Probabilism in Legal Interpretation


Rules that appear in constitutions and statutes are cast in words. Do those words have an objectively ascertainable meaning?

For many decades, this fundamental question has been the subject of an ongoing debate implicating disciplines as diverse as philosophy, political theory, and sociology.1 The debate’s participants split into five distinct schools of legal interpretation, identified as textualism, intentionalism, purposivism, pragmatism, and fiatism. Textualism follows the “meaning at birth” approach: it interprets a statutory or constitutional provision in accordance with how the provision was generally understood when it was put in place.2 For example, in applying a statutory rule that prohibits the use of “vehicles in parks,” textualist judges ought to find out what means of transportation and types of land were ordinarily included in the words “vehicle” and “park” at the time of the rule’s enactment. Intentionalism, by contrast, focuses on the subjective intent of the provision’s drafters.3 Correspondingly, in ascertaining what “no vehicles in parks” means, an intentionalist judge ought to find out what the rule’s drafters intended to communicate by the words “vehicle” and “park.” If they find credible evidence indicating that the drafters intended to limit the prohibition’s scope to automobiles and public recreational lands, she should interpret the words “vehicle” and “park” in accordance with this intent, notwithstanding the misalignment between that narrow interpretation and ordinary language.

Under both textualism and intentionalism, words and sentences of a statutory or constitutional provision must be traced back and tied to the authority—the provision’s drafters—that brought those words and sentences into existence. For that reason, these schools of thought connect to the broader notion of originalism.4 Textualism proceeds on an irrebuttable presumption that the provision’s drafters communicated their command to the subjects by using ordinary language.5 Under intentionalism, this presumption is rebuttable: Credible evidence indicating that the drafters intended to communicate something different will override the ordinary language.6

Purposivism, in contrast, derives the meaning of legal words and sentences from the unstated—yet, underlying—purpose of the provision that contains those words and sentences.7 This approach employs the so-called “mischief rule” that purports to reconstruct the lawmaker’s goal or intent by prompting judges to identify—analytically or empirically—the mischief that the provision aims to quell or abate.8 For example, when the history, the rationale, or other credible background information indicates that the provision “no vehicles in parks” was enacted to prevent damage to the environment, purposivist judges will tend to interpret “vehicle” as limited to transportation using fossil fuels and possibly to expand the meaning of “park” to include privately owned recreational lands—regardless of whether the drafters of the rule had all that in mind.

Textualism, intentionalism, and purposivism are being challenged by pragmatism and fiatism. From a pragmatist perspective, words such as “vehicle,” “park,” and all other legal terms have no fixed or objectively ascertainable meaning and are open to multiple interpretations.9 As a result, a judge ought to assume the role of a benevolent legislator or trustee and inject meaning into the words of constitutions and statutes in a way that best serves societal interests at a given point in time.10 For example, when a pragmatist judge finds out that the parks’ massive invasion by bicycle riders creates a socially inefficient congestion while denying other visitors enjoyment of the park, she can interpret “vehicle” to include bicycles.11

Fiatism is a concept defined by the philosopher Borden Parker Bowne as “the fancy that we can make things over to suit ourselves by renaming them.”12 The phenomenon it describes has been—and still is—the raison d’être of the critical schools of thought that reject textualism, intentionalism, purposivism, and pragmatism altogether.13 These schools include critical legal studies, critical race theory, as well as some of the Marxist and feminist approaches to law.14 The common denominator of these schools of thought is the ever-present nexus between language and power.15 Theorists who affiliate themselves with these schools maintain that legal concepts—such as “offer,” “acceptance,” and “intent”—and many other legal words—such as “vehicle” and “park”—are inherently ambiguous because they are abstract and have no direct referents in the real world. This feature opens the words to multiple interpretations, thereby giving judges a broad power to determine the actual meaning of the law.16 However, unlike pragmatists, critical legal scholars maintain that judges do not act as trustees for society as a whole. On the contrary, according to those scholars, judges act for the benefit of the rulers who put them in that position and for the benefit of the well-to-do classes that the rulers—and judges themselves represent.17

Within the framework of fiatism, critical legal scholars also maintain that legal terms are arranged into dichotomies: good faith vs. bad faith; negligence vs. adequate care; adverse vs. non-adverse possession; and so forth.18 According to their argument, such dichotomies are false because any of them can be deconstructed in a way that will show that the realities they purport to categorize on a yes-or-no basis escape that categorization.19 In fact, those realities form a continuum situated between the two dichotomized extremes. As a result, judges are free to categorize them either way.20 This feature, too, makes words of the law radically indeterminate; and when indeterminacy carries the day, fiatism triumphs over reason and, as a result, might makes a right. From that perspective, what judges say about the meaning of “no vehicles in parks” and other words of the law turns out to be a political determination catered to the ruling classes while being dressed up—and thereby legitimized—as a legal ruling.21

This Article offers an alternative theory of interpretation identified as probabilism. At its core, probabilism maintains that judges should view legal rules as a communication between individuals. The parties to that communication are the lawmakers, who give their commands in the form of rules, and the individuals who receive those commands. What these commands say and do not say is a purely empirical fact. Correspondingly, in order to identify the content of such communications, judges should proceed in the same way they carry out factfinding in a bench trial. Under probabilism, the question “What was communicated by the individuals who drafted the statute prohibiting the use of vehicles in the park?” is a question of a potentially uncertain fact. This question is conceptually no different from the empirical question “What was communicated by Jane Roe to John Doe on January 1, 2020, when she wrote him a note saying that he should not use any of his vehicles in the park?”22

To answer any such question under probabilism, judges ought to proceed in three steps. First, judges ought to consider all available information that pertains to the interpretive task they are facing. Second, judges ought to specify and analyze all the probable meanings of the rule as a communication coming from its maker—the legislator or the constitution’s founders. Third and finally, among all probable meanings of the rule, judges should adopt the one they deem most probable to be factually correct. To that end, judges should determine the probability that the provision’s drafters have chosen its wording to communicate the meaning under consideration. As I explain in the pages ahead, this approach uses the conventional evidentiary principles to resolve questions of statutory and constitutional interpretation. Specifically, probabilism combines the original public meaning of the underlying statutory or constitutional provision with the evidence concerning the drafters’ intent and the applicable default rules to find out what the drafters likely meant to say. By following probabilism, judges will enhance the accuracy of their decisions and accomplish their mission as faithful agents of the legislature and the people.

En route to explaining the theory of probabilism, I reevaluate and criticize the competing theories of interpretation, namely, textualism, intentionalism, purposivism, pragmatism, and fiatism. These five approaches, taken together, present a false choice between two philosophical extremes: the worldview promulgated as “objectivism” and adopted by the originalist schools of thought—textualism and intentionalism—and the opposite worldview that goes under the name of “relativism” and is followed by the critical schools of thought and, to a lesser degree, by purposivism and pragmatism as well.23 Objectivism holds that words by which people communicate with each other have ascertainable true meanings.24 Relativism, on the other hand, maintains that most words have no such meanings: They are empty shells that need to be filled in and are consequently up for grabs.25 The choice between the objectivist and the relativist worldviews is false because it ignores yet another plausible worldview: Words in general, and legal terms in particular, may have factual meanings that are probably, rather than ascertainably, true; and if so, treating those words as empty shells is wrong.

After introducing probabilism and illustrating how it works, I proceed to conduct a comparative investigation into its merits and shortcomings relative to the alternative schools of thought: textualism, intentionalism, purposivism, pragmatism, and fiatism. This investigation reveals that probabilism outperforms both the text-centered and the intent-centered variants of originalism as a working theory of interpretation. Originalism suffers from a serious methodological flaw, thus far unacknowledged. Every theory of meaning must develop secondary rules on how to resolve uncertainties and genuine disagreements over what a given expression says or purports to say.26 As a corollary, a workable theory of legal interpretation should provide both explicit and comprehensive guidance on how to interpret statutory and constitutional provisions in the face of uncertainty and disagreement. Originalism provides no such guidance. It incorporates no standards of proof or other secondary rules and consequently gives judges no instructions on how to interpret a statutory or constitutional provision that does not speak for itself and does not reveal its drafters’ intent. This omission is particularly problematic in constitutional law where the dated language and original intent of the Framers are both unclear and temporally distant from current affairs.

To fix this flaw in originalism, instead of confining the scope of judicial inquiries into what the drafters of the unclear provision actually said or had in mind, originalists must switch to an inclusionary approach that allows judges to hear all evidence that shows indicia of trustworthiness and can help ascertain the provision’s most probable meaning.27 The transition from the pure originalism to the “public meaning” originalism28 has moved the theory in that direction, but not sufficiently so: the move took place without setting up evidentiary rules that include standards of proof, or probability thresholds, for making decisions under uncertainty as to what the “public meaning” of a statutory or constitutional text likely is. This methodological shortcoming can be fixed by introducing the requisite rules of evidence and probability thresholds. However, this fix will not merely repair originalism: It will also transform both versions of originalism into probabilism.

Probabilism also outperforms fiatism, albeit, for a different reason. Fiatism embraces no theory of meaning to begin with. Rather, it offers a negative account that rejects the interpretive enterprise altogether.29 Under fiatism, legal rules have no author-related meanings discoverable by impartial, fact-centered inquiries but rather are empty shells that can be filled in by virtually any content chosen by the people in power.30 Vindicating this extreme position against textualism, intentionalism, purposivism, and pragmatism is a nontrivial task. Yet, this task is also not insurmountable, since none of those four theories imposes effective external constraints upon judges’ decisions in cases involving an unclear statutory or constitutional provision. Defending the “empty shell” account of legal rules against probabilism is more difficult. Doing so requires one to establish that probabilities of factual correctness ascribable to the meanings of legal rules are, in fact, unreal. For example, to prove that the prohibition on “vehicles in parks” is an empty shell, one must establish the futility of the probabilistic claim that the meaning of the word “vehicle” is more likely to include cars than wheelchairs. Such endeavors can hardly succeed.

To make a comprehensive assessment of how purposivism, as a theory that derives a word’s legal meaning from the underlying purpose of the provision, and pragmatism—a theory that allows judges to infuse meanings into legal words in a way that best serves society’s interest—stand against probabilism, one needs to notice their forced fusion.31 Because judges interpreting a legal rule can virtually never assume that the rule’s drafters intended its designated goal to be attained at all costs, as opposed to the costs economically adjusted to the social value of the goal, the judges’ purposivist interpretation of the rule must always incorporate a cost-benefit tradeoff. As a result, purposivism transforms into pragmatism. By the same token, a pragmatist judge who cares about nothing but social welfare ought to identify and rank the social valuables she ought to maximize. Eliciting those valuables from the text and the underlying purpose of statutory and constitutional rules, instead of contriving them by using one’s personal judgment, is the best way to accomplish that task. By taking this path in interpreting legal rules, however, the pragmatist judge will be implementing the core idea of purposivism.

The inevitability of cost-benefit tradeoffs is a problem for both purposivism and pragmatism. Deriving the meaning of a legal rule from a cost-benefit tradeoff, instead of ascertaining what the rule actually or probably means, ignores and often sacrifices the entitlement that the rule confers on the individual or on the group of people.32 Any such sacrifice amounts to an extra burden that the individual or the group is forced to bear, in addition to paying their regular taxes, for the benefit of society as a whole. By forcing the individual or the group to bear such additional burdens, the state denies them equal concern and respect.33 To avoid this ill effect, the legal system must adopt probabilism—the school of thought geared toward maximizing the accuracy of courts’ decisions that determine the factual meanings of statutory and constitutional provisions and the corresponding rights, duties, and obligations. Probabilism, of course, is not the only school of thought that takes positive law seriously: The two versions of originalism do so as well. But as discussed above, originalism has shortcomings that make it methodologically inferior to probabilism.

Structurally, this Article unfolds as follows. In Part II, I introduce probabilism and explain its mechanics. In Part III, I analyze the strengths and weaknesses of probabilism relative to textualism and intentionalism, and demonstrate that probabilism outperforms these originalist methodologies. As part of that demonstration, I revisit three canonical Supreme Court decisions that allude to originalism: Kasten v. Saint-Gobain Performance Plastics Corp.,34 Staples v. United States,35 and Crawford v. Washington.36 Contrary to the conventional wisdom, I show that each of those decisions implicitly relies upon and is best explained by probabilism. In Part IV, I analyze the recent U.S. Supreme Court decision, Bostock v. Clayton County,37 that spurred controversy among the originalist justices as to how to determine the meaning of the word “sex” in Title VII of the Civil Rights Act of 1964—that outlawed workplace discrimination on the basis of “race, color, religion, sex, or national origin”38—and whether it encompasses sexual orientation as well. The following analysis questions the Justices’ fundamental premise associating the originalist text-oriented and intent-based “fixation thesis”39 with the certainty of meaning. Specifically, I show that this association is responsible for the logical inconsistencies that plague both the majority opinion and the dissent, and then move on to demonstrate that statutory and constitutional meanings need not be certain and can be determined probabilistically. Furthermore, I demonstrate that under the probabilistic approach that utilizes all relevant evidence—including the information generated by applying Ludwig Wittgenstein’s concept of “family resemblance”40—the word “sex” unquestionably includes the lesbian, gay, bisexual, and transgender ways of life, thereby extending the Title VII protection to the LGBT community. In Part V, I demonstrate that probabilism outperforms purposivism, pragmatism, and fiatism on both descriptive and normative grounds. A short Conclusion follows.

Introducing Probabilism

Evidence theory has developed three fundamental insights that can improve the ways in which judges determine the factual meanings of ambiguous statutory and constitutional provisions. The first and most basic insight is the need to set up standards of proof: rules that determine the probability thresholds for factual findings under uncertainty, such as “preponderance of the evidence,” “clear and convincing evidence,” and proof “beyond a reasonable doubt.”41 The second insight is the social utility in setting up default rules, or tiebreakers, in the form of rebuttable presumptions that will instruct courts on how to resolve disagreements in certain predetermined types of cases.42 The third insight is about evidence-selection: Rules ensure that courts use evidence that rationally facilitates factfinding and do not consider evidence that is wasteful, irrelevant, or misleading.43

These three insights define probabilism—a method that courts have been using in carrying out factfinding tasks since time immemorial.44 This method highlights the necessity of having secondary rules that will instruct courts on how to determine individuals’ primary rights, duties, and obligations under conditions of uncertainty. As stated in the Introduction, I posit that judges should use the same factfinding method—and an analogous set of secondary rules—in determining the meanings of unclear legal provisions: human communications that appear in statutes and constitutions.45

The use of probabilism as a theory of legal interpretation is not as simple as it appears to be because it depends on the meaning of the legal “meaning.”46 As Professor Richard Fallon put it, “[w]e need to know what we are looking for before we can ascertain whether the evidence sufficiently establishes what needs to be proved.”47 Under probabilism, the meaning of a statutory or constitutional provision that needs to be ascertained is the communication intended by the provision’s drafters as encapsulated in the words and the sentences they chose to use. As a result, the interpreters need to discover, as an empirical fact, what the drafters’ intent was and how it was manifested in the provision’s words and sentences.

Empirical facts are virtually never certain and can only be probable.48 Communications encapsulated in the words and sentences of the lawmaker are no exception to this general observation.49 Judges effectively never have enough information to determine the meaning of those communications with certainty. Consequently, they have no choice but to base their interpretive decisions upon probability. Under such circumstances, judges facing an unclear statutory or constitutional provision would have to identify its possible meanings by considering every relevant evidence and then determine which of those meanings is most probable. To make that determination, judges would have to ask themselves and answer the following question: Which of the provision’s available meanings represents what its drafters most likely intended to communicate through the language they chose to use?

