Printer Friendly


Courts look at text differently in high-stakes cases. Statutory language that would otherwise be "unambiguous" suddenly becomes "less than clear." This, in turn, frees up courts to sidestep constitutional conflicts, avoid dramatic policy changes, and, more generally, get around undesirable outcomes. The standard account of this behavior is that courts' failure to recognize "clear" or "unambiguous" meanings in such cases is motivated or disingenuous, and, at best, justified on instrumentalist grounds.

This Article challenges that account. It argues instead that, as a purely epistemic matter, it is more difficult to "know" what a text means--and, hence, more difficult to regard that text as "clear" or "unambiguous"--when the practical stakes are raised. For that reason, this Article insists, it is entirely rational for courts to be more cautious when interpreting text in high-stakes cases than they would be if the stakes were low. Drawing on contemporary work in philosophy of language and epistemology, this Article grounds its argument in the observation that ordinary speakers' willingness to attribute "knowledge" or "clarity" decreases as the practical stakes increase. And while the technical explanations of this phenomenon vary, they all reflect a basic insight: that one needs greater epistemic justification to act on some premise the higher the practical stakes.

To illustrate, this Article applies the above insight to various interpretive settings. Considering judicial review, for example, this Article explains that it makes good epistemic sense for a court to wait until it is really sure that a statute means what it thinks it means before taking the extraordinary step of invalidating that statute as unconstitutional. Similarly, this Article urges that it is just sound epistemic practice for a court is to construe a statute in a way that would unsettle an existing implementation regime only if it is especially well justified in its reading of the statutory text--that is, only if it really knows that its reading is correct.

This Article thus offers at least a partial justification of courts' seemingly loose treatment of statutory text when the practical stakes are raised. And it does so, in contrast to prior scholarly efforts, by appeal to reasons that both formalists and instrumentalists can accept.
Table of Contents

    I. Prior Explanations
       A. cynicism
       B. The New Holy Trinity
       C. Constructed Constraint
   II. High/Low Stakes
       A. Motivating Examples
       B. Technical Explanations
       C. Basic Insight
  III. Applications
       A. Constitutional Avoidance
          1. Bond v. United States
          2. Northwest Austin
       B. Nonconstitutional Challenges
          1. King v. Burwell
          2. Inclusive Communities
       C. Rule of Lenity
       D. Chevron
   IV. Coda: On Constitutional Interpretation .

We're all textualists now[, except in June].

--Justice Elena Kagan (1)


"We must enforce plain and unambiguous statutory language according to its terms." (2) Courts recite such maxims again and again. (3) And, in run- of-the-mill cases, they pretty much do as they say. As John Manning has observed, gone are the days when courts would openly rewrite statutory language in the service of Congress's apparent policy aims. (4) More still, courts (somewhat) reliably give effect to "plain" or "clear" language, ostensible (or perhaps conceivable) practical downsides notwithstanding. (5) All of this suggests a new consensus that courts should prioritize Congress's specific instructions over its general policy ambitions--the reason being that those instructions are the best indication of "Congress's specific choices about the means to carry [its policy] ends into effect." (6) Courts thus agree that a statute's precise contribution to the law is (at a minimum) what Congress communicates through that statute precisely--at least, that is, where what Congress communicates is "clear." (7)

The above story does reasonably well with ordinary cases. More worrisome is how it seems to fare when the practical stakes are raised. As different scholars have noted, courts treat statutory text as more malleable in big cases. When considering constitutional challenges, for example, courts frequently bend over backwards to avoid reading statutes in ways that would raise "serious constitutional doubts." (8) The result is the adoption of what Neal Katyal and Thomas Schmidt disparage as "tortured constructions of statutes ... bearing] little resemblance to laws actually passed." (9) So too in cases involving nonconstitutional challenges to major statutes, (10) where courts--and, in particular, Chief Justice Roberts--are routinely criticized for "ignor[ing]" statutory text outright in an effort to uphold existing implementation regimes. (11)

So what to make of the disparity? Is it just that courts stick to the text in low-stakes cases but are textually unbound when it matters? (12) Or, only slightly more charitably, is it that courts care about text only so much, and that, at some point, practical or institutional interests simply outweigh? Something like this cynical (or semicynical) explanation is familiar, especially as the Supreme Court issues its late-Term decisions. Hence, Adrian Vermeule's remark: "We have two Supreme Courts--roughly, constrained legalism October through May, and then a free-for-all." (13)

