressed intent,” more or less correspond to the types of originalism developed above in Parts I.A and I.B, respectively.
  • 63. Id at 352 (Posner concurring).
  • 64. Hively, 853 F3d at 352.
  • 65. Id, citing William Blackstone, 1 Commentaries on the Laws of England 60
    (Chicago 1979).
  • 66. Hively, 853 F3d at 352 (Posner concurring).
  • 67. Id.
  • 68. Id at 353.
  • 69. Id.
  • 70. 135 S Ct 2584 (2015).
  • 71. Hively, 853 F3d at 356 (Posner concurring).
  • 72. Id at 360 (Sykes dissenting).
  • 73. Id at 362–63.
  • 74. Id at 362.
  • 75. See, for example, Matthew W. Green Jr, Same-Sex Sex and Immutable Traits: Why Obergefell v. Hodges Clears a Path to Protecting Gay and Lesbian Employees from Workplace Discrimination under Title VII, 20 J Gender, Race & Just 1, 36–39 (2017).
  • 76. See generally, for example, Brief for the United States as Amicus Curiae, Zarda v. Altitude Express, Inc, No 15-3775 (2d Cir filed July 26, 2017) (making similar arguments in favor of a narrow understanding of “sex discrimination”).
  • 77. See note 36 and accompanying text.
  • 78. “There, then, are two ways of thinking about various things.” Richard Rorty, Philosophy as a Kind of Writing: An Essay on Derrida, 39 New Lit Hist 101, 103 (2008).
  • 79. See Jeremy Wanderer, Robert Brandom 96 (McGill-Queen’s 2008).
  • 80. See generally, for example, Gottlob Frege, On Sense and Reference (Max Black, trans), in Peter Geach and Max Black, eds, Translations from the Philosophical Writings of Gottlob Frege 56 (Basil Blackwell 1960). But see Brandom, Making It Explicit at 94 (cited in note 8) (explaining that, although Gottlob Frege is “usually thought of as the father of the contemporary way of working out the representationalist order of explanation,” one could “read . . . back into Frege” inferentialist themes).
  • 81. Robert B. Brandom, Reason in Philosophy: Animating Ideas 40–41 (Harvard 2009).
  • 82. Id at 10.
  • 83. See id at 95–99 for further discussion of the representationalist and inferentialist distinction. See also Part III.B, which describes in detail one such language game.
  • 84. Id at 29–33.
  • 85. Brandom, Reason in Philosophy at 33 (cited in note 81).
  • 86. Id.
  • 87. Brandom, Making It Explicit at 88 (cited in note 8).
  • 88. Id at 88–89.
  • 89. Brandom, Reason in Philosophy at 135 (cited in note 81).
  • 90. See id.
  • 91. Id at 8.
  • 92. Id at 10.
  • 93. Brandom, Reason in Philosophy at 35 (cited in note 81) (emphasis omitted). See also Brandom, Making It Explicit at 172–73 (cited in note 8).
  • 94. Brandom, Reason in Philosophy at 36 (cited in note 81).
  • 95. See id at 11–14.
  • 96. Id at 37. Apperception is the process of integrating a concept or commitment into the broader constellation of commitments one holds. Brandom emphasizes the “synthetic unity” of apperception because there is an active synthesis of concepts by the rational agent that produces a unified web of concepts subject to norms of criticism, amplification, and justification.
  • 97. Id at 41.
  • 98. This is the heart of Brandom’s inferential account in Making It Explicit. For a summary, see Wanderer, Robert Brandom at 41–53 (cited in note 79).
  • 99. For more on challenging, see generally Jeremy Wanderer, Brandom’s Challenges, in Bernhard Weiss and Jeremy Wanderer, eds, Reading Brandom: On Making It Explicit, 96 (Routledge 2010).
  • 100. Brandom, Making It Explicit at 5 (cited in note 8).
  • 101. The game is “perspectival” in that “it essentially involves a distinction of social perspective, between what one is doing in acknowledging a commitment (oneself) and attributing a commitment (to someone else).” Robert Brandom, Reply to Allan Gibbard’s “Thought, Norms, and Discursive Practice”, in Weiss and Wanderer, eds, Reading
    Brandom 297, 298 (cited in note 99). The game is thus “social” in what Brandom calls an “I-thou” sense because of this distinction between the scorekeeping/acknowledgement and gameplaying/attribution perspective. “I-thou” sociality can be contrasted with “I-we” conceptions of the social, which involve a distinction between an individual and a broader community. Id at 297–300.
  • 102. See Brandom, Making It Explicit at 185–91 (cited in note 8).
  • 103. Id at 185.
  • 104. To see this discrepancy, remember that a gameplayer, who keeps score on herself, will assign to herself some set of entitlements and incompatibilities given her commitments. Her self-assigned set of inferential consequences may not match those assigned based on other scorekeepers’ standards of correctness. This possibility—that the commitments of a gameplayer entail commitments of which the gameplayer is not herself aware—will become important for thinking about scorekeeping activity in judging. See note 119–21 and accompanying text.
  • 105. Brandom, A Hegelian Model of Legal Concept Determination at 19 (cited in note 6).
  • 106. See Wanderer, Robert Brandom at 202–06 (cited in note 79) (explaining that multiple possible understandings of appropriateness could characterize the instances of scorekeeping).
  • 107. See id at 146–47.
  • 108. Recall “communication as conveyance.” See note 8.
  • 109. See Wanderer, Robert Brandom at 152–53 (cited in note 79).