To answer this question, judges must proceed in two steps. First, they need to identify the provision’s plausible meanings. To this end, the judges need to determine what the drafters’ language could communicate to people at the time it was used. Second, after identifying the provision’s plausible meanings, judges need to determine, in relation to each identified meaning, how probable it is that the drafters chose to use the language they used to communicate this particular meaning, as opposed to an alternative meaning. In carrying out this task, judges ought to consider every relevant and potentially credible evidence, including the provision’s history and purpose, and how the provision’s words and sentences were understood at the time of its enactment. This interpretive methodology has two advantages. The first of these advantages is factualism50: This methodology enhances the accuracy of courts’ decisions that ascribe meanings to unclear statutory and constitutional provisions. The second and equally important advantage is separation of powers: Probabilism forces judges to act as faithful agents of the legislature and the people.

As a threshold matter, the interpreters of statutory and constitutional provisions ought to acknowledge that they would not be able to carry out their task in the absence of standards of proof, presumptions, and other secondary rules. When secondary rules do not exist and the meaning of the underlying legal provision is unclear, there is no way to ascertain what that provision likely means. The absence of secondary rules is the principal shortcoming of the two versions of originalism: textualism and intentionalism. Each of these schools of thought endorses the “fixation thesis” postulating that words of the statute and the constitution acquire their fixed meanings at the time of the enactment.51 As a result, textualism only works well when the underlying legal provision has a readily ascertainable ordinary meaning, while leaving the judges in a state of impasse in all other cases.52 By the same token, intentionalism works well when the drafters’ intent as to what the underlying provision means is crystal clear, but it offers no tiebreaking rules for cases in which this intent is uncertain. As a result, judges who practice originalism find themselves at an impasse when they encounter a statutory or constitutional provision that can be plausibly interpreted in more than one way.

To avoid such an impasse, judges should stop using textualism and intentionalism in their pure forms. Instead, they should try to establish which of the available meanings of the underlying legal provision represents what its drafters most likely intended to communicate by using the language they chose to use. Hence, judges should follow probabilism.

Standard of Proof

Acknowledging that secondary rules are indispensable for legal interpretation does not yet establish what those rules should ordain. However, the initial task faced by the legal system is unmistakably clear: The system ought to set up standards of proof in the form of probability thresholds that will allow judges to make factual findings necessary for ascertaining the meanings of unclear laws. The process of formulating those standards turns out to be much easier than designing the standards of proof for the general law of evidence. Under the general law of evidence, standards of proof come in the form of probability thresholds for decisions that determine the risks of error that factfinders can and cannot assume: “preponderance of the evidence,” “clear and convincing evidence,” and “proof beyond a reasonable doubt.”53 The nature of the risk of error varies from one decisional context to another. In criminal cases, the risk of convicting an innocent defendant differs from the risk of acquitting a guilty offender, as well as from the risk of denying an excusatory defense to a defendant proven beyond a reasonable doubt to have committed the underlying offense.54 In civil cases, the risk of erroneously depriving a person of her money or property differs from the risk of a mistaken deportation, denial of parental rights, confinement to a mental institution, and so forth.55 This variability complicates the lawmakers’ task.

In the domain of statutory and constitutional interpretation, however, things are different because the risks of error on both ends are stable and symmetrical. To see why, consider a hypothetical scenario in which facts supporting the ascription of meaning A to the underlying statutory or constitutional provision are as likely to be present as the facts indicating that the provision actually means to say B rather than A. To stay within the bounds of this hypothesis, assume further that in arriving at the conclusion that A is as probable as B, judges did everything they were supposed to do: They considered every piece of available evidence pertaining to the question “A or B?” while ignoring misleading and irrelevant evidence. Under such circumstances, probabilism tells us that the judges can decide the case either way. Decision A would be as good as decision B: both decisions are equally justified on factual grounds. Importantly, the fact that there is, or might be, a reason to believe that adopting meaning A over B, or vice versa, will make society better off is immaterial. This reason has nothing to do with the factual grounds of the underlying legal provision that have been fully accounted for. Similar to all other factfinding endeavors, the inquiry into the provision’s meaning is empirical rather than normative: It is about the “is,” not the “ought.”

This abstract scenario lays the groundwork for setting up the general probability threshold, or standard of proof, for the domain of legal interpretation. The requisite standard should be set at the preponderance of the evidence. That is, when the facts indicating that the provision in question means A are more probable than not, judges should interpret the provision as saying A rather than anything else. Under probabilism, A wins the day because it is comparatively the most probable meaning of the provision. That is, A wins the day because the provision’s drafters most likely intended to communicate A by using the language they chose to use. Choosing A over B thus enhances the accuracy of the judges’ decisions and aligns with the judges’ duty to act as faithful agents for the legislature and society as a whole. Notably, this standard of proof was implicitly present in the old rule that instructed courts to interpret an ambiguous statutory or constitutional provision by adopting the provision’s better-attested historical meaning.56 This rule has been uncovered and discussed in a recent article by Professors John McGinnis and Michael Rappaport, who claim, on originalist grounds, that “at the time of the Constitution’s enactment a legal interpretive rule . . . required interpreters to choose the better supported interpretation,” and that “the 51–49 rule—the rule that directs the interpreter to choose the better interpretation of a provision, even if it is only slightly better—is a key to reducing uncertainty.”57

This formulation of the requisite standard of proof raises the question about the nature of the probability that judges should use in determining the meanings of unclear statutory and constitutional provisions. Probability is often understood as referring to statistical information cast in mathematical language.58 For the most part, this understanding has been rejected by the courts of law.59 In courts, probability predominantly refers to the relative plausibility of the competing concrete accounts of factual events, formulated as stories.60 This feature derives from the irreducibly second personal nature of legal entitlements, obligations, and responsibilities that attach to the rightholder, on one side, and to the duty-bearer, on the opposite side. Correspondingly, facts that courts need to ascertain ought to be second personal as well. Courts must only rely upon second personal evidence, that is: upon case-specific information concerning the alleged jural relationship between the holder of the underlying entitlement and the bearer of the correlative duty or obligation.61 Statistical distributions will not do.

Courts should proceed in the same way in determining the probabilities of factual accounts that ascribe meanings to the lawmaker’s communications. Whether one such account is more probable than its competitors depends on the accounts’ relative plausibility. “There is no mathematical algorithm for ‘plausibility’: What informs courts’ evaluations of plausibility are all the experience-based factors that convince people that some story may be true” and another story may be false.62 These factors include coherence, consistency, completeness, articulation, simplicity, and evidential support or consilience.63 Based on these factors, courts consider the competing accounts and decide which is superior and wins the day. To win the plausibility contest under the preponderance standard, evidence must unfold a narrative that makes the most sense.64 Among the prima facie plausible accounts as to what the underlying statutory or constitutional provision means to communicate, the court should select the account that gets the overall highest score on coherence, consistency, completeness, articulation, simplicity, and evidential support.


For cases involving two or more equally plausible understandings of the underlying provision, the legal system needs to set up default or tiebreaking rules.65 One such tiebreaker, which many legal systems actually use, comes in the form of the rule that instructs interpreters to adopt the provision’s current ordinary meaning when everything else is equal.66 This tiebreaker is related to the drafters’ forward-looking standpoint.67 When the provision in question is open to two or more equally plausible understandings, selecting the understanding that reflects the current ordinary meaning of the provision has a better chance to hit the mark because the drafters may have intended the provision to track the changes in the phenomena and the social practices to which the provision’s language refers.68 For example, an old statute referring to “documents” must be understood, when everything else is equal, as extending to computerized or digital records that did not exist at the time. That is, when nothing in the statute indicates its particular preference for paper records, it is safe enough to assume that the drafters used the word “documents” to reflect the future changes in society’s documentation practices. By the same token, the word “vehicles” in the “no vehicles in the park” prohibition would include Teslas and other cars powered by electricity when everything else is equal, that is: when evidence shows that the prohibition was put in place to advance different goals that include the safety of walkers and joggers in the park. The fact that the prohibition’s drafters did not anticipate the advent of vehicles not using fossil fuels would be of no consequence because the lawmakers clearly had in mind all kinds of transportation, including those not yet developed. By contrast, things would be different with respect to a law imposing a “clean air tax” upon owners of “all vehicles” in order to reduce air pollution caused by the use of fossil fuels. This provision cannot be interpreted to cover nonpolluting cars powered by electricity.

Another tiebreaker, or default rule, established by numerous legal systems, is the so-called “lenity rule” that applies in criminal cases, where it requires judges to prefer a narrow over a broad interpretation of the underlying offense upon finding both interpretations equally plausible.69 Importantly, this rule does not apply when the tiebreaker is not required.70 Thus, when the broad interpretation of the criminal prohibition is more likely to be correct than the narrow one, the court should simply adopt the broad interpretation. That is, the court must proceed on the assumption that the risks of error on both ends of the interpretive disagreement are symmetrical. Tiebreakers are being used merely to avoid impasse71 and, as a supplementary reason, to motivate drafters of statutes to be as precise as possible.72 Remarkably, the tiebreaking rules set up for the domain of legal interpretation operate similarly to presumptions under the law of evidence.73 They identify the preferred interpretation for ambiguous statutory and constitutional provisions and shift the burden of proof to the party advancing a different interpretation.

Furthermore, the tiebreakers menu may also include a rule that instructs judges facing equally probable interpretations of the underlying statutory or constitutional provision to choose the interpretation that best promotes overall social welfare.74 This tiebreaker may create some common ground between probabilism, textualism, intentionalism, purposivism, and pragmatism. It is important to understand, however, that this common ground will not transform probabilism into a different school of thought.

Under probabilism, welfare maximization can only function as a tiebreaker rather than as the ultimate goal of the process by which judges ascertain the meanings of unclear laws. Put differently by using my terminology: Welfare maximization can only be a secondary default rule of legal interpretation. This critical factor separates probabilism from purposivism and pragmatism that oftentimes put welfare maximization ahead of all other considerations. For slightly different reasons, textualism, too, cannot untie judges’ decisions from the words of the law by substituting those words with an open-ended standard such as welfare maximization. This substitution will unravel textualism. Therefore, textualism can only use welfare maximization (or another chosen value, such as equality) as a tiebreaker, similarly to probabilism.

As far as intentionalism is concerned, this approach cannot simply adopt welfare maximization or another tiebreaker for cases in which the drafters’ intent is unclear. Thus, applying a tiebreaker will not fix the fundamental problem that intentionalism suffers from because tiebreakers cannot determine what was an “unclear intent.” To separate the “clear intent” cases from the “unclear intent” cases governed by the tiebreaker, intentionalism would need to set up a standard of proof for the uncertain evidence to determine what was the drafters’ intent. Critically, the adoption of any such standard would undo intentionalism. Setting the requisite standard too high would make the drafters’ intent virtually unprovable, which would let the welfare-maximizing tiebreaker seize the domain of interpretation, thereby transforming intentionalism into some version of pragmatism or purposivism. On the other hand, setting the standard at preponderance of the evidence, in tune with my proposal, would transform intentionalism into probabilism.75

Language, “Family Resemblances,” and Other Evidence

I now move on to discuss the mechanisms for selecting evidence under probabilism. These mechanisms track the admissibility, the judicial notice, and the permissive inference rules operating under the general law of evidence.76 Those rules prioritize evidence that provides empirical grounds for making factual findings, while excluding from the factfinders’ consideration nonfalsifiable evidence that falls into the “self-asserting, self-serving, and speculative” categories.77

Begin with the inclusionary rule that instructs courts to consider and form their decisions upon all potentially credible information.78 This rule provides that, in ascertaining the meaning of an unclear statutory or constitutional provision, judges must take into consideration all available evidence that can help understand the provision’s language and context. As far as language is concerned, such evidence will relate to the drafters’ intent in communicating the provision the way they did; the way in which the provision’s words and expressions have come to be understood in other legal contexts, as well as in people’s ordinary speech, at a given point in time; and so forth. Critically, in ascribing meanings to general and open-ended concepts that appear in statutes and constitutions, judges will also have to account for “family resemblances”—generalized descriptions as to what the given concept normally encompasses or encompassed at a given point in time.79 Such general descriptions include facts “generally known within the trial court’s territorial jurisdiction” that can be judicially noticed.80 They can also be proven by evidence. Either way, “family resemblance” will function as a dependable secondary rule in cases in which the word’s or the concept’s true meaning is uncertain.

Accounting for “family resemblances” involves an empirical, as well as logical, investigation into people’s linguistic and conversational practices. This important point was made decades ago by Ludwig Wittgenstein,81 whose philosophical work focused on the rule-governed aspects of people’s language—specifically, on the connection between words and their particular uses or instantiations.82 As part of this work, Wittgenstein coined the term “family resemblance” as a substitute for the words’ essential meanings.83 This term captures the ways in which people speaking the same language use and understand words as referring to “a complicated network of similarities overlapping and criss-crossing.”84

The word “game,” Wittgenstein’s prime example of a family resemblance, vividly illustrates this concept as an intricate network of meanings.85 The word “game” captures activities as diverse as soccer and poker, whose only common denominator is competition that produces winners and losers. At the same time, when a child throws a ball against the wall in order to catch it, she also engages in a game: albeit, a noncompetitive game. The games I mentioned may still have a common denominator: for example, recreation. However, once we think of a professional soccer, this common denominator disappears as well, yet the word “game” stays unmodified as a proper description of the underlying activity. Similar reasoning applies to the word “vehicles” in the “no vehicles in the park rule.” As observed long ago by H. L. A. Hart:

Faced with the question whether the rule prohibiting the use of vehicles in the park is applicable to some combination of circumstances in which it appears indeterminate, all that the person called upon to answer can do is to consider . . . whether the present case resembles the plain case ‘sufficiently’ in ‘relevant’ respects.86

Hence, some words that people use have different meanings with no unifying common denominator. This language feature, however, does not produce open-endedness and indeterminacy because words are not communicated in a vacuum. Rather, they are communicated within a given conversational context. Together with the family resemblance, this context allows the person to whom the communication is addressed to determine the word’s likely meaning. For example, when the expression “John is playing games” is made in the context of a discussion of John’s dealings with people, the expression will likely mean to say that John is being insincere or not serious. In this example, family resemblance and context reduce the range of interpretive possibilities to two: “insincere” and “not serious.” Thus, the listener’s next step will be to use other available information to determine which of those interpretations is more likely than the other to be correct.

When words with a high level of generality are used in statutory and constitutional provisions, the interplay of context and family resemblances becomes critical. In any such case, the relevant background information, including the provision’s purpose and history, comes in handy.87 Another important contextual factor is the default and other rules of the law surrounding the provision88 and how it came to be understood and applied by governmental agencies.89 Under probabilism, courts will thus have to admit any evidence shedding light on the textual meaning or, alternatively, the motivation of the underlying statutory or constitutional provision. That is, any evidence capable of either increasing or decreasing the probability of a factual account as to what the provision means to say will be admissible, and judges will have to consider it. This rule of statutory and constitutional interpretation aligns with the relevancy doctrine set up by the general law of evidence.90

Setting up the exclusionary rules for the domain of legal interpretation is fairly easy. First, and most importantly, those rules should generally prioritize official records and peer-reviewed historical publications over any private opinion evidence offered by a self-motivated litigant.91 This preference aligns with the general evidentiary standards for admitting hearsay evidence92 and expert testimony.93 Private opinion evidence as to what a statutory or constitutional rule means to say generally fails to satisfy the credibility standards met by official records and professional expert testimony. Such evidence is both self-asserting and speculative.94 More often than not, it is offered on a “believe me” basis and can neither be verified nor refuted on empirical grounds. For that reason, judges should not rely on it in ascertaining the meaning of the law, save for very exceptional cases.