More recently, a handful of scholars have offered limited justifications of the disparity, ultimately on instrumentalist grounds. Richard Re, for instance, has suggested that judges consistently adhere to "clear" text, but that, for some, "purposive and pragmatic considerations" partially determine just how clear a text needs to be to command respect. (14) Re's explanation is that "when a statute's central objective is at risk or an otherwise plausible reading leads to alarming results," it only makes sense to "hold the text to a higher- than-normal standard." (15) Somewhat differently, Curtis Bradley and Neil Siegel have argued in the constitutional context that whether a text is perceived as "clear" or "ambiguous" depends in part on historical practice. (16) According to Bradley and Siegel, even if a text is "clear" at the time of enactment, subsequent activity to the contrary (e.g., a "[l]ong-settled and established practice" of congressional acquiescence) (17) can actually render that text "ambiguous," thereby freeing courts from textual constraint. (18) In support of this striking claim, Bradley and Siegel cite practical and institutional interests, claiming, for example, that crediting historical practice shows respect for coordinate branches and helps keep old texts up to date. (19)

Both of the justifications of the high-stakes--low-stakes disparity just mentioned are limited in that each only somewhat maps onto the high-stakes--low- stakes distinction. Re's account, for example, predicts that courts will treat text more loosely if either pragmatic or purposive reasons (20) cut against the otherwise "clear" meaning of the text. (21) Therefore, that account would have it that, even in low-stakes cases, the presence of purposive reasons will result in more casual reading--a prediction that runs contrary to the pattern of judicial behavior observed at the outset. (22) Bradley and Siegel's account, by contrast, applies principally, in high-stakes cases: the possible invalidation of a long-established practice will, after all, typically render a case high stakes. (23) At the same time, the universe of high-stakes cases is plainly much larger than that of cases involving challenges to long-established practice (e.g., major cases involving recently enacted statutes or regulations), making that account incomplete. (24)

More fundamentally, though, both Re's and Bradley and Siegel's accounts require that one accept the sort of instrumentalist reasoning that most proponents of careful statutory reading reject. Re's suggestion, for instance, that some texts be held to a "higher-than-normal standard" sets off alarm bells for those for whom the role of a court when interpreting a statute is to determine what Congress meant by the words that it used. (25) Similarly, Bradley and Siegel's suggestion that texts need to be kept up to date seems to run contrary to what Larry Solum calls the "fixation thesis," the claim that the meaning of a text is fixed at the time of enactment, a basic assumption of most any version of textualism. (26)

As an alternative, this Article contends that one can plausibly make sense of how courts handle text in high-stakes cases by appeal to epistemological considerations cognizable by formalists and instrumentalists alike. The argument is as follows: to say that the meaning of a statute is "clear" or "plain" is, in effect, to say (27) that one knows what that statute means. (28) As numerous philosophers have observed, however, ordinary speakers attribute "knowledge"--and, in turn, "clarity"--more freely or less freely depending upon the practical stakes. (29) In low-stakes situations, speakers are willing to concede that a person "knows" this or that given only a moderate level of justification. Suppose, for example, Jane has checked the train schedule, uses the train system with some regularity, and is in no particular rush; in that situation, it is plausible for Jane to say that she "knows" that the train will arrive at 7:00 a.m. as scheduled. By contrast, if the practical stakes are high, speakers require greater justification before allowing that someone "knows" that same thing, holding constant that person's evidence: if, say, Jane has the same evidence as above but absolutely cannot afford to be late, Jane's claim to "know" that the train will arrive at 7:00 a.m. is more doubtful. (30)

As this Article explains, philosophers differ in their technical explanations of the above phenomenon: some attribute it to the semantic connection between "knowledge" and action, (31) others suggest that we mean different things by "know" in different practical contexts, (32) and others still contend that it has to do not with semantics but with pragmatics--very roughly, what we imply, as opposed to what we say, when attributing "knowledge." (33) Technical disagreements notwithstanding, most agree that this pattern of linguistic behavior reflects a basic insight concerning the relationship between epistemological and practical reason--namely, that one needs greater epistemic justification to act on some premise the higher the practical stakes. (34)

Applying this basic insight, this Article urges that courts' seemingly loose treatment of statutory text in high-stakes cases is partially attributable to (or at least justified by) the heightened epistemological standards that apply in high-stakes settings. Put more colloquially, because it is more difficult to "know" what statutes mean in high-stakes cases, it makes perfect sense that courts find "clear" or "plain" meaning less often. And, as a result, courts will more often have license to resort to "gap-filling" (i.e., nonlinguistic) measures in those cases. (35) Consider cases involving constitutional challenges. As Justice Brandeis observed, "The Court has frequently called attention to the 'great gravity and delicacy' of its function in passing upon the validity of an act of Congress...." (36) Thus, it comes as no surprise that courts require a great deal of epistemological justification before acting on the premise that a statute means X where reading the statute to mean X would raise serious constitutional concerns. Because of the gravity of acting on that premise, it is, for the reasons articulated above, more difficult for courts to "know" that the statute means X in the context of an adjudication. This, in turn, makes it more difficult for courts to regard X as the statute's "clear" or "plain" meaning. The range of "fairly possible" readings for that statute thus proves greater than it would absent the looming constitutional concern, in the sense the statute turns out, in that context, to admit of readings other than X. (37) Readings that would otherwise be reasonably regarded as "tortured" (38) thus become epistemologically available owing to the heightened practical stakes.