  • 110. See id.
  • 111. See note 104.
  • 112. Robert B. Brandom, Articulating Reasons: An Introduction to Inferentialism 29 (Harvard 2000).
  • 113. Brandom, A Hegelian Model of Legal Concept Determination at 36 (cited in note 6) (emphasis added).
  • 114. Brandom, Reason in Philosophy at 84 (cited in note 81).
  • 115. Id at 84–85.
  • 116. See Brandom, A Hegelian Model of Legal Concept Determination at 33–38 (cited in note 6).
  • 117. Brandom, Reason in Philosophy at 86 (cited in note 81).
  • 118. Id at 87.
  • 119. See id at 92–93.
  • 120. Brandom, A Hegelian Model of Legal Concept Determination at 36–37 (cited in note 6).
  • 121. See the discussion below in Part IV.B.
  • 122. Brandom, A Hegelian Model of Legal Concept Determination at 36 (cited in note 6).
  • 123. Hively, 853 F3d at 357 (Posner concurring).
  • 124. See Part III.B.
  • 125. See, for example, Robert H. Bork, The Judge’s Role in Law and Culture, 1 Ave Maria L Rev 19, 22–24 (2003).
  • 126. See, for example, Romans 5:12–21 (King James Version).
  • 127. See Part II.B.3.
  • 128. Hively, 853 F3d at 361 (Sykes dissenting).
  • 129. See Part I.A.
  • 130. See notes 108–12 and accompanying text.
  • 131. Molybdenum (Royal Society of Chemistry, 2017), archived at http://perma.cc/T3A2-3NY2.
  • 132. See Wanderer, Robert Brandom at 21 (cited in note 79).
  • 133. See id.
  • 134. See id.
  • 135. See id.
  • 136. As discussed above, the public concepts in question are not “the same” in the sense of consisting in a discrete packet of content that is shared by all of the participants in a conversation. The inferential significance of the concept is necessarily perspectival, but the norms for evaluating correct or incorrect scorekeeping performances associated with the public concepts being considered here are, like the norms associated with “molybdenum,” meaningfully external to a particular speaker—part of an already-up-and-
    running linguistic practice rather than a narrowly localized one. See notes 107–13 and accompanying text.
  • 137. Hively, 853 F3d at 360, 362–63 (Sykes dissenting).
  • 138. Id at 352–55 (Posner concurring).
  • 139. See Noel Canning, 134 S Ct at 2564.
  • 140. See generally Frank H. Easterbrook, Pragmatism’s Role in Interpretation, 31 Harv J L & Pub Pol 901 (2008).
  • 141. Hively, 853 F3d at 360 (Sykes dissenting).
  • 142. Id.
  • 143. See Part IV.C. Think, for example, of Justice Elena Kagan’s textualist dissent in Yates v United States, in which she argues that a fish was a “tangible object” even in the context of a post-Enron corporate responsibility statute. 135 S Ct 1074, 1091 (2015) (Kagan dissenting). Kagan styles herself as vindicating the true meaning of “tangible object” and as faithful to the statutory text. See id. A Brandomian can be thought of as faithful to the text in much the same way but with potentially different views about how to go about determining the true meaning of “tangible object.”
  • 144. Brandom, Making It Explicit at 83 (cited in note 8).
  • 145. Hively, 853 F3d at 360 (Sykes dissenting).
  • 146. Id at 373.
  • 147. 871 F3d 1260 (11th Cir 2017).
  • 148. Id at 1263.
  • 149. Id.
  • 150. Id at 1264, citing 26 USC § 213(d).
  • 151. Morrissey, 871 F3d at 1264–68.
  • 152. Id at 1265.
  • 153. Id at 1267.
  • 154. Id at 1272.
  • 155. I refer here principally to the work of philosophers like Professor Ruth Millikan, who attempts to develop a naturalized account of biological function tied to evolutionarily selected-for traits. See generally Ruth Garrett Millikan, Language, Thought, and Other Biological Categories: New Foundations for Realism (MIT 1984). This turns out to be an extremely complicated question, and full engagement with it is beyond the scope of this Comment.
  • 156. See generally, for example, Lockhart v United States, 136 S Ct 958 (2016) (featuring application of different originalist statutory techniques in both the majority and dissenting opinions).
  • 157. See Obergefell, 135 S Ct at 2596.
  • 158. See, for example, Motor Vehicle Manufacturers Association of the United States, Inc v State Farm Mutual Automobile Insurance Co, 463 US 29, 53–56 (1983) (evaluating costs and benefits of seat belt regulations).
  • 159. See, for example, Daubert v Merrell Dow Pharmaceuticals, Inc, 509 US 579, 592–93 (1993) (“Faced with a proffer of expert scientific testimony . . . the trial judge . . . must make a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and whether that reasoning properly can be applied to the facts in issue.”).
  • 160. See, for example, Brown v Board of Education of Topeka, 347 US 483, 494 n 11 (1954) (citing to detrimental psychological effects of segregation).
  • 161. See, for example, Carpenter v United States, 138 S Ct 2206, 2214–16 (2018).
  • 162. See Allison Orr Larsen, The Trouble with Amicus Facts, 100 Va L Rev 1757, 1784–1808 (2014).
  • 163. See, for example, the historical analysis in Michael H. v Gerald D., 491 US 110, 124–27 (1989). If judges are capable of tracing conceptions of the unitary family through centuries of common-law doctrine, then they are probably capable of taking stock of how those conceptions stand in the present day.