Under probabilism, judges should also be wary of potentially self-serving evidence coming from staffers entrusted with drafting congressional committee reports and other documents that move the legislative process forward. Those staffers have an opportunity to fabricate statutes’ backgrounds in order to influence their interpretation down the road. Mindful of that opportunity and its distortionary consequences, judges should generally refrain from using staffers’ reports as evidence documenting the statute’s history.95

Probabilism at Work

In this Part, I provide further details on how probabilism works. To that end, I analyze three textbook Supreme Court decisions that deal with statutory and constitutional interpretation: Kasten v. Saint-Gobain,96 Staples v. United States,97 and Crawford v. Washington.98 My analysis of these decisions substantiates the theoretical insights set forth in Part II. Specifically, the following analysis provides a concrete demonstration of how secondary and other probabilistic rules should operate, while underscoring their advantage over other methodologies, as well as their necessity in every case that features a genuine disagreement about the meaning of the underlying statutory or constitutional provision. This analysis also highlights the advantages of the preponderance standard and the evidence-selection rules, recommended in Part II, for decisions aiming to ascertain the meanings of unclear laws. Moreover, the analysis reveals that the Supreme Court Justices implicitly resort to probabilism in interpreting unclear statutory and constitutional provisions.

Kasten v. Saint-Gobain

In this case, the Supreme Court determined the meaning of the “antiretaliation provision” of the Fair Labor Standards Act of 1938 (“FLSA”) “that forbids employers ‘to discharge or in any other manner discriminate against any employee because such employee has filed any complaint or instituted or caused to be instituted any proceeding under or related to [the Act], or has testified or is about to testify in such proceeding.’”99 The case involved Kevin Kasten, an employee of Saint-Gobain who was laid off.100 Kasten claimed that the company terminated his employment in retaliation for numerous complaints he made to his supervisor about the location of the workers’ timeclocks.101 These complaints pointed out that the company installed the timeclocks between the area where Kasten and other workers put on and take off their protective gear and the area where they carry out their assigned tasks.102 Workers consequently received no credit and no pay for the time they spent putting on and taking off their work clothes, in violation of the FLSA.103 The company offered a different account: According to it, Kasten was dismissed because he failed to record his comings and goings on the timeclock after being repeatedly warned.104 The company also moved for summary judgment on the theory that Kasten, according to his own admission, has filed no written complaints.105 According to the company, the FLSA’s antiretaliation provision does not cover oral complaints, as indicated by its filing requirement.106 The District Court and, subsequently, the Seventh Circuit agreed with this argument: Both courts decided that oral complaints could not be filed.107

The Supreme Court disagreed with the lower courts.108 The Court’s decision, written by Justice Breyer, started off by identifying the prevalent meaning of the verb “to file” in both ordinary language and legal contexts.109 According to that meaning, “filing” refers to an official submission of a petition, claim, complaint, and the like.110 Predominantly, but not always, “filing” refers to submission of a document;111 on a smaller number of occasions, it also refers to a formal submission of an oral complaint or argument.112 By making these findings, the Court effectively determined that the probability of Kasten’s claim that the filing requirement under the FLSA can be satisfied by submitting an oral complaint, instead of a document, is relatively low, but still not negligible. The Court’s resort to frequentist probability can hardly be justified by textualism or intentionalism. This interpretive move aligns with probabilism more than with any other school of thought.

The Court then moved on to update this probability by introducing three additional sets of facts. These sets of facts included the FLSA’s historical background, underlying policy, and parallel statutes.113 As far as history is concerned, the Court found that “[i]n the years prior to the passage of the Act, illiteracy rates were particularly high among the poor.”114 This fact supported the broad interpretation of “filing” as including a worker’s submission of an oral complaint or grievance.115 With regard to policy, the Court noticed that effective implementation of Congress’ antiretaliation policy critically depends on the government’s ability to “us[e] hotlines, interviews, and other oral methods of receiving complaints”116 and on the presence of “informal workplace grievance procedures.”117 This fact, too, supported the proposition that Congress did not intend to impose a documentary filing requirement as a prerequisite for the worker’s protection against her employer’s retaliation. As far as parallel statutes are concerned, the Court found out that their antiretaliation provisions use a different language that is broader than the words “has filed any complaint” that appear in the FLSA.118 According to the Court, this difference lends support to two equally plausible scenarios: “(1) that Congress wanted to limit the scope of the phrase before us to writings, or (2) . . . Congress did not believe the different phraseology made a significant difference in this respect.”119 Because these scenarios are equally plausible, the Court concluded that the FLSA’s neighbor statutes neither increase nor decrease the updated probability of Kasten’s argument that “filing” can be oral as well.120

The outcome of the Court’s analysis was that “filing” can be oral and need not necessarily be documentary.121 The Court also decided to accord Skidmore deference122 to the Department of Labor’s view “that the words ‘filed any complaint’ cover oral, as well as written, complaints.”123

Against the Court’s conclusion, the company argued that “filing” should be interpreted as referring to the submission of a document to ensure that the employer receives a fair notice about the employee’s complaint.124 The Court rejected this argument on the theory that the filing of an oral complaint puts the employer on notice as well.125

Another argument made by the company merits special attention. Arguably, because retaliation against the complaining employee is a criminal offense, the lenity rule requires that “filing” be construed according to its narrow meaning as encompassing documentary filing only.126 In evaluating this argument, the Court noticed that the lenity rule “leads us to favor a more lenient interpretation of a criminal statute ‘when, after consulting traditional canons of statutory construction, we are left with an ambiguous statute,’”127 and that it “can apply when a statute with criminal sanctions is applied in a noncriminal context.”128 The Court nonetheless held that the lenity rule was not applicable, for the following reason: “[A]fter engaging in traditional methods of statutory interpretation, we cannot find that the statute remains sufficiently ambiguous to warrant application of the rule of lenity here.”129 In other words, the lenity rule does not apply because the meaning of “filing” as covering filings of oral complaints as well was more probable than the narrow meaning’s requirement of writing. This part of the Court’s decision therefore comes very close to expressly endorsing probabilism.

Remarkably, the elements of probabilism are also present in Justice Scalia’s dissent, joined in part by Justice Thomas.130 Justice Scalia agreed, arguendo, with the Court’s interpretation of “filing” as including oral complaints submitted by employees.131 Yet, according to him, the FLSA does not cover complaints to the employer at all because “every definition of the verb ‘filed’ that the Court’s opinion provides, whether it supports the inclusion of oral content or not, envisions a formal, prescribed process of delivery or submission”;132 and if so, the FLSA refers to complaints submitted to governmental agencies, as opposed to intracompany grievances and complaints.133 Aware of the fact that the statutory provision speaks about “any complaint,” Justice Scalia stated that although “it is . . . possible to speak of ‘filing a complaint’ with an employer, . . . that is assuredly not common usage.”134 Hence, the dissenters, too, based their decision upon probability. According to the dissenters, the Court’s interpretation of that requirement has a probability of being factually correct, but that probability fell below the probability that attached to their understanding of “filing.”

Both the Court’s decision and the dissent illustrate the indispensability of secondary rules for the process of ascertaining the meanings of unclear statutory and constitutional provisions. The majority Justices and the dissenters sharply disagreed about the meaning of “filing” for purposes of the FLSA, yet all of them proceeded on the implicit assumption that determination of that meaning can only be accomplished under a predetermined probability threshold and that such thresholds are necessary to avoid impasse. Moreover, both the majority and the dissent seem to have implicitly adopted in this connection the “more likely than not” standard. This adoption is manifested in the Court’s refusal to grant the company the lenity rule’s protection after finding that the two competing meanings of “filing” are not equally probable;135 and it is also implicit in the dissent’s preference for the common over the uncommon usage of the same term when everything else is equal.136

As far as evidence is concerned, the FLSA’s historical background that included widespread illiteracy among poor workers at the onset of the twentieth century played a critical role in the Court’s decision. This evidence tilted the scales in favor of the Court’s decision to interpret “filing” as covering oral complaints as well. Notice that this interpretive move was decidedly purposivist, if not pragmatist, given the FLSA’s list of employees’ actions that could not be retaliated against. Reminiscent of a catalog,137 this list included: (1) filing any complaint; (2) instituting any proceeding; and (3) testifying in any such proceeding. The common denominator of complaints filed and proceedings instituted or carried on is formal process. From this angle, only a formal complaint could satisfy the statute’s “filing” requirement—an interpretation virtually dictated by the ejusdem generis canon138 that both textualism and intentionalism endorse.139 By allowing the worker illiteracy factor to override this canon of statutory interpretation, the Court’s decision separated itself from these originalist schools of thought altogether.

The Court’s decision further exposed the inherent weakness of purposivism: overdetermination. The FLSA unquestionably aims to enhance workers’ welfare by protecting against retaliation the whistleblowers who complain about improper employment practices. Still, why assume that making informal complaints eligible for that protection would enhance workers’ welfare? Arguably, recognizing informal complaints as properly “filed” diverts employees from the proper filing practice and opens up an opportunity for making opportunistic retaliation claims. Moreover, given that worker illiteracy is no longer a big problem in the United States, purposivism seems to support the dissent’s interpretation of the statute as limited to complaints submitted to governmental agencies, if not the more formal requirement of writing. This interpretation, however, runs contrary to the evidence as to what the statute intended to communicate as a matter of fact.

The dissenting Justices seem to have lived up to their reputation as textualists: They confined the focus of their inquiry to the relevant statutory texts and the Webster’s, Cambridge, and Oxford dictionaries.140 As I have already noted, however, the dissenters were still unable to base their decision on these materials alone.141 Ambiguous texts cannot disambiguate themselves on their own. Their interpreters need to resort to some external rule or procedure that will tell them how to go about textual uncertainties, and this is exactly what Justice Scalia did. After acknowledging that “filing” in the FLSA may well encompass intracompany filings of employees’ complaints, he rejected that understanding of the term “filing” for being uncommon and hence not as probable as the common understanding.142 For Justice Scalia, the meaning of “filing” as limited to complaints submitted to governmental agencies consequently won the interpretation tournament on points. Notice, however, that Justice Scalia’s points system was extraneous to the statutory text. Similarly to the Court’s criteria for ascertaining what “filing” means, this system incorporated the laws of probability, not the laws of Congress.

Regarding the merits of the case, the Court’s and the dissent’s interpretations of the term “filing” are equally plausible. Under such circumstances, probabilism requires judges to apply the relevant tiebreaking rule. In the case at bar, one could think of two tiebreakers: (1) a rebuttable presumption that favors a common over an uncommon meaning of the disputed word or provision;143 and (2) the lenity rule that limits the scope of vague criminal prohibitions.144 These tiebreakers tilt the scales in favor of the dissenters’ interpretation of “filing” as referring to complaints submitted to a government’s agency.

Staples v. United States

In this case, the Supreme Court considered whether a gun owner unaware of his gun’s automatic capability is criminally liable for failing to register the gun as a “machinegun” in the National Firearms Registration, in violation of the National Firearms Act (“NFA”).145 The Court answered this question in the negative and vacated the conviction of the petitioner, Harold Staples, whose claim that he did not know that his unregistered rifle, AR-15, was converted into a machinegun stood unrefuted.146

Justice Thomas wrote the Court’s decision. He reasoned that the common law mens rea requirement for convictions, which applies to all criminal statutes that do not rule it out expressly or implicitly, equally applies to the “failure to register a machinegun” offense under the NFA.147 Justice Thomas’ starting point was that whether or not a defendant must be aware of the characteristics of his unregistered weapon that made it a “machinegun” under the NFA is a question of statutory construction.148 Justice Thomas then went on to reason that since the statute was silent on that question, the Court should “construe [it] in light of the background rules of the common law,”149 according to which the “mens rea [requirement] is the rule of, rather than the exception to, the principles of Anglo–American criminal jurisprudence.”150 That is, there is a general “presumption favoring mens rea,”151 and the question that the Court needs to answer is whether the NFA implicitly negated that presumption.152

The government argued that it did.153 According to the government, the NFA has created a regulatory offense under which defendants are strictly liable for failure to register machineguns and other enumerated firearms.154 This offense—so went the argument—is akin to undocumented sales of addictive narcotics punishable under the Narcotic Act of 1914, which the Court interpreted as imposing strict liability in United States v. Balint155 and other decisions.156 The government also relied on the Court’s decision in United States v. Freed,157 which interpreted the underlying provision of the NFA158 as imposing strict liability for possession of unregistered grenades.159

Justice Thomas flatly rejected these analogies. Gun ownership, he explained, is not comparable with possessing dangerous narcotics or grenades because, “despite their potential for harm, guns generally can be owned in perfect innocence.”160 Furthermore, “a long tradition of widespread lawful gun ownership by private individuals” across the United States “did not apply to the possession of hand grenades[, as] in Freed[,] or to the selling of dangerous drugs[, as] in Balint.”161 Under this tradition and the practices evolving therefrom, “buying a shotgun or rifle is a simple transaction that would not alert a person to regulation any more than would buying a car.”162 Based on these facts, Justice Thomas concluded that a person must be aware of the automated weapon characteristics of her unregistered gun in order to be guilty of the “failure to register” offense.163 According to him, this reading of the relevant NFA provisions receives further confirmation from the potentially harsh penalty—up to ten years of imprisonment—faced by the offenders.164

The Staples decision provides an even more vivid illustration of probabilism than Kasten. To begin, review the presumption favoring the mens rea requirement that constitutes one of the pillars of the Supreme Court’s criminal jurisprudence.165 This rebuttable presumption is a secondary rule, according to my terminology. Under this presumption, when a statutory offense can be interpreted as both requiring and not requiring mens rea, then, with all other factors being equal, the court should interpret the offense as incorporating the mens rea requirement among its elements.166 As a corollary, when the relevant background facts—such as history, tradition, and the legislator’s usage—indicate that the legislator intended to create a regulatory offense, the court should interpret the offense as imposing strict liability.167

This interpretive mechanism resonates with the preponderance requirement under the law of evidence, and Justice Thomas applied it precisely in this way. As he expressly recognized, dangerousness of the device motivating Congress to require its registration and to criminalize failure to register is a matter of degree. From that perspective, he estimated that, as per American tradition, guns are closer to automobiles than to dangerous narcotics and grenades168 and concluded that failure to register a particular type of gun cannot be a regulatory strict-liability offense absent an explicit statutory provision to that effect.169 This reasoning indicates that courts should interpret the statute as intending to impose strict liability for failure to register a device when the device gets closer to grenades or narcotics than to automobiles on the traditional scale of dangerousness. In other words, in order for the court to set aside the presumption in favor of mens rea and read strict liability into a statutory offense, it must make a factual finding that the disputed gun falls within the “abnormally” rather than “normally” dangerous category.

Justice Thomas was well aware of the fact that this reasoning required further support. After all, Congress could also have been understood as imposing its registration requirement on the enumerated types of guns on the theory that such weapons fall within an in-between category of dangerous devices that stands between the “normally” and the “abnormally” dangerous categories.170 This understanding may be good enough to override the mens rea presumption.171 For that reason, presumably, Justice Thomas also threw the severe punishment factor on the scales.172 As he properly observed, a severe prison sentence is generally not consistent with Congress’s intent to impose strict liability.173 This factor, too, was not decisive because Congress also prescribed a severe prison sentence for possession of grenades—an offense interpreted by the Court in Freed as requiring no mens rea.174 Yet, as indicated by Justice Thomas, the interpretation adopted by Freed was merely a shortcut to the proof of a guilty mind because possession of grenades, unlike that of machineguns, could “not [be] entirely ‘innocent’ in and of itself.”175 This part of Justice Thomas’s reasoning, too, relied on probability. Although grenade possession is not innocent in and of itself in a greater number of cases than possession of machineguns, the probability of a grenade possessor being criminally motivated does not appear to be significantly greater than that of machinegun owners. Justice Thomas’s interpretation of the offense consequently won the day on the balance of probabilities: It was factually more probable than not that Congress did not intend to impose strict criminal liability coupled with a lengthy prison sentence upon those who inadvertently fail to register their machineguns in the National Firearms Registration.

Under probabilism, Justice Thomas’s decision was correct in its bottom line. Yet, its implicit reliance on probabilism within a purportedly textualist analysis of the statute was problematic. This reliance was problematic not because of its methodological mismatch, which had no effect on the resolution of the case at bar. Rather, it was problematic because it allows judges to avoid calling probabilities by their true name. As a consequence, probabilities resorted to by judges as part of their effort at ascertaining the meanings of unclear statutory and constitutional provisions often remain unanalyzed, unassessed, and unchecked.