To be clear, to say that it is more difficult to "know" what a statute means in a high-stakes case is not to say that to do so is impossible. To illustrate, this Article contrasts two relatively recent Supreme Court decisions: Bond v. United States (39) and Northwest Austin Municipal Utility District No. One v. Holder. (40) In Bond, this Article argues, the Supreme Court relied explicitly upon the epistemological principle at issue here, (41) but did so in the service of a plainly implausible reading--plainly implausible even considering the heightened practical stakes of the case. In Northwest Austin, by contrast, the Court's reliance on that principle was implicit, but its prima facie strained reading of the statute at issue was one arguably made plausible by the raised stakes. (42) In comparing these two cases, this Article attempts to show that, even in high-stakes situations, courts remain at least somewhat textually constrained. Further, the contrast is intended as a concession that some but not all of the observed disparity between high- and low-stakes cases might be justified on epistemological grounds.

In terms of which cases count as "high stakes," this Article takes no position except to say that a case is high stakes just in case it matters a great deal to the deciding court and those to whom its opinion is addressed--that is, a "high-stakes" case matters more to the pertinent individuals than does the typical case. (43) For that reason, the arguments below depend in part on apparent subjective evaluation: only if it seems likely that the deciding court thinks that a case matters specially should one expect that court to proceed with the corresponding epistemic caution. On the other hand, this Article leaves open the possibility of objective critique: even if courts do (or don't) regard some class of cases as high stakes, one can still argue they should (or shouldn't).

This Article has four Parts. Part I considers prior explanations of the disparity between high-stakes and low-stakes interpretation. Part II offers an alternative, epistemological explanation of that phenomenon, building upon the basic insight that the degree of epistemological justification required to act on a premise increases as do the practical stakes. Part III applies that basic insight to uncontroversially high-stakes adjudicatory situations. It also considers situations one might think should qualify as high stakes--in particular, criminal adjudication--but that courts appear to regard as low stakes, given the ease with which they identify "plain" or "unambiguous" meaning. In so doing, this Part shows that the insight this Article leverages has critical as well as justificatory potential. Last, Part IV considers the applicability of the Article's thesis to constitutional interpretation. As that Part observes, the inherently high-stakes nature of constitutional interpretation might seem to explain why courts treat constitutional text much more loosely than statutory text. Be that as it may, the Part concedes that courts treat constitutional text so loosely in certain instances that it becomes fair to ask whether courts are engaged in constitutional interpretation at all. (44)


This Part discusses prior accounts of courts' disparate treatment of text in high- and low-stakes cases. Section I.A sets out the standard account of the disparity, according to which courts' reluctance to identify "clear" or "unambiguous" meanings is either disingenuous or the product of motivated reasoning. Section I.B examines a recent, uncynical account that attributes the disparity to the rise of a new form of purposivism. Section I.C borrows from constitutional law, considering the "historical gloss" approach to interpretation, pursuant to which postenactment practice can render a previously "clear" text "unclear" (or vice versa). As this Part argues, all of these accounts are limited in that each promises to justify at most a subset of high-stakes decisions. More still, the justifying reasons each account offers are ones that formalists--the strongest proponents of adherence to "clear" or "unambiguous" text--would reject.

A. Cynicism

According to the standard, cynical account, courts are reluctant to identify "clear" or "unambiguous" meanings in high-stakes cases because the practical concerns raised in those cases overwhelm any commitment to textual fidelity. In its starkest form, the standard account has it that courts outright "ignore" statutory text in an effort to advance some political or institutional agenda. (45) Or, in a slightly milder form, the suggestion is that such agendas skew courts' perception of statutory text in high-stakes cases, causing them to fail to perceive "clear" or "unambiguous" meanings apparent to the unmotivated reader. To the extent that the standard account is justificatory, it is so in virtue of the principle that "the [practical] ends justify the [interpretive] means." With, for example, the Paris Climate Agreement at stake, one might (might!) forgive courts for attending to text less carefully than normal. (46)