Crawford v. Washington

The Supreme Court’s decision in Crawford v. Washington176 ended a century of pragmatism in the Sixth Amendment Confrontation Clause jurisprudence and moved it to the domain of principle. The Confrontation Clause commands that “[i]n all criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him.”177 The words “to be confronted” unquestionably include cross-examination of the prosecution’s witnesses.178 The word “witnesses” clearly refers to people communicating information upon which the jury (or the judge in a bench trial) can determine that the defendant committed the crime that the prosecution accuses him of. Whether this word refers to any such communicator or only to people who formally bear testimony against the defendant is less clear. Under the broad interpretation of the word “witness,” the Sixth Amendment would ban not only in-court testimony of a person whom the defendant was not given an opportunity to cross-examine, but also any factual account given by a person who does not testify in court and whose out-of-court statement incriminating the defendant is adduced into evidence by the prosecution. As a consequence, all exceptions to the rule against hearsay that render admissible inculpatory statements given out of court by non-testifying declarants would become unconstitutional and void.179 Under the narrow interpretation of the word “witness,” the cross-examination requirement as a constitutional prerequisite for admitting inculpatory statements into evidence would only apply to statements that possess the formal characteristics of testimony.180 Admitting any such statement into evidence would consequently be constitutional only when the defendant gets the opportunity to cross-examine the person who made it.181 Absent such opportunity, the prosecution would not be able to use the statement as evidence against the defendant.182 At the same time, hearsay exceptions allowing the prosecution to rely upon informal—and hence nontestimonial—statements as evidence against the defendant would be altogether exempt from constitutional scrutiny.183

Led by Justice Scalia, the Court in Crawford ascribed to the word “witness” the narrow meaning.184 Remarkably, this interpretation of the word did not come instead of the broad-interpretation alternative. Rather, it replaced the pragmatic approach to inculpatory hearsay statements that the Court followed for decades without ever attempting to find out what the word “witness” means. By following that approach, the Court upheld the constitutionality of hearsay exceptions that rendered admissible inculpatory statements on the combined grounds of necessity and reliability.185

Initially, the Court allowed trial judges to admit a facially reliable out-of-court statement as evidence against the defendant when the person who made that statement—the declarant—could not testify in court.186 Subsequently, it refined this general approach by engineering three constitutionality standards that functioned as alternatives. Under one of those standards, any time-honored, or “firmly rooted,” exception to the hearsay rule was deemed constitutional on the theory that it proved to work well and so the Framers did not intend to repeal it.187 Under another standard, the defendant’s confrontation right could be satisfied by a functionally equivalent substitute for cross-examination: The defendant’s opportunity to cross-examine the declarant before or during the trial was deemed sufficient.188 As a consequence, the Court upheld the constitutionality of hearsay exceptions for former testimony189 and for the declarant’s prior statements inconsistent with her testimony in court.190 Under yet another standard, the Confrontation requirement could be satisfied by a hearsay exception that allowed an inculpatory hearsay statement into evidence when the statement showed indicia of reliability and the declarant was unavailable to testify in court.191 Those standards operated alongside the old forfeiture doctrine that deemed the defendant whose wrongdoing deliberately prevented the declarant from testifying to have forfeited his right to cross-examination.192

Justice Scalia repealed the pragmatic approach for what it was: This approach followed pragmatism instead of being faithful to the constitutional text and the Framers’ intent.193 According to Justice Scalia, the confrontation requirement as applied to “witnesses” aimed at eliminating the objectionable practice of ex parte affidavits and the various sworn and unsworn depositions elicited by the government from its witnesses.194 The Sixth Amendment prohibits all such “flagrant inquisitorial practices”195 regardless of whether they do or do not elicit reliable statements.196 The gravamen of the Confrontation Clause is procedure, rather than reliability or whether the underlying hearsay exception is firmly rooted.197 The constitutionally mandated procedure for all inculpatory testimonial evidence is cross-examination in open court.198 Exceptionally, when the declarant is unavailable, the defendant’s ability to adequately cross-examine her in a prior proceeding would also open the doors for admitting her statement into evidence.199 And, as before, the defendant can forfeit her right to cross-examine a government’s witness by foregoing the opportunity to cross-examine her or by committing a wrongdoing that deliberately prevents the witness from testifying in court.200

Within this originalist framework, a person becomes a “witness” for the government and her statement classifies as “testimonial” whenever the law-enforcement agents subject her to an ex parte questioning. For example, when, as in Crawford, the defendant’s accomplice reveals at the police questioning that she and the defendant committed a crime, the fact that the “against penal interest” exception to the hearsay rule renders that admission admissible is immaterial.201 Because the accomplice’s statement is “testimonial,”202 the Confrontation Clause does not permit the government to use it as evidence against the defendant.203

Crawford often serves as a classic example of textualism, and understandably so: It reconnected the Confrontation Clause to the word “witness” as understood by the Framers.204 On close examination, however, this decision turns out to align with probabilism. As a purely factual matter, the Framers who drafted the Sixth Amendment were unambiguously concerned about sworn (and unsworn) affidavits and depositions elicited by the government from witnesses motivated to incriminate the accused.205 Yet, as noted by Chief Justice Rehnquist in his concurring opinion in Crawford, “it does not follow that they were similarly concerned about the Court’s broader category of testimonial statements.”206 The Framers did not envision the modern-day interrogation featuring audio- and video-recorded statements made by witnesses, nor did they account for public scrutiny and other constraints under which law-enforcement agencies carry out their investigations.207 For that reason, it is not possible to establish on factual grounds that the word “witnesses” used by the Framers in the distant past encompassed people interviewed by the police under the modern-day conditions.

At the same time, however, it is still possible to reduce the Confrontation Clause into a broad principle that arguably motivated the Framers; and I posit that Justice Scalia did exactly that.208 This principle separates event-statements that incriminate the defendant from proceeding-statements that tend to do the same.209 The event-statement category includes any statement that the declarant makes, explicitly or implicitly, during any event outside the framework of police questioning and other legal proceedings.210 Reflecting an unmediated interaction between the declarant and the events on trial, such statements fall outside the purview of the Sixth Amendment. The proceeding-statement category, on the other hand, encompasses statements originating from the declarant’s interaction with the government.211 The confrontation requirement should apply to any proceeding-statement offered as evidence of guilt because the Framers put it in place to enable criminal defendants to counterbalance the government’s excessive control over such statements.212 Hence, any person making a proceeding-statement that implicates the defendant counts as a “witness” under the Confrontation Clause.

  There is no certainty the Framers had this outcome in mind. Yet, there is a strong indication that they proceeded on the “counterbalance” principle when they drafted the Sixth Amendment.213 Historical evidence214 cited by Justice Scalia,215 indicates that the Framers were deeply concerned about the expansion of the government’s power over criminal trial.216 The core of that concern was the government’s control over inculpatory witness accounts.217 This evidence indicates that Justice Scalia’s finding as to what the word “witness” means was most likely correct. The reasoning upon which Justice Scalia based that finding was virtually flawless, too; and it would have been absolutely—rather than virtually—flawless had he not claimed to have identified the Framers’ intent with certainty and relied on the intent’s probability instead.

Probabilism and Bostock

This part of the Article illustrates probabilism from yet another angle through discussion of the recent Supreme Court decision, Bostock v. Clayton County.218 This decision has interpreted the word “sex” in the Title VII prohibition of workplace discrimination on the basis of “race, color, religion, sex, or national origin” as encompassing sexual orientation as well.219 According to the Court, the prohibition of workplace discrimination against a member of the LGBT community is an unintended—yet logically inevitable—consequence of the drafters’ broad wording that outlawed all forms of employment-related discrimination of males as males and of females as females. An employer’s adverse treatment of a male employee for being gay, transgender, or bisexual penalizes that male employee for not behaving as the employer expects males to behave.220 By the same token, an employer’s adverse treatment of a female employee for being lesbian, transgender, or bisexual penalizes that female employee for not behaving as the employer expects females to behave.221 In each of these cases, the employee’s maleness or femaleness plays a role in the employer’s decision, in violation of Title VII.222

The Court based this decision on textualism.223 This path-breaking ruling was described by the dissent, written by Justice Alito, as breathtaking arrogance.224 The dissent analogized the Court’s allusion to textualism to a “pirate ship” that “sails under a textualist flag, but what it actually represents is a theory of statutory interpretation . . . [authorizing] courts [to] ‘update’ old statutes so that they better reflect the current values of society.”225 According to the dissent, the word “sex,” as used back in 1964 when the Civil Rights Act was put in place, could not, and hence did not, refer to a person’s sexual orientation.226

As far as logic is concerned, to demonstrate that sex discrimination and discrimination on the basis of sexual orientation are not overlapping, the dissent compiled the following table:

Table 1

According to the dissent, this table demonstrates that men and women in Bostock have received equal treatment from their employers. In the employers’ eyes, sexual orientation motivating disparate treatment of employees is a reason separate from sex. The dissent held that discrimination based upon sexual orientation is not a sex-discrimination prohibited by Title VII, however morally wrong it may be.227 To further support this argument, the dissent mentioned a few unaccomplished legislative initiatives aimed at protecting employees from being discriminated against for being lesbian, gay, bisexual, or transgender.228 According to the dissent, these initiatives represent the general understanding that the word “sex” in Title VII does not encompass sexual orientation.229

As far as pure logic is concerned, the dissent was right. From a purely logical standpoint, an employer who discriminates against lesbian, gay, bisexual, and transgender employees does not treat a male employee differently from a similarly situated female employee. Rather, it discriminates among males and among females alike. Such discrimination is morally repugnant, but arguably not unlawful.

Furthermore, the dissent’s claim that the word “sex” does not logically include “sexual orientation” could be made even stronger by considering the community of asexual people.230 Consider an employer who fires every employee, male or female, discovered to be asexual.231 This form of discrimination is depicted in the table below:

Table 2


Under this scenario, the employer treats male and female employees equally, while penalizing both males and females for being asexual, that is, for behaving “unmanly” and “unfemininely” at once. If so, what makes the LGBT discrimination different from this scenario? In the LGBT discrimination case, men are discriminated against for behaving “unmanly” and women for behaving “unfemininely”—but why should it make a difference? Could it be the case that Title VII prohibits employers from penalizing female employees for acting “unfemininely” while allowing them to adversely treat the same employees for acting “unfemininely” while exhibiting no maleness either?

This weakness in the Court’s reasoning originates from its implicit assumption that the nexus between “sex” and “sexual orientation” should be determined analytically rather than empirically. Yet, the fact that the analytical ground underneath the Court’s decision is shaky does not establish that this decision was wrong. The Court’s decision could still be vindicated on empirical grounds. Arguably, by banning all kinds of employment-related sex discrimination, Title VII meant to say—quite simply, and as a matter of fact—that employers should treat all sex-related reasons as not relevant for hiring, for assigning jobs to, as well as for promoting and for firing employees. This prohibition originates from the simple factual understanding of the word “sex.” According to this understanding, an employer cannot discriminate against an employee for reasons that have to do with his or her maleness or femininity, both conventional and unconventional. This interpretation would extend the Title VII protection not only to the LGBT community, but also to asexual employees.

According to the dissent, this broad interpretation of Title VII would be wrong as well, albeit not analytically wrong. To reject this interpretation, the dissent relied on Title VII’s history: absence of any references to sexual orientation in the Congress discussions, as well as the fact that the word “sex” in 1964 predominantly referred to the “male vs. female” biology.232 In the dissenters’ opinion, these historical facts indicate that Congress did not intend to extend its antidiscrimination protection to males and females whose sexual orientation might motivate their employers to treat them adversely.233

Notice that by making this argument the dissent moved the debate from textualist grounds to intentionalism.234 This move was far from accidental. The dissent could easily establish, as it did, that the Court’s decision does not really align with textualism, but it was unable to fully vindicate its own interpretation of what “sex” does and does not encompass by simply stating that “[i]t defies belief to suggest that the public meaning of discrimination because of sex in 1964 encompassed discrimination on the basis of a concept that was essentially unknown to the public at that time.”235 To start with, homosexuality was far from unknown in 1964.236 Moreover, the dissent did not dispute the ban on gender-related stereotyping formulated in Price Waterhouse v. Hopkins237—a Title VII precedent featuring an accomplished female accountant not “feminine” enough to become a partner at Price Waterhouse238; and it also could not dispute that not hiring, not promoting, or firing an employee for reasons that have to do with his or her affiliation to the LGBT community entails gender stereotyping. In other words, the dissent resorted to textualism only offensively, rather than defensively, because Title VII’s words “because of sex” were open to multiple textual interpretations back in 1964, as they are now.

Does the dissent win the debate on intentionalist grounds upon demonstrating that “there is not a shred of evidence that any Member of Congress interpreted the statutory text [the] way [the Court interpreted it] when Title VII was enacted[?]”239 I believe it does not. The fact that Congress did not envision Bostock-type discrimination back in 1964 surely does not establish that it affirmatively intended not to extend its antidiscrimination protection to employees whose intimate life, or lack thereof, does not fit their employer’s taste. Congress may well have intended to set up a broad prohibition that bans every type of employment decision implicating the employee’s maleness or femininity, both conventional and unconventional.240 Note that when a legislator enacts a standard by using a broad concept, it does not envision many, if not most, of the standard’s future applications ahead of time.241 Regulating conducts unforeseeable ahead of time is the whole point of setting up a standard, as opposed to a rule.242 For example, when a legislator lays down a “good faith” requirement for business relationships,243 it may not anticipate the scenario featuring a firm willing to forego minor deviations from its agreements with suppliers to all suppliers but gay and/or lesbian suppliers, yet such a practice would most certainly constitute bad faith.244 When a legislator enacts a broad standard, it intends the standard to apply in many different circumstances, including circumstances altogether unknown at the time of the standard’s enactment. This is how standards work; and this is how they probably should work in the Bostock type of cases under the “because of sex” formula.245

The qualifier “probably” in my last sentence suggests that there is no certainty as to whether Congress intended to use the “because of sex” expression as a broad standard. Yet, the dissent’s claim that Congress intended to set up a bright-line rule is equally uncertain. To resolve this puzzle, the statute’s interpreters consequently need secondary rules that will tell them how to proceed in the face of uncertainty. Alas, intentionalism provides no such rules, and textualism, as we just saw, does not provide them either.

Enter probabilism. Under this methodology, since Title VII’s expression “because of sex” includes a general concept—“sex”—its meaning should be determined by considering family resemblance as well. Specifically, one ought to find out whether this concept, as understood in the sixties, referred not strictly to the male-female biology, as suggested by some evidence, but also, more generally, to the intimate activities associated with maleness, as opposed to femininity, or vice versa.246

Answering this empirical question turns out not to be very difficult. First and most importantly, it ought to be acknowledged at the outset that the expression “because of sex” is open textured and that its open texture—as Frederick Schauer put it, for all words and expressions, in his classic work Playing by the Rules—“is [an] indelible feature of language, a consequence of the confrontation between fixed language and a continuously changing and unknown world.”247 Empirical generalizations used at ascertaining open-textured words and expressions such as “sex” and “because of sex” are bound to be probabilistic and nonexclusive, rather than certain and exclusive.248

Based on this understanding, consider now the fundamental egalitarian premise of Title VII.249 By enacting this antidiscrimination statute, Congress clearly intended to outlaw gender-related stereotypes and prejudices that attached to women as opposed to men, as a reason for not hiring, not promoting or firing an employee.250 Importantly, the outlawed stereotypes and prejudices included not only those that were common back in 1964, but also those that were yet to develop. To secure this inclusion, Congress preferred the “because of sex” standard to a fully itemized list of specific prohibitions. Furthermore, at the time of the Civil Rights Act’s enactment, homosexuality was already a widely known, albeit not socially accepted, sexual orientation.251 These factors establish the family-resemblance relationship between Title VII’s prohibition of employment discrimination and an adverse treatment of LGBT employees. This form of discrimination does not treat female and male employees differently from each other, but it penalizes women for being “unfeminine” and men for acting “unmanly”—a morally repugnant employment practice that resembles the core sex discrimination sufficiently enough. Hence, the drafters’ ban on discrimination “because of sex” was likely enough to encompass disparate treatment of LGBT employees. This argument establishes the case for adopting the Court’s interpretation of “because of sex” by a preponderance of the evidence.

Anti-Factualism and Its Failures

Thus far, this Article proceeded on the assumption that probabilism only needs to compete with textualism and intentionalism—schools of thought premised on the principle of legislative supremacy that requires judges to act as faithful agents of the legislature.252 In large part, this assumption originated from the Supreme Court’s formulation of the playing field on which these methodologies compete against each other.253 Under that formulation, judges say what they mean and mean what they say, and so the ways in which they explain and justify their interpretive decisions determine the methodologies that are up for examination. The methodologies employed by the Supreme Court justices in ascertaining the meanings of statutory and constitutional provisions are genuine. They combine textualist and intentionalist methodologies with fact-based purposivism;254 and, as I suggested, they implicitly rely on probabilism as well. The use of these methodologies indicates that the Supreme Court did not seize the lawmakers’ power by reading its own value preferences into acts of Congress and the Constitution under the guise of interpretation.

Still, how do we know that all this is true and that there is no fiatism, pragmatism, or unqualified, fact-free, purposivism at play? This question brings me to my second and more principled ground for rejecting the descriptive accounts of judging offered by fiatism, pragmatism, and unqualified, fact-free, purposivism. Each of these accounts is anti-factual at its core, ignoring or denying judges’ displayed commitment to empirical facts and principled reasons.255

Begin with facts. Humans have been remarkably successful at ascertaining the facts needed for their survival and well-being. Alongside their other epistemic accomplishments, this success is evidenced by the scientific progress that accelerated in the past two centuries and helped people vastly improve their well-being by developing effective means of transportation, communication, education, and medical care.256 Probability, in one form or another,257 massively contributed to this success. Throughout the years, it became a standard tool for making decisions under uncertainty in virtually every human endeavor, from air-traffic control to education policy, urban planning, equalization of opportunities, prevention of global warming, and development of vaccines.258 As such, it helped people make good decisions grounded upon incomplete, yet entirely dependable, empirical evidence.259 Adjudication was no exception. As part of that process, judges have become accustomed to resolve cases and controversies based upon incomplete, yet dependable, empirical evidence pertaining to the events they are called upon to determine and to the legal commands they are called upon to interpret and apply.

For the most part, judges interpret statutes and constitutional provisions on the basis of reasons that are exposed to scrutiny. Critically, these reasons are principled: They repeat themselves in multiple cases because judges are pre-committed to apply them across the board. The reasons judges use thus have precedential value.260 As importantly, the judges’ pre-commitment to applying these reasons in all of their interpretive decisions, including decisions in cases yet to be filed, makes it very difficult for them to skew their decisions in a chosen ideological direction. By and large, such skews are not even possible. When judges pre-commit themselves to general reasons, they usually cannot foretell which cases will be filed and when. As importantly, ambiguities of language spread across all kinds of issues and ideologies—a fact of life that helps courts maintain evenhandedness in applying the chosen reasons.261

These realities make it difficult to defend fiatism as a descriptively sound account of judging. Being a skeptical theory, fiatism can only be as valid as its underlying philosophical assumptions about the unattainability of facts, the radical indeterminacy of language and the interplay between the indeterminate language and the interpreters’ power. These assumptions, however, are ill-founded because they do not account for humans’ epistemic success and fail to explain the mutual understandings and coordination that millions of people have achieved in interacting and communicating with each other. Ambiguity and the corresponding multiplicity of meanings can be ascribed to words, when viewed in isolation from each other. Yet, words and interpersonal communications are not the same. The ambiguity of words does not establish, ipso facto, that people’s communications are indeterminate rather than meaningful. Meanings are not generated by individual words. Rather, they are generated by combinations of words that form sentences and, critically, by the conversational context and conventions.262 Fiatism offers no empirical rebuttal to this simple reality of human communication.263 Worse yet: Because fiatism purports to successfully communicate its own ideas, it turns out to be self-refuting as a matter of logic.264

With pragmatism, the issue is more complicated because it does have a certain normative appeal that may project on what judges actually do in practice. Despite this appeal, pragmatism does not fit in as a description of what judges actually do. Unlike fiatism, pragmatism proceeds on the assumption that judges have enough epistemic competence to formulate sound social policies and implement those policies in their decisions. Arguably, therefore, the same epistemic competence should enable judges to properly ascertain the actual or most likely meanings of statutory and constitutional provisions. If so, why should judges base their decisions on the vision of the law that they deem normatively attractive rather than factually correct? Pragmatism offer no satisfactory answer to this question. In fact, it avoids this question altogether by postulating that judges decide cases on normative grounds when the law is unascertainable. This assumption, however, is self-contradictory. To make a normatively sound decision that augments social welfare, judges need to process evidence and form a probabilistic judgment about relevant facts. To make a factually sound decision about the meaning of the underlying statutory or constitutional provision, judges need to do the same. Hence, if probabilistic understanding of what the law says is not within judges’ reach, judges should be equally unable to make pragmatically sound decisions. Further, if pragmatist judging is a viable endeavor, as I believe is the case, the probabilistically sound understanding of what the law says should remain a viable result as well.

To illustrate this point, consider whether pragmatism can explain the evolvement of the Sixth Amendment Confrontation doctrine. The reliability approach that reigned before Crawford seems to have satisfied the pragmatist demands, and yet it was replaced by the text-centered distinction between testimonial hearsay statements, against which criminal defendants receive constitutional protection, and nontestimonial hearsay statements altogether exempt from constitutional scrutiny.265 To be sure, this distinction can be explained on pragmatic grounds as well: for example, as economizing on the appellate courts’ efforts at reviewing criminal convictions. Yet, this line of rationalization offers no satisfactory explanation for the successful effort made by the Supreme Court at ascertaining the actual meaning of the word “witness” and in deciding to shift away from the reliability approach that worked reasonably well.266 Worse yet, the pragmatist rationalization portrays the Court as camouflaging pragmatism that it did not previously hide from the public view. Indeed, if society is better off under Crawford than under the previous regime, why hide this critical advantage behind the word “witness”? Assuming, arguendo, that there is nothing wrong about deciding hard cases on pragmatic grounds, had the Court actually decided to follow pragmatism, it would have done so expressly rather than clandestinely, as it did before Crawford.

The same critique applies to unqualified, fact-free, purposivism as well. Purposivism holds that judges need to identify the rule’s implicit purpose and determine the rule’s meaning in accordance with that purpose—a methodology that involves a fair measure of self-policing.267 Thus, if there are empirical indications that the rule banning vehicles from parks was put in place to prevent recreational lands from being polluted, a purposivist judge should interpret the ban as not applying to “green” vehicles powered by clean energy. This interpretation aligns with the intentionalist methodology and is therefore easy to defend. Things become more complicated and, indeed, more controversial when there are no empirical indications as to what the rule’s underlying purpose is or might be. How should a purposivist judge proceed in the absence of such indications? Under the unqualified or “strong” version of purposivism,268 the judge should identify the rule’s purpose on normative grounds.269 Unpacking these grounds, however, shows that there is virtually no difference between fact-free purposivism and pragmatism.

Assume that there are solid normative reasons for believing that parks should be open for “green” vehicles. To be normatively justifiable, however, these reasons must account for congestion and other social costs incurred by allowing “green” vehicles into parks. That is, the judge must step into the lawmaker’s shoes and interpret the word “vehicle” not by reference to its factual meaning, but, instead, by juxtaposing the social benefits against the social costs of the narrow interpretation separating “green” vehicles from the means of transportation that use fossil fuels. This methodology would be no different from pragmatism. A pragmatist judge in my present example will adopt the same narrow interpretation of the word “vehicle” in order to maximize social welfare.270 As with pragmatism, this approach can only be attractive if the meanings of unclear statutory and constitutional provisions could not be determined probabilistically, and if judges had good reasons for hiding their inability or unwillingness to ascertain the factual meaning of the law. My analysis of the Crawford decision, however, shows that neither of these assumptions holds true. First, Justices do not shy away from openly deciding certain cases on pragmatist grounds. Second, as a matter of sheer logic, probability attaches to any factual proposition,271 including propositions about the meaning of a communication coming from the lawmaker.

The takeaway from this discussion is straightforward. To establish their viability as a theory, pragmatism, unabashed purposivism, and fiatism need to prove that factual meanings of legal rules are unascertainable. The requisite proof might involve some form of global epistemic skepticism. Global epistemic skepticism, however, is a descriptively unsustainable, as well as potentially self-refuting, worldview.272 To make their claims plausible, pragmatism, unabashed purposivism, and fiatism consequently need to develop, in one form or another, an exceptionalist account of adjudication that somehow relates epistemic skepticism to judging. Specifically, they need to establish that there is something special, if not altogether unique, about legal interpretation that makes it impossible for judges to make sound probabilistic decisions about the factual meaning of the law.

An attempt in that direction was made by critical legal theorists that claim to have uncovered contradictions and inconsistencies in legal doctrine.273 Some of these theorists even offered an explanatory account of those contradictions and inconsistencies by associating them with people’s mutually incompatible demands for freedom and security.274 This association may or may not be true and needs to be proven rather than simply postulated. Furthermore, even if it were factually accurate, using it as a basis for an observation that statutory and constitutional rules embody contradictory ideas and are consequently indeterminable would involve a plain logical error. People’s failure to reconcile between mutually incompatible demands is a feature of the open-ended normative discourse that need not be transformed into the law. People need not coalesce around a single coherent vision of the good in order to set up legal rules that will secure their mutually beneficial coexistence and enable them to coordinate their endeavors.275 They can design for themselves legal rules without forming an agreement about moral fundamentals and other deep worldviews. Whether they choose to do so or not is a matter of empirical fact. Understanding what legal rules say as a matter of fact consequently requires a strictly empirical investigation that extrapolates probability from the evidence. The law is what this empirical investigation shows it to be.


Alongside their differences, textualism, intentionalism, purposivism, pragmatism, and fiatism share one common feature, identified as absolutism. Each of these schools of thought views legal interpretation as a special type of decision-making endeavor that only succeeds when it manages to ascertain the true meaning of an unclear statutory or constitutional provision. Under textualism, the requisite meaning must be extracted from the provision’s words and sentences, as understood at the time of its enactment. Under intentionalism, the requisite meaning is present in the drafters’ intent, while under purposivism it is embedded in the provision’s purpose. Under pragmatism and fiatism, ascertainment of the provision’s true meaning is doomed to fail because the requisite meaning does not exist (as under fiatism) or is beyond the courts’ reach, except when the provision’s words are straightforward and require no interpretation (as under pragmatism).

The probabilistic approach to legal interpretation, advocated in this Article, separates itself from these schools of thought in two ways. First, it denies the schools’ absolutist assumption and the consequent “true or false” approach to the meaning of the law. Under probabilism, the meanings of statutory and constitutional provisions admit to different degrees of probability and thus can be more and less likely true. Second, probabilism repudiates other schools’ underlying assumption that legal interpretation is a special type of decision-making. Under probabilism, legal interpretation is not special in any sense. Rather, it is just another factfinding endeavor—one that courts should carry out in the same way in which they determine disputed facts under conditions of uncertainty. Correspondingly, when the meaning of the underlying statutory or constitutional provision is uncertain, courts ought to abandon the absolutist quest for certainty. Instead of trying to find out the provision’s true meaning, courts will do well to identify the meaning that is most likely to be true as a matter of fact. Under this framework, courts will have to answer the following question on a balance of probabilities: “What was communicated by the individuals who drafted the provision by the language they chose to use?” In answering this factual question, courts should take advantage of all relevant information: the provision’s wording, including “family resemblances,” as understood at the time of the enactment; as well as the provision’s history, background and purpose. In cases of impasse, courts should resort to tie-breaking rules such as lenity and the presumption that the drafters authorized courts to update the provision’s original public meaning after a passage of time. Admittedly, this methodology will not always allow courts to deliver the right answer. Yet, courts following probabilism will deliver the right answer in the majority of the cases. For that simple reason, probabilism is preferable to all other schools of legal interpretation.

  1. [1]. See generally Philosophical Foundations of Language in the Law (Andrei Marmor & Scott Soames eds., 2011) (outlining different theoretical frameworks for understanding the relationship between language and law).

  2. [2]. See, e.g., Victoria Nourse, Textualism 3.0: Statutory Interpretation After Justice Scalia, 70 Ala. L. Rev. 667, 676–84 (2019) (explaining textualism and illustrating it by court decisions); Caleb Nelson, What Is Textualism?, 91 Va. L. Rev. 347, 351–53 (2005) (separating textualism from intentionalism as a school of interpretation that underscores the significance of ordinary language and fair notice).

  3. [3]. See, e.g., Larry Alexander, Connecting the Rule of Recognition and Intentionalist Interpretation: An Essay in Honor of Richard Kay, 52 Conn. L. Rev. 1513, 1522–25 (2021) (describing intentionalism as a school of thought that treats interpretation of legal texts as an empirical endeavor focusing on the immanent nexus between text and authorial intent).

  4. [4]. See generally Ilan Wurman, A Debt Against the Living: An Introduction to Originalism 11–24 (2017) (describing the origins and development of originalism); see also John O. McGinnis & Michael B. Rappaport, Unifying Original Intent and Original Public Meaning, 113 Nw. U.L. Rev. 1371, 1375–88 (2019) (explaining textualism and intentionalism as branches of one broad originalist school of thought).

  5. [5]. See, e.g., Kevin Tobia & John Mikhail, Two Types of Empirical Textualism, 86 Brook. L. Rev. 461, 461–62 (2021) (“Underpinning ... ‘empirical textualism’ is a set of observations about the relationship between ordinary meaning and ordinary language users. First, interpreting a legal text in line with its ordinary meaning promotes rule of law values like publicity and fair notice. The law should be publicly available to ordinary people, in other words, and it should enable members of the public to rely upon and form reasonable expectations about it. Ordinary meaning analysis is thus often taken to promote democracy; as such, its focus is naturally placed on the understanding of the demos. Similarly, ordinary meaning analysis is taken to prevent judicial overreach. It is the public’s common understanding of the text that matters, so the logic runs, not the potentially biased views of unelected judges. More broadly, in centering interpretation on ordinary meaning, empirical textualism promises an alluring objectivity.”).

  6. [6]. See, e.g., Neil H. Buchanan & Michael C. Dorf,A Tale of Two Formalisms: How Law and Economics Mirrors Originalism and Textualism, 106 Cornell L. Rev. 591, 628 (2021) (observing that “taking account of the nature of distinctlylegaltexts tends to reinforce the appeal of intentionalism[ because] ... (1) The People choose our lawmakers, whether via special processes for constitutions or through periodic elections for legislators; (2) the enactments of our lawmakers are legitimately law as a consequence of that democratic/republican pedigree; (3) thus, when uncertainty about the content of the law arises, it should be resolved in favor of the original intentions and expectations of the lawmakers and the People they represented, rather than in accordance with some implication of the words they used, at least if that implication would have surprised them”).