Proponents of the standard account include Neal Katyal and Thomas Schmidt, who, in a recent article, criticize the Roberts Court sharply for its treatment of statutory text in cases involving constitutional challenges. (47) As explained above, the canon of constitutional avoidance in its modern form permits courts to reject the "most natural reading" of a statute (48) if that reading would "raise serious constitutional questions." (49) A court may, however, adopt a less natural but also less constitutionally doubtful reading only if that alternate reading is "fairly possible." (50) If, by contrast, the meaning of the challenged statute is "clear," a court must accept it and address any constitutional questions directly. (51)

Katyal and Schmidt argue that, through application of the avoidance canon, the Roberts Court has engaged in an aggressive campaign of judicial "rewriting" of statutes in constitutional cases, "usher[ing] in legal change" under the banner of judicial restraint. (52) According to Katyal and Schmidt, the Roberts Court has appeared "indifferent" to whether the statutory readings it adopts in constitutional cases are "at all plausible." (53) Instead, they continue, the Court has freely endorsed interpretations that are otherwise "unthinkable!,] ... abandoning] normal principles of statutory interpretation whenever a serious constitutional issue looms." (54) The result is that the Court "leaves in place ... law[s] that Congress never passed and may never have wanted to pass," (55) a problem made all the worse by the reality of partisan gridlock and, hence, the implausibility of legislative override. (56) Katyal and Schmidt allow that the Court's motivations in these cases may be innocent, driven by, for example, a "desire to have narrower rulings." (57) Be that as it may, Katyal and Schmidt insist that sometimes "distorting a statute in the name of avoidance does more violence to congressional intent--and is therefore more countermajoritarian--than outright invalidation." (58)

Sounding a more optimistic note, Jonathan Adler characterizes the Roberts Court's (59) apparent loose treatment of text as an exercise of "Burkean minimalism." (60) Like Katyal and Schmidt, Adler describes the Court as ignoring the "plain meaning" of statutory text in cases involving constitutional challenges, as well as cases involving nonconstitutional challenges to major statutes. (61) According to Adler, though, the Court's "willing[ness] to stretch or massage relevant statutory provisions" in such cases reflects a desire to "avoid interpretations that would require invalidating federal statutes on constitutional grounds or would otherwise prove disruptive to the status quo." (62) In Adler's view, the Court seems committed in these cases not only to respecting and deferring to the political branches--the traditional understanding of judicial minimalism--but also to "reducing the practical impact of [its] rulings." (63) Drawing on Chief Justice Roberts's famous analogy between judges and umpires, Adler suggests that, in the chief justice's view, a good judge, like a good umpire, is one who "avoid[s] making calls that control the outcome of the game." (64) And while Adler is careful not to endorse this understanding of judging, (65) he does offer a prima facie justification for the Court's "willing[ness] ... to stretch statutory text," namely the "avoid[ance of] disruptive consequences." (66)

Needless to say, the justification that Adler highlights is both highly contestable and of limited appeal. As Adler himself hints, "strain[ing]" to read text in ways that promote desirable outcomes--status quo preserving or no--threatens institutional legitimacy. (67) In addition, as Katyal and Schmidt observe, efforts to minimize judicial impact can have unintended practical consequences. (68) More fundamentally, whatever appeal this sort of practical justification has to instrumentalists, formalists reject it out of hand. (69) As mentioned above, any justification resulting from the standard, cynical (or semicynical) account must be of the "ends-justify-the-means" variety. And that sort of justification is, for the formalist, per se uncognizable.

B. The New Holy Trinity

The cynical account has it that courts abandon interpretive principle in high-stakes cases. (70) Richard Re argues, by contrast, that interpretive principle has "evolved" in ways that helps explain the high-stakes-low-stakes disparity. (71) According to Re, although courts no longer "rewrite" statutory language openly in the service of Congress's apparent policy aims, the Roberts Court in particular has come to assign significant weight to such aims--along with other pragmatic considerations--when determining "how much clarity is required for a text to be clear" and so to command a particular outcome. (72) The result, Re continues, is a new form of purposivism, pursuant to which text constrains, but the degree to which it constrains depends upon nontextual factors. (73) As Re explains it, "If a reading has no textual support, then no amount of pragmatism or purpose can carry the day." (74) If, on the other hand, "a statute's central objective is at risk or an otherwise plausible reading leads to alarming results," (75) then the approach requires overwhelming textual evidence for text to control--assuming, that is, some minimally textually plausible alternative. (76) Re dubs this approach "The New Holy Trinity," (77) contrasting it with the old, open-rewriting brand of purposivism associated with the now-infamous Holy Trinity Church v. United States. (78)