  7. [7]. See William N. Eskridge, Jr. & Philip P. Frickey, Statutory Interpretation as Practical Reasoning, 42 Stan. L. Rev. 321, 324 (1990) (“The three main theories today emphasize (1) the actual or presumed intent of the legislature enacting the statute (‘intentionalism’); (2) the actual or presumed purpose of the statute (‘purposivism’ or ‘modified intentionalism’); and (3) the literal commands of the statutory text (‘textualism’).”); John F. Manning & Matthew C. Stephenson, Legislation and Regulation: Cases and Materials 24 (3d ed. 2017) (connecting textualism, intentionalism, and purposivism to “the principle of legislative supremacy, which encapsulates the related ideas that in the U.S. constitutional system, acts of Congress enjoy primacy as long as they remain within constitutional bounds, and that judges must act as Congress’s faithful agents”); see also John F. Manning, What Divides Textualists from Purposivists?, 106 Colum. L. Rev. 70, 85–91 (2006) [hereinafter What Divides Textualists from Purposivists?] (explaining purposivism and its proximity to the modern-day textualism); Anita S. Krishnakumar, Backdoor Purposivism, 69 Duke L.J. 1275, 1295–304 (2020) (arguing that interpretive canons applied by the U.S. Supreme Court entrench purposivism disguised as textualism).

  8. [8]. See Samuel L. Bray, The Mischief Rule, 109 Geo. L.J. 967, 970 (2021) (“The mischief rule serves two functions. First, a stopping-point function: it offers a rationale for an interpreter’s choice about how broadly to read a term or provision in a legal text. Second, a clever-evasion function: it allows an interpreter to read a legal text a little more broadly to prevent a clever evasion that would perpetuate the mischief. Of these two, the stopping-point function is much more common.”)(footnote omitted); see also Henry M. Hart, Jr. & Albert M. Sacks, The Legal Process: Basic Problems in the Making and Application of Law, in The Canon of American Legal Thought 241 (David Kennedy & William W. Fisher eds., 2007) (stating the mischief-focused method of interpreting statutes); Stuart Minor Benjamin & Kristen M. Renberg, The Paradoxical Impact of Scalia’s Campaign Against Legislative History, 105 Cornell L. Rev. 1023, 1033–45 (2020) (rationalizing purposivism as a method leading to enhancement of the government’s regulatory powers and hence disfavored by conservative textualists).

  9. [9]. See Richard A. Posner, What Has Pragmatism to Offer Law?, 63 S. Cal. L. Rev. 1653, 1656
    –60 (1990) (explaining pragmatism and its focus on augmentation of social welfare); see also Cass R. Sunstein, Justice Breyer’s Democratic Pragmatism, 115 Yale L.J. 1719, 1729–31 (2006) (discussing Justice Breyer’s view that social consequences matter to the choice of a theory of legal interpretation).

  10. [10]. See Benjamin N. Cardozo, The Nature of the Judicial Process 66 (1921) (“The final cause of law is the welfare of society.”).

  11. [11]. Cf. Thomas W. Merrill, Legitimate Interpretation—Or Legitimate Adjudication?, 105 Cornell L. Rev. 1395, 1397–400 (2020) (setting forth a model of legitimate adjudication featuring judges who act as faithful agents by integrating statutory and constitutional texts, precedent, settled practices, as well as qualified considerations of morality and social consequences).

  12. [12]. Borden Parker Bowne, The Passing of Educational Fiatism, 4 Pacific Phil. Quart. 77, 77 (1923). For additional examples of how this concept has been used, see the canonical—yet, still controversial—article by Herbert Wechsler, Toward Neutral Principles of Constitutional Law, 73 Harv. L. Rev. 1, 11 (1959), and see generally Lon L. Fuller, Reason and Fiat in Case Law, 59 Harv. L. Rev. 376 (1946), discussing fiat and its interaction with the law.

  13. [13]. See, e.g., Allan C. Hutchinson & Patrick J. Monahan, Law, Politics, and the Critical Legal Scholars: The Unfolding Drama of American Legal Thought, 36 Stan. L. Rev. 199, 211 (1984) (arguing that “every doctrinal dispute is reducible to the contradictory claims of communal security and individual freedom” and that “[a]ny particular resolution of that conflict is simply an arbitrary choice”).

  14. [14]. See generally John Henry Schlegel, Critical Legal Studies: An Afterword, 36 Stan. L. Rev. 673 (1984) (offering concluding thoughts on a critical legal studies symposium); Duncan Kennedy & Karl E. Klare, A Bibliography of Critical Legal Studies, 94 Yale L.J. 461 (1984) (compiling a list of critical legal studies scholarship); Nate Holdren & Eric Tucker, Marxist Theories of Law Past and Present: A Meditation Occasioned by the 25th Anniversary of Law, Labor, and Ideology, 45 Law & Soc. Inquiry 1142 (2020) (providing a survey of Marxist scholarship in the law); Richard Delgado & Jean Stefancic, Critical Race Theory: The Cutting Edge (3d ed. 2013) (discussing critical race theory); Martha L. A. Fineman, Feminist Theory and Law, 18 Harv. J.L. & Pub. Pol’y 349 (1995) (providing a discussion on feminist legal scholarship).

  15. [15]. See Steven L. Winter, The “Power” Thing, 82 Va. L. Rev. 721, 793–804 (1996) (discussing the nexus between language and power through the philosophy of Michel Foucault).

  16. [16]. Scholars have identified this type of broad judicial discretion in many areas of the law. See generally Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685 (1976) [hereinafter Kennedy, Form and Substance]; Duncan Kennedy, The Stages of the Decline of the Public/Private Distinction, 130 U. Pa. L. Rev. 1349 (1982); Jay M. Feinman, Essay, Critical Approaches to Contract Law, 30 UCLA L. Rev. 829 (1983); Joseph William Singer, The Player and the Cards: Nihilism and Legal Theory, 94 Yale L.J. 1 (1984); James Boyle, The Politics of Reason: Critical Legal Theory and Local Social Thought, 133 U. Pa. L. Rev. 685 (1985); Gary Peller, The Metaphysics of American Law, 73 Calif. L. Rev. 1151 (1985).

  17. [17]. See, e.g., Boyle, supra note 16, at 697–702.

  18. [18]. See Kennedy, Form and Substance, supra note 16, at 1732–37, 1766–76.

  19. [19]. Id. at 1732–37. See generally J. M. Balkin, Deconstructive Practice and Legal Theory, 96 Yale L.J. 743 (1987) (discussing and applying deconstruction); Clare Dalton, An Essay in the Deconstruction of Contract Doctrine, 94 Yale L.J. 997 (1985) (same).

  20. [20]. See Kennedy, Form and Substance, supra note 16, at 1732–37.

  21. [21]. See id.

  22. [22]. Cf. Nicholas S. Zeppos, Legislative History and the Interpretation of Statutes: Toward a Fact-Finding Model of Statutory Interpretation, 76 Va. L. Rev. 1295, 1335–41 (1990) (advocating a fact-focused model of statutory interpretation without resorting to probability).

  23. [23]. See Richard J. Bernstein, Beyond Objectivism and Relativism: Science, Hermeneutics, and Praxis 8–15 (1983) (claiming that objectivism vs. relativism is a false dichotomy).

  24. [24]. Id. at 8.

  25. [25]. Id.

  26. [26]. See Stephen E. Sachs, Originalism: Standard and Procedure, 135 Harv. L. Rev. 777, 780 (2022) (“Still, in practice we can’t do without a decision procedure, and good procedures are hard to find. This problem undermines some popular arguments for originalism based on its consequences, either for particular policies (say, gun rights) or for the legal system at large (say, constraining judges). The less we can find out about the original law, the less likely we are to benefit from looking for it.”).

  27. [27]. Cf. Lawrence B. Solum, The Fixation Thesis: The Role of Historical Fact in Original Meaning, 91 Notre Dame L. Rev. 1, 12 (2015) (“Interpretation is an empirical inquiry. The communicative content of a text is determined by linguistic facts (facts about conventional semantic meanings and syntax) and by facts about the context in which the text was written. Interpretations are either true or false—although in some cases we may not have sufficient evidence to show that a particular interpretation is true or false.”). Under probabilism, interpretation involves a broader empirical inquiry. As part of that inquiry, interpreters should consider not only the linguistic facts, but also the drafters’ intent and every relevant evidence. Based on the evidence they gathered, interpreters should decide—under the premise that evidence is virtually never sufficient—which of the available interpretations is most likely to be true. See infra Part II.

  28. [28]. See Solum, supra note 27, at 15, 27–30; see also Richard S. Kay, Original Intention and Public Meaning in Constitutional Interpretation, 103 Nw. U.L. Rev. 703, 719–25 (2009) (discussing “public meaning” originalism); John O. McGinnis & Michael B. Rappaport, The Power of Interpretation: Minimizing the Construction Zone, 96 Notre Dame L. Rev. 919, 921–22 (2021) (arguing that reading the Constitution as a formal legal document can substantially reduce its indeterminacy).

  29. [29]. See Roberto Mangabeira Unger,The Critical Legal Studies Movement, 96 Harv. L. Rev. 561, 566–67, 576 (1983) (“If the criticism of formalism and objectivism is the first characteristic theme of leftist movements in modern legalthought, the purely instrumental use of legal practice and legal doctrine to advance leftist aims is the second. The connection between these two activities—the skeptical critique and the strategic militancy—seems both negative and sporadic. It is negative because it remains almost entirely limited to the claim that nothing in the nature of law or in the conceptual structure of legal thought—neither objectivist nor formalist assumptions —constitutes a true obstacle to the advancement of leftist aims ... . When we took the negative ideas relentlessly to their final conclusions, we were rewarded by seeing these ideas turn into the starting points of a constructive program.”).

  30. [30]. See Andrew Altman, Critical Legal Studies: A Liberal Critique 90 (1990) (discussing critical scholars’ view that “legal rules are ‘empty vessels’ into which individuals can pour virtually any content they please”).

  31. [31]. See sources cited supra notes 7, 9.

  32. [32]. See Ronald Dworkin, Taking Rights Seriously xi, at 365–66 (1977) (laying out a theory of rights that prohibits such sacrifices and allows rights to override collective goals).

  33. [33]. Id. at 274–77.

  34. [34]. See generally Kasten v. Saint-Gobain Performance Plastics Corp., 563 U.S. 1 (2011) (analyzing the meaning of the word “filed” over time).

  35. [35]. See generally Staples v. United States, 511 U.S. 600 (1994) (considering the meaning of the word “firearm”).

  36. [36]. See generally Crawford v. Washington, 541 U.S. 36 (2004) (using historical sources to determine “the original meaning of the Confrontation Clause”).

  37. [37]. See Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1737, 1754 (2020).

  38. [38]. 42 U.S.C. § 2000e–2 (2018).

  39. [39]. See Solum, supra note 27, at 6–7.

  40. [40]. See Ludwig Wittgenstein, Philosophical Investigations: The English Text of the Third Edition 65e–71e (G.E.M. Anscombe trans., 3d ed. 1968); see also Michael Forster, Wittgenstein on Family Resemblance Concepts, in Wittgenstein’s Philosophical Investigations: A Critical Guide 66, 66–67 (Arif Ahmed ed., 2010) (identifying “Wittgenstein’s conception of ‘family resemblance’ as a characteristic of concepts”); Nicholas Griffin, Wittgenstein, Universals and Family Resemblances, 3 Canadian J. Phil. 635, 635–37 (1974) (analyzing the concept of “family resemblance”).

  41. [41]. For exposition and analysis of these standards, see Alex Stein, Foundations of Evidence Law 143–53 (2005).

  42. [42]. Id. at 150–52.

  43. [43]. Id. at 154–67; see also generally Alex Stein, Inefficient Evidence, 66 Ala. L. Rev. 423 (2015) (asserting that evidence rules create a legal system that avoids inefficiency).

  44. [44]. See Ronald J. Allen & Alex Stein, Evidence, Probability, and the Burden of Proof, 55 Ariz. L. Rev. 557, 565-75 (2013) (describing the probabilistic nature of adjudicative factfinding).

  45. [45]. See infra Part II.

  46. [46]. See Richard H. Fallon Jr., The Meaning of Legal “Meaning” and Its Implications for Theories of Legal Interpretation, 82 U. Chi. L. Rev. 1235, 1241 (2015).

  47. [47]. Id. at 1252.

  48. [48]. See A.J. Ayer, Language, Truth and Logic 8 (1936) (observing that “however strong the evidence in favour of historical statements may be, their truth can never become more than highly probable”).

  49. [49]. This observation lies at the heart of philosophical empiricism. See id.

  50. [50]. Id. at 11 (persuasively arguing that “all propositions which have factual content are empirical hypotheses; and that [a] function of an empirical hypothesis is to provide a rule for the anticipation of experience,” as well as that propositions that are neither factual nor certain a priori are nonsensical).

  51. [51]. See Solum, supra note 27, at 15-16.

  52. [52]. Id. at 12 (acknowledging that interpreters may not have sufficient evidence for ascertaining the meaning of the enactment).

  53. [53]. See Stein, supra note 41, at 143–53.

  54. [54]. Id. at 148–51.

  55. [55]. Id. at 143–48, 151–53.

  56. [56]. See McGinnis & Rappaport, supra note 28, at 938.

  57. [57]. Id. at 938, 942 (arguing that “legal interpretive rules required that the judge select the meaning that was more likely than not”).

  58. [58]. See Alex Stein, The Flawed Probabilistic Foundation of Law and Economics, 105 Nw. U.L. Rev. 199, 207–22 (2011).

  59. [59]. See id. at 235–46; see also Allen & Stein, supra note 44, at 567–71 (illustrating that courts “endorse[] the relative plausibility criterion for factual findings”).

  60. [60]. Allen & Stein, supra note 44, at 567–71.

  61. [61]. See Alex Stein, Second-Personal Evidence, in Philosophical Foundations of Evidence Law 96, 96–107 (Christian Dahlman, Alex Stein & Giovanni Tuzet, eds., 2021) (explaining how the second personal nature of legal rights, duties, and responsibilities connects to factfinding and explains the predominance of the “relative plausibility” method).

  62. [62]. Id. at 103.

  63. [63]. Allen & Stein, supra note 44, at 568.

  64. [64]. Id.

  65. [65]. See William Baude & Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079, 1099 (2017) (arguing that “canons [of interpretation] are best understood as unwritten interpretive rules — setting interpretive defaults, establishing the priority of different sources, and instructing judges in cases of uncertainty”). See generally Adam M. Samaha, On Law’s Tiebreakers, 77 U. Chi. L. Rev. 1661 (2010) (discussing tiebreaking rules and presenting justifications for their use).

  66. [66]. Cf. Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 23 (1997) (“[T]ext should not be construed strictly, and it should not be construed leniently; it should be construed reasonably, to contain all that it fairly means.”); Keith E. Whittington, Originalism: A Critical Introduction, 82 Fordham L. Rev. 375, 383–85 (2013) (discussing this method of construction); Fallon, supra note 46, at 1250–51 (same).

  67. [67]. See Whittington, supra note 66, at 385–86 (“[E]xpected applications might be helpful to later interpreters in clarifying the substantive content of the embodied constitutional rule
    ... . [K]nowing the application and its relationship to possible principles would allow us to infer the otherwise obscure rule. The insight to be gleaned is not the authoritative status of the expected application, but the apparent rule at play given that such an application is expected to follow from it.”).

  68. [68]. Cf. id. at 383; Robert H. Bork, The Constitution, Original Intent, and Economic Rights, 23 San Diego L. Rev. 823, 826 (1986) (“[A]ll an intentionalist requires is that the text, structure, and history of the Constitution provide him not with a conclusion but with a major premise. That premise states a core value that the Framers intended to protect. The intentionalist judge must then supply the minor premise in order to protect the constitutional freedom in circumstances the Framers could not foresee.”).

  69. [69]. See Cleveland v. United States, 531 U.S. 12, 25 (2000) (“[A]mbiguity concerning the ambit of criminal statutes should be resolved in favor of lenity.” (quoting Rewis v. United States, 401 U.S. 808, 812 (1971))); see also Note, The New Rule of Lenity, 119 Harv. L. Rev. 2420, 2422–23 (2006) (describing rule of lenity).

  70. [70]. See Skilling v. United States, 561 U.S. 358, 410 (2010) (interpreting criminal statute by using lenity as a fallback principle preceded by language, context and the rule against vagueness).

  71. [71]. See Samaha, supra note 65, at 1661–63.

  72. [72]. See Scott Baker & Kimberly D. Krawiec, The Penalty Default Canon, 72 Geo. Wash. L. Rev. 663, 664 (2004) (“When faced with an interpretation dispute regarding an incomplete statutory provision, courts should first endeavor to discover the reasons for statutory incompleteness. If the provision is incomplete for strategic reasons, meaning that lawmakers created an intentionally incomplete statute in an attempt to shift responsibility for the negative impacts of law to other governmental branches, then the courts should penalize lawmakers by holding that the provision is so incomplete that it amounts to an unconstitutional delegation of legislative authority.”).

  73. [73]. See Ronald J. Allen, David S. Schwartz, Michael S. Pardo & Alex Stein, An Analytical Approach to Evidence: Text, Problems, and Cases 755–61 (7th ed. 2022) [hereinafter Analytical Approach to Evidence] (explaining how presumptions operate).

  74. [74]. See Posner, supra note 9, at 1656–57, 1667 (alluding to Cardozo’s vision of common law as promoting social welfare as a guide to statutory interpretation that advises judges to proceed on the assumption that law strives “to support competitive markets and to simulate their results in situations in which market-transaction costs are prohibitive”).

  75. [75]. The inevitability of this transformation was implicitly acknowledged by two leading originalists. See John O. McGinnis & Michael B. Rappaport, Original Methods Originalism: A New Theory of Interpretation and the Case Against Construction, 103 Nw. U. L. Rev. 751, 774 (2009) (“Under the original interpretive rules, we believe that interpreters were required to select the interpretation of ambiguous and vague terms that had the stronger evidence in its favor. When the interpretation of language was unclear, the interpreter would consider the relevant originalist evidence—evidence based on text, structure, history, and intent—and select the interpretation that was supported more strongly by the evidence.” (footnotes omitted)).

  76. [76]. See Analytical Approach to Evidence, supra note 73, at 56–58, 763, 801–02, 815–16 (explaining admissibility requirements, permissive inferences, and judicial notice).

  77. [77]. See Stein, supra note 43, at 432, 443–60.

  78. [78]. See Fed. R. Evid. 401 (“Evidence is relevant if: (a) it has any tendency to make a fact more or less probable than it would be without the evidence; and (b) the fact is of consequence in determining the action.”); Fed. R. Evid. 402 (stating that relevant evidence is admissible unless otherwise excluded by the Constitution, an Act of Congress, the Federal Rules of Evidence, “or other rules prescribed by the Supreme Court”).

  79. [79]. See supra note 40 and materials cited therein.

  80. [80]. See Fed. R. Evid. 201(b)(1).

  81. [81]. See Wittgenstein, supra note 40.

  82. [82]. Id.

  83. [83]. Id. §§ 65–66.

  84. [84]. Id. § 66.

  85. [85]. Id. §§ 65–66.

  86. [86]. H. L. A. Hart, The Concept of Law 127 (Paul Craig ed., 3d ed. 2012). The first edition of this book was published in 1961. See id. at iv.

  87. [87]. See John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 682
    –83 (1997) (illustrating and explaining courts’ use of statutory purpose and history in interpreting opaque statutory provisions).

  88. [88]. See Baude & Sachs, supra note 65, at 1108–12 (rationalizing the use of interpretive defaults).

  89. [89]. See, e.g., City of Arlington v. Fed. Commc’ns Comm’n, 569 U.S. 290, 302 (2013) (“We ... deferred under Chevron to the Commission’s ‘eminently reasonable ... interpretation of the statute it is entrusted to administer’ . . . .” (second omission in original) (quoting Commodity Futures Trading Comm’n v. Schor, 478 U.S. 833, 844 (1986))).

  90. [90]. See, e.g., Analytical Approach to Evidence, supra note 73, at 65–70.

  91. [91]. This factfinding method originates from the “best evidence” principle. See generally Dale A. Nance, The Best Evidence Principle, 73 Iowa L. Rev. 227 (1988).

  92. [92]. See Stein, supra note 43, at 444–50.

  93. [93]. Id. at 459–60.

  94. [94]. Id. at 444, 455.

  95. [95]. See Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring) (“As anyone familiar with modern-day drafting of congressional committee reports is well aware, the references to the cases were inserted, at best by a committee staff member on his or her own initiative, and at worst by a committee staff member at the suggestion of a lawyer-lobbyist; and the purpose of those references was not primarily to inform the Members of Congress what the bill meant
    ... but rather to influence judicial construction. What a heady feeling it must be for a young staffer, to know that his or her citation of obscure district court cases can transform them into the law of the land, thereafter dutifully to be observed by the Supreme Court itself.”). For a more nuanced statement of the same principle, see Manning, supra note 87, at 681 (“In using such forms of legislative history, the Court often attributes a committee’s or sponsor’s declarations of intent to Congress as a whole. In the period prior to the emergence of textualism, this method of imputing legislative intent went largely unquestioned. The profusion of modern interest in the legitimacy of this practice owes much to the Court’s more controversial uses of legislative history—to alter the apparent meaning of statutory texts. Yet the more common practice, affecting a far broader range of cases, has involved the judicial use of legislative history to supply the ordinary details of meaning that all statutes occasionally leave indeterminate. Because the latter use of legislative history is both less controversial and more routine, it provides a better context for examining the most basic textualist objections to legislative history as an authoritative source of legislative intent.” (footnotes omitted)).

  96. [96]. Kasten v. Saint-Gobain Performance Plastics Corp., 563 U.S. 1, 4 (2011).

  97. [97]. Staples v. United States, 511 U.S. 600, 604 (1994).

  98. [98]. Crawford v. Washington, 541 U.S. 36, 38 (2004).

  99. [99]. Kasten, 563 U.S. at 4 (emphasis omitted) (alteration in original) (quoting 29 U.S.C. § 215(a)(3) (2018)).

  100. [100]. Id.

  101. [101]. Id. at 4–5.

  102. [102]. See id.

  103. [103]. Id. at 5.

  104. [104]. Id. at 6.

  105. [105]. Id.

  106. [106]. Id. at 6–7.

  107. [107]. Id. at 6.

  108. [108]. Id. at 4, 17.

  109. [109]. Id. at 7–11.

  110. [110]. Id.

  111. [111]. Id. 9–10.

  112. [112]. Id. at 8–9.

  113. [113]. Id. at 12–13.

  114. [114]. Id. at 12 (citing E. Gordon & E. Gordon, Literacy in America 273 (2003); Dept. of Com., Bureau of Census, Sixteenth Census of the United States, 1940, Population: The Labor Force (Sample Statistics): Occupational Characteristics 60 (1943)).

  115. [115]. See id. at 12–14 (“[A] complaint must be sufficiently clear and detailed for a reasonable employer to understand it ... as an assertion of rights protected by the statute and ... can be met ... by oral complaints, as well as by written ones.”).

  116. [116]. Id. at 13.

  117. [117]. Id.

  118. [118]. Id. at 10–14.

  119. [119]. Id. at 11.

  120. [120]. See id. at 13.

  121. [121]. Id.

  122. [122]. Id. at 14–16; see also Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944) (giving weight to a persuasive interpretive view falling within an authorized agency’s area of expertise).

  123. [123]. Kasten, 563 U.S. at 15.

  124. [124]. Id. at 13–14.

  125. [125]. Id.

  126. [126]. Id. at 16.

  127. [127]. Id. (quoting United States v. Shabani, 513 U.S. 10, 17 (1994)).

  128. [128]. Id. (citing Leocal v. Ashcroft, 543 U.S. 1, 11 n.8 (2004)).

  129. [129]. Id.

  130. [130]. Id. at 17–21 (Scalia, J., dissenting).

  131. [131]. Id. at 18–19.

  132. [132]. Id. at 19–20.

  133. [133]. Id. at 21–22.

  134. [134]. Id. at 20 (emphasis added).

  135. [135]. Id. at 16 (majority opinion) (“[W]e cannot find that the statute remains sufficiently ambiguous to warrant application of the rule of lenity here.”).

  136. [136]. Id. at 20 (Scalia, J., dissenting).

  137. [137]. See Gideon Parchomovsky & Alex Stein, Essay, Catalogs, 115 Colum. L. Rev. 165, 168 (2015) (conceptualizing as “catalogs” rules that “contain a specific enumeration of proscribed conduct and a general provision empowering courts to penalize or enjoin other similar activities”).

  138. [138]. Id. at 168 n.14 (describing the ejusdem generis canon of statutory interpretation and citing sources).

  139. [139]. See Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 200 (2012) (describing the ejusdem generis canon as related to originalism).

  140. [140]. See Kasten, 563 U.S. at 18–19 (Scalia, J., dissenting) (citing dictionaries to interpret the word “complaint”).

  141. [141]. See text accompanying notes 129–33.

  142. [142]. Kasten, 563 U.S. at 20 (Scalia, J., dissenting) (“It is, I suppose, possible to speak of ‘filing a complaint’ with an employer, but that is assuredly not common usage.”).

  143. [143]. See supra notes 66–68 and accompanying text.

  144. [144]. See supra notes 69–70 and accompanying text.

  145. [145]. This Act criminalizes, inter alia, possession of an unregistered “machinegun” defined as a “weapon which shoots ... automatically more than one shot, without manual reloading, by a single function of the trigger.” See 26 U.S.C. §§ 5845(a)(6) & (b), 5861(d) (2018).

  146. [146]. Staples v. United States, 511 U.S. 600, 602–04 (1994).

  147. [147]. Id. at 604, 606.

  148. [148]. Id. at 603–04.

  149. [149]. Id. at 605 (citing United States v. U.S. Gypsum Co., 438 U.S. 422, 436–37 (1978)).

  150. [150]. Id. (quoting U.S. Gypsum Co., 438 U.S. at 436).

  151. [151]. Id. at 606.

  152. [152]. Id.

  153. [153]. Id.

  154. [154]. Id. at 606–07.

  155. [155]. United States v. Balint, 258 U.S. 250, 251–54 (1922).

  156. [156]. Staples, 511 U.S. at 605–09 (describing past cases interpreting and applying the Narcotic Act of 1914).

  157. [157]. United States v. Freed, 401 U.S. 601, 616 (1971).

  158. [158]. 26 U.S.C. § 5861(d) (2018).

  159. [159]. Staples, 511 U.S. at 608–10, 614 (noting that Freed does not require “that a defendant ... know that his weapon is unregistered” to impose only unregistered weapon NFA liability).

  160. [160]. Id. at 611.

  161. [161]. Id. at 610.

  162. [162]. Id. at 614.

  163. [163]. Id. at 619–20.

  164. [164]. Id. at 616–18.

  165. [165]. Id. at 605–06; Posters ‘N’ Things, Ltd. v. United States, 511 U.S. 513, 522–23 (1994); Liparota v. United States, 471 U.S. 419, 426 (1985); United States v. U.S. Gypsum Co., 438 U.S. 422, 436–37 (1978).

  166. [166]. See Staples, 511 U.S. at 618–19.

  167. [167]. Id. at 616–18; see also id. at 622 (Ginsburg, J., concurring) (“The Nation’s legislators chose to place under a registration requirement only a very limited class of firearms, those they considered especially dangerous.”).

  168. [168]. Id. at 609–10, 614 (majority opinion).

  169. [169]. Id. at 613–19.

  170. [170]. This type of counterargument was put forth by Justice Stevens in his dissenting opinion:

    In 1934, when Congress originally enacted the statute, it limited the coverage of the 1934 Act to a relatively narrow category of weapons such as submachine-guns and sawed-off shotguns—weapons characteristically used only by professional gangsters like Al Capone, Pretty Boy Floyd, and their henchmen. At the time, the Act would have had little application to guns used by hunters or guns kept at home as protection against unwelcome intruders. Congress therefore could reasonably presume that a person found in possession of an unregistered machinegun or sawed-off shotgun intended to use it for criminal purposes. The statute as a whole, and particularly the decision to criminalize mere possession, reflected a legislative judgment that the likelihood of innocent possession of such an unregistered weapon was remote, and far less significant than the interest in depriving gangsters of their use.

  171. [171]. Id. at 627–30.

  172. [172]. Id. at 617–18 (majority opinion).

  173. [173]. Id.

  174. [174]. See United States v. Freed, 401 U.S. 601, 609–10 (1971).

  175. [175]. Staples, 511 U.S. at 609 (citing Freed, 401 U.S. at 609).

  176. [176]. Crawford v. Washington, 541 U.S. 36, 36 (2004).

  177. [177]. U.S. Const. amend. VI.

  178. [178]. Crawford, 541 U.S. at 45–50.

  179. [179]. See Analytical Approach to Evidence, supra note 73, at 578–80.

  180. [180]. Id.; see also Crawford, 541 U.S. at 50–51 (noting that the Confrontation Clause applies to testimonial statements).

  181. [181]. See Crawford, 541 U.S. at 50–51.

  182. [182]. See id.

  183. [183]. Id. at 51 (“An off-hand, overheard remark might be unreliable evidence and thus a good candidate for exclusion under hearsay rules, but it bears little resemblance to the civil-law abuses the Confrontation Clause targeted.”).

  184. [184]. Id. at 50–56.

  185. [185]. See Alex Stein, Constitutional Evidence Law, 61 Vand. L. Rev. 65, 67 n.6, 72–74 (2008).

  186. [186]. Id.

  187. [187]. Id.

  188. [188]. Id.

  189. [189]. See Mattox v. United States, 156 U.S. 237, 243 (1895).

  190. [190]. See California v. Green, 399 U.S. 149, 164 (1970) (holding that admission of a witness’s prior statement inconsistent with his trial testimony does not violate the Confrontation Clause when the defendant can cross-examine the witness).

  191. [191]. See Ohio v. Roberts, 448 U.S. 56, 65–66 (1980), abrogated byCrawford v. Washington, 541 U.S. 36 (2004).

  192. [192]. See Reynolds v. United States, 98 U.S. 145, 158–59 (1878).

  193. [193]. Crawford, 541 U.S. at 62–65; see also id. at 66 (“The Framers would be astounded to learn that ex parte testimony could be admitted against a criminal defendant because it was elicited by ‘neutral’ government officers.”).

  194. [194]. Id. at 62–66.

  195. [195]. Id. at 51.

  196. [196]. Id.

  197. [197]. Id. at 60–63.

  198. [198]. Id.

  199. [199]. Id. at 68–69.

  200. [200]. Id. at 62.

  201. [201]. Id. at 67–69 (overriding Washington’s “against penal interest” exception to the hearsay rule and vacating defendants’ conviction that relied on unconfronted statement admitted under that exception).

  202. [202]. Id.

  203. [203]. Id.

  204. [204]. See Stephanos Bibas, Essay, Originalism and Formalism in Criminal Procedure: The Triumph of Justice Scalia, the Unlikely Friend of Criminal Defendants?, 94 Geo. L.J. 183, 192 (2005) (“Crawford was a successful blend of originalism and formalism.”); see also, e.g., State v. Hailes, 92 A.3d 544, 559 (Md. Ct. Spec. App. 2014), aff’d, 113 A.3d 608 (Md. 2015) (“[T]he opinion of Justice Scalia for a unanimous Supreme Court in Crawford ... is rife with originalism.”).

  205. [205]. See Richard D. Friedman, Confrontation: The Search for Basic Principles, 86 Geo. L.J. 1011, 1014–22 (1998) (identifying the Framers’ concern about out-of-court affidavits).

  206. [206]. Crawford, 541 U.S. at 71 (Rehnquist, J., concurring).

  207. [207]. See, e.g., Lawrence Rosenthal, Originalism in Practice, 87 Ind. L.J. 1183, 1233–34 (2012) (analyzing Crawford and noting that in order to account for modern police “practice[s], Justice Scalia invoked the now-familiar problem of changed circumstances which ... so often requires originalists to retreat to some form of weak original-expected-applications originalism or semantic originalism”).

  208. [208]. Id. at 1232–35 (identifying this approach as “expected applications” originalism).

  209. [209]. See Stein, supra note 41, at 190–93.

  210. [210]. Id. at 190.

  211. [211]. Id.

  212. [212]. Id. at 191.

  213. [213]. See Friedman, supra note 205, at 1014–22.

  214. [214]. Id.

  215. [215]. See Crawford v. Washington, 541 U.S. 36, 42–51, 50 (2004) (assembling and analyzing historical evidence to find out that “the principal evil at which the Confrontation Clause was directed was the civil-law mode of criminal procedure, and particularly its use of ex parte examinations as evidence against the accused” and that “[t]he Sixth Amendment must be interpreted with this focus in mind”).

  216. [216]. See Friedman, supra note 205, at 1014–22.

  217. [217]. Id. at 1018–20.

  218. [218]. See Bostock v. Clayton Cnty., 140 S. Ct. 1731 (2020). For scholarly analyses of this decision, see Tara Leigh Grove, Comment, Which Textualism?, 134 Harv. L. Rev. 265, 266 (2020) (arguing that the “Bostock [decision] revealed ... tensions within textualism” (emphasis omitted)); William N. Eskridge Jr., Brian G. Slocum & Stefan Th. Gries, The Meaning of Sex: Dynamic Words, Novel Applications, and Original Public Meaning, 119 Mich. L. Rev. 1503, 1512 (2021) (arguing that the Bostock decision missed out an opportunity to introduce linguistic dynamism as an interpretive method that recognizes and incorporates the rights of novel social groups such as gays and lesbians).

  219. [219]. Bostock, 140 S. Ct. at 1737 (“Today, we must decide whether an employer can fire someone simply for being homosexual or transgender. The answer is clear. An employer who fires an individual for being homosexual or transgender fires that person for traits or actions it would not have questioned in members of a different sex. Sex plays a necessary and undisguisable role in the decision, exactly what Title VII forbids.”).

  220. [220]. Id. at 1741 (“So an employer who fires a woman, Hannah, because she is insufficiently feminine and also fires a man, Bob, for being insufficiently masculine may treat men and women as groups more or less equally. But in both cases the employer fires an individual in part because of sex. Instead of avoiding Title VII exposure, this employer doubles it.”).

  221. [221]. Id.

  222. [222]. Id.

  223. [223]. See id. at 1738–42, 1741 (grounding the chosen interpretation of Title VII on “the ordinary public meaning of the statute’s language at the time of the law’s adoption”).

  224. [224]. Id. at 1757 (Alito, J., dissenting) (“The arrogance of this argument is breathtaking.”).

  225. [225]. Id. at 1755–56.

  226. [226]. Id. at 1767 (“In 1964, ordinary Americans reading the text of Title VII would not have dreamed that discrimination because of sex meant discrimination because of sexual orientation, much less gender identity. The ordinary meaning of discrimination because of ‘sex’ was discrimination because of a person’s biological sex, not sexual orientation or gender identity. The possibility that discrimination on either of these grounds might fit within some exotic understanding of sex discrimination would not have crossed their minds.”).

  227. [227]. Id. at 1769 (“For most 21st-century Americans, it is painful to be reminded of the way our society once treated gays and lesbians, but any honest effort to understand what the terms of Title VII were understood to mean when enacted must take into account the societal norms of that time. And the plain truth is that in 1964 homosexuality was thought to be a mental disorder, and homosexual conduct was regarded as morally culpable and worthy of punishment.”).

  228. [228]. Id. at 1770–73, 1776–78.

  229. [229]. Id. at 1783–84 (“The updating desire to which the Court succumbs no doubt arises from humane and generous impulses. Today, many Americans know individuals who are gay, lesbian, or transgender and want them to be treated with the dignity, consideration, and fairness that everyone deserves. But the authority of this Court is limited to saying what the law is.”).

  230. [230]. See Elizabeth F. Emens, Compulsory Sexuality, 66 Stan. L. Rev. 303, 308–09 (2014) (describing the emergence of asexuality as the fourth sexual orientation); see also id. at 365 (indicating that the Title VII antidiscrimination provision that refers to “sex” does not protect asexuals).

  231. [231]. Note that in New York, such practices constitute unlawful discrimination. See id. at 362–66.

  232. [232]. See Bostock, 140 S. Ct. at 1756–58 (Alito, J., dissenting).

  233. [233]. Id. at 1767.

  234. [234]. Id. at 1757 (“As I will show, there is not a shred of evidence that any Member of Congress interpreted the statutory text that way when Title VII was enacted.”).

  235. [235]. Id. at 1773.

  236. [236]. See Jonathan Katz, Gay American History: Lesbians and Gay Men in the U.S.A.:
    A Documentary 276–83 (1976) (observing that homosexuality was well-known in the second half of the twentieth century).

  237. [237]. Price Waterhouse v. Hopkins, 490 U.S. 228, 258 (1989) (“[W]hen a plaintiff in a Title VII case proves that her gender played a motivating part in an employment decision, the defendant may avoid a finding of liability only by proving by a preponderance of the evidence that it would have made the same decision even if it had not taken the plaintiff’s gender into account.”).

  238. [238]. Id. at 233–35.

  239. [239]. Bostock, 140 S. Ct. at 1757 (Alito, J., dissenting).

  240. [240]. For a persuasive argument that supports this observation, see Andrew Koppelman, Why Discrimination Against Lesbians and Gay Men is Sex Discrimination, 69 N.Y.U. L. Rev. 197, 208–19 (1994).

  241. [241]. See Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557, 569–77 (1992) (explaining lawmakers’ preferences of standards over rules when future is unknown).

  242. [242]. See id. at 564.

  243. [243]. See, e.g., U.C.C. § 1-304 (Am. L. Inst. & Unif. L. Comm’n 2001).

  244. [244]. See Neil G. Williams, Offer, Acceptance, and Improper Considerations: A Common-Law Model for the Prohibition of Racial Discrimination in the Contracting Process, 62 Geo. Wash. L. Rev. 183, 186 (1994) (attesting that the “equal treatment rule” codified in “42 U.S.C. § 1981 ... broadly prohibits racial discrimination in the making and enforcement of all contracts” and “bars racial discrimination at all stages of the contracting process” (footnote omitted)). Arguably, the equality protection bestowed by this rule on “all persons” also prohibits discrimination on the basis of gender and sexual orientation.

  245. [245]. See Parchomovsky & Stein, supra note 137, at 172–78 (explaining how legal standards work).

  246. [246]. See Grove, supra note 218, at 278 (mentioning the presence of a similar interpretation in late seventies).

  247. [247]. Frederick Schauer, Playing by the Rules: A Philosophical Examination of Rule-Based Decision-Making in Law and in Life 36 (1991).

  248. [248]. Id. at 35–37.

  249. [249]. See Price Waterhouse v. Hopkins, 490 U.S. 228, 239 (1989) (“In passing Title VII, Congress made the simple but momentous announcement that sex ... [is] not relevant to the selection, evaluation, or compensation of employees.”).

  250. [250]. Id. at 250–52 (holding that Title VII’s words “because of sex” outlaw all kinds of sex-based stereotyping in employment decisions and that it is unlawful for an employer to penalize a female employee for behaving non-femininely—e.g., “on the basis of a belief that a woman cannot be aggressive”); see also id. at 244–45 (“[O]nce a plaintiff in a Title VII case shows that gender played a motivating part in an employment decision, the defendant may avoid a finding of liability only by proving that it would have made the same decision even if it had not allowed gender to play such a role. This balance of burdens is the direct result of Title VII’s balance of rights.” (footnote omitted)).

  251. [251]. See Katz, supra note 236, at 276–83.

  252. [252]. See Manning & Stephenson, supra note 7, at 24 (stating the principle of legislative supremacy and linking textualism, intentionalism, and purposivism to that principle).

  253. [253]. See, e.g., Brian A. Lichter & David P. Baltmanis, Foreword: Original Ideas on Originalism, 103 Nw. U.L. Rev. 491, 491 (2009) (“The impact of originalism is significant, and originalism now represents a dominant—perhaps the dominant—method of constitutional interpretation.”). See generally Randy E. Barnett, The Gravitational Force of Originalism, 82 Fordham L. Rev. 411 (2013) (arguing that originalism exerts pervasive “gravitational force” on legal doctrine); Grove, supra note 218, at 265 (“[T]extualism has in recent decades gained considerable prominence within the federal judiciary ... .”).

  254. [254]. See Krishnakumar, supra note 7, at 1288–304 (showing that the U.S. Supreme Court decisions implicitly invoke purposivist interpretation).

  255. [255]. See generally Frederick Schauer, Giving Reasons, 47 Stan. L. Rev. 633 (1995) (articulating common-law judges’ commitment to principle).

  256. [256]. See, e.g., Larry Laudan, Progress and Its Problems: Towards a Theory of Scientific Growth 1–8 (1977) (noting the pervasiveness of scientific progress and arguing for its development—rather than a rationality-focused explanation).

  257. [257]. See generally L. Jonathan Cohen, An Introduction to the Philosophy of Induction and Probability (1989) (identifying and analyzing different conceptions of probability, mathematical and non-mathematical).

  258. [258]. See Ian Hacking, The Taming of Chance 95–104 (1990) (linking human progress to probability).

  259. [259]. Id.

  260. [260]. See Baude & Sachs, supra note 65, at 1135–36 (analogizing accepted methods of interpretation to precedent).

  261. [261]. These and other factors are part of the judiciary’s “passive virtues” that make it “the least dangerous branch” of government. See generally Alexander M. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics (1962).

  262. [262]. See Andrei Marmor, Social Conventions: From Language to Law 106–30 (2009) (analyzing the role of social conventions in language and speech acts).

  263. [263]. See Lawrence B. Solum, On the Indeterminacy Crisis: Critiquing Critical Dogma, 54 U. Chi. L. Rev. 462, 479–80 (1987) (“Kripke suggests that Wittgenstein’s skeptical paradox may be resolved using the notion of ‘community’: our actual community does agree on how many rules are applied—including, for example, the red stoplight rule. Anyone who claims to have mastered a rule will be judged by the community to have done so if his particular responses agree with those of the community in enough cases. Those who deviate are corrected. One who is incorrigibly deviant with respect to enough rules cannot participate in community life. There is certainly nothing remarkable about this solution to the skeptical paradox in the context of legal rules; community agreement about at least some applications of legal rules is, at bottom, the explanation for how they can be consistently applied.” (footnotes omitted)).

  264. [264]. Id. at 483–84; see also id. at 479 (“[R]ule-skepticism can be shown to be toothless for the same reason that this sort of epistemological skepticism is toothless: worrying about being a brain in a vat will not have any effect on what you do. Likewise, worrying about rule-skepticism will not have any effect on the way cases are decided. The skeptical possibilities invoked by both rule-skepticism and epistemological skepticism are not practical possibilities, and only practical possibilities affect the way one acts.”).

  265. [265]. See supra notes 176–83 and accompanying text.

  266. [266]. See Brief for Respondent at 11–12, Crawford v. Washington, 541 U.S. 36 (2004) (No. 02-9410) (“During the ensuing 23 years since the Roberts decision, the trend of this Court has been to adhere to and refine this framework. Petitioner and Amici, however, would have this Court discard both this past 23 years of successful application of the Roberts framework
    ... .” (footnote omitted)); G. Ross Anderson Jr., Returning to Confrontation Clause Sanity: The Supreme Court (Finally) Retreats from Melendez-Diaz and Bullcoming, 60 Fed. Law. 67, 67, 71 n.98 (2013) .

  267. [267]. See What Divides Textualists from Purposivists?, supra note 7, at 96–99 (describing purposivist methods of interpretation that adhere to legislative supremacy); Richard A. Posner, The Problems of Jurisprudence 270 (1990) (“American judges are frequently in the position of [a] platoon commander ... . The legislature’s commands are unclear and the judges cannot ask the legislators for clarification. In that situation the judges should not think of themselves as failed archeologists or antiquarians. They are part of a living enterprise—the enterprise of governing the United States—and when the orders of their superiors are unclear this does not absolve them from responsibility for helping to make the enterprise succeed. The responsible platoon commander will ask himself what his captain would have wanted him to do if communications should fail, and similarly judges should ask themselves, when the message imparted by a statute is unclear, what the legislature would have wanted them to do in such a case of failed communication.” (footnote omitted)).

  268. [268]. The concept of “strong purposivism,” identified here as fact-free, was coined by John F. Manning in Textualism and the Equity of the Statute, 101 Colum. L. Rev. 1, 10 (2001).

  269. [269]. See id. at 11 (observing that “[t]he distinguishing feature of strong purposivism is that when a specific statutory text produces ‘an unreasonable [result] “plainly at variance with the policy of the legislation as a whole,”‘ federal judges may (and must) alter even the clearest statutory text to serve the statute’s ‘purpose’” (second alteration in original)).

  270. [270]. See Posner, supra note 267, at 276–80; Posner, supra note 9, at 1664 (“In approaching an issue that has been posed as one of statutory ‘interpretation,’ pragmatists will ask which of the possible resolutions has the best consequences, all things (that lawyers are or should be interested in) considered, including the importance of preserving language as a medium of effective communication and of preserving the separation of powers. Except as may be implied by the last clause, pragmatists are not interested in the authenticity of a suggested interpretation as an expression of the intent of legislators or of the framers of constitutions. They are interested in using the legislative or constitutional text as a resource in the fashioning of a pragmatically attractive result.”).

  271. [271]. See Cohen, supra note 257, at 47–60, 121–23 (explaining how different forms of probability attach to the universe of factually uncertain events).

  272. [272]. See, e.g., Aziz Z. Huq, A Right to a Human Decision, 106 Va. L. Rev. 611, 645 n.167 (2020) (attesting “that ... thoroughgoing [epistemic] skepticism is self-refuting” and citing sources).

  273. [273]. See supra note 16 and sources cites.

  274. [274]. See Joseph William Singer, The Legal Rights Debate in Analytical Jurisprudence from Bentham to Hohfeld, 1982 Wis. L. Rev. 975, 980 (“The contradiction between freedom ... and security
    ... translates into the contradiction between individual rights and state powers. We must determine the extent to which individual freedom of action may legitimately be limited by collective coercion over the individual in the name of security.”); Betty Mensch, Freedom of Contract as Ideology, 33 Stan. L. Rev. 753, 761 (1981) (reviewing P.S. Atiyah, The Rise and Fall of Freedom of Contract (1979)) (“Despite the mystical moment of formation posited by the classical model [of contract], when the offeror’s and the acceptor’s free wills meet in full accord, those decisions are not decisions about freedom, but between the claim for freedom and the need for security.”).

  275. [275]. See Joseph Raz, The Morality of Freedom 181 (1988) (defending the vision of law as a unifying accommodation of different moral outlooks held by individuals willing to live together in a diverse, but orderly, society).


Justice of the Israel Supreme Court.

I am grateful to Daphne Barak-Erez, David Cheshin, Gideon Parchomovsky, Rinat Sangero-Kitai, participants in the seminar on judging at the Hebrew University of Jerusalem School of Law, and colleagues who attended my lecture on statutory interpretation at the Center for Judicial Training and Education for helpful input. A point of disclosure: I actually practice probabilism. See, e.g., PCA 10011/17 Meytal Engineering & Services Co. v. Salman (Isr. Sup. Ct. 19.8.2019) (Hebrew) (Stein, J., concurring) (using legislative background, drafters’ intent, choices of words in parallel statutes, and other relevant evidence to determine the most probable meaning of the word “product” in a statute prohibiting discrimination in sales of products and whether it includes real estate as well); PAA 2283/18 Zoning and Planning Comm. of Tel Aviv v. Yad Charuzim Properties Ltd. (Isr. Sup. Ct. 1.1.2019) (Hebrew) (Stein, J., concurring) (accounting for legislative background, statutory language, history, and purpose to determine the most probable meaning of the word “appreciation” in a statute imposing land appreciation taxes upon property owners realizing building rights).