Much like Adler, Re suggests that loose treatment of text in "unusual but pivotal" cases helps to avoid "alarming results." (79) Unlike Adler, however, Re does his best to translate that justification into noninstrumentalist terms. Re hypothesizes that, by holding text to a higher standard when a surprising or harmful result looms, courts are attempting to "adhere to clear text when it's the product of deliberate compromise" but to set it aside "when it springs from an inattentive mistake." (80) So articulated, Re's defense of the New Holy Trinity mirrors the standard defense of the old doctrine that "judges may deviate from even the clearest statutory text when a given application would otherwise produce 'absurd' results" (81)--a doctrine associated, fittingly, with the old Holy Trinity. (82) As Manning explains, the basic thought underlying the absurdity doctrine is that "legislators necessarily draff statutes within the constraints of bounded foresight, limited resources, and imperfect language." (83) For that reason, if a given statutory application is absurd, a court should "presume [ ] that this absurd result reflects imprecise drafting that Congress could and would have corrected had the issue come up during the enactment process." (84)

Because Re's defense of the New Holy Trinity is familiar, so too is the basic objection. It is, at this point, widely recognized that Congress legislates means as well as ends. Enacting legislation (when it happens) requires compromise, and implementing compromise often requires adopting otherwise suboptimal means. (85) As textualists have long argued, the most feasible way for Congress to identify specific means is for it to use specific words. (86) For that reason, if courts treat precise statutory language as a mere "proxy" of Congress's general policy aims, they make it infeasible for members of Congress to negotiate compromises with binding force. (87) This is why there is now a consensus that courts must enforce "clear" statutory text. (88) And, as Manning has argued, the reasons that support that consensus do not support an exception for cases in which enforcing "clear" text would entail an "absurd" practical outcome. (89) It is, of course, entirely sensible for courts to consider the "absurdity" of practical outcome as evidence against a particular reading of some text. (90) But once a court, having taken that outcome into account, deems the corresponding reading "best" nonetheless, that court would exceed its authority by setting that reading aside, dismissing it as a mistake. (91)

Unlike the old Holy Trinity, Re insists that courts may adopt an unnatural reading only if that reading has "non-frivolous textual support." (92) Be that as it may, what Re advocates is that courts reject a statute's most natural reading on the grounds that Congress's choice of language is more likely owing to "inattent[ion]" than to compromise. (93) Re appears, in so doing, to suggest that courts treat precise statutory language as a mere "proxy" for purpose--so long, that is, as that language is anything less than crystal clear. If that's right, the New Holy Trinity threatens to undermine legislative bargains in just the same way as the old. Although it is more limited in scope-leaving untouched textually unassailable readings--the kind of damage caused is the same.

Another way of putting the objection is that the New Holy Trinity double counts legislative purpose. Again, pursuant to the New Holy Trinity, courts consider Congress's apparent policy aims when deciding "how much clarity is required" in order for text to control--if a text's most natural reading runs contrary to purpose, that reading must be especially clear for courts to give it effect. (94) The problem with tying "clarity" to purpose, however, is that contemporary courts already consider Congress's apparent policy aims when determining a text's most natural reading. As Manning observes, in the bad old days, textualist judges were "literalists," equating fidelity to text with enforcement of "ordinary" meaning. (95) Over time, though, textualists came to recognize that language has meaning only in context and so began to consider the practical setting when making sense of a particular text. (96) In a slogan, textualists now accept that, to understand "what Congress is trying to say," courts must have some grasp of "what Congress is trying to do." (97) But if courts now consider legislative purpose when figuring out how a statute is most naturally read, the question for Re becomes: Why consider purpose again when deciding how clear that most natural reading must be to control?

A second objection, already mentioned, is that Re's account predicts loose treatment of text even in cases with low practical stakes. (98) Again, according to Re, courts treat text more loosely when purposive considerations cut against. (99) This, in turn, suggests that courts will read text "creatively" even in low-stakes cases so long as Congress's apparent purpose runs contrary to a statute's most straightforward reading. As indicated above, that prediction appears inconsistent with judicial behavior in recent years. (100)

Both of the objections above pertain to what one might call the "purposive" branch of the New Holy Trinity--that is, the claim that, where a text's most natural reading runs contrary to Congress's apparent policy aims, a heightened standard of clarity goes into effect. This leaves the proposal's "pragmatic" branch, the claim that potentially dramatic practical consequences raise the bar for what counts as "clear." For reasons articulated below, there is something fundamentally correct about the pragmatic branch of the New Holy Trinity, even if it is not fully theorized--the justification for such a doctrine has not to do with risk of congressional mistake, but rather judicial mistake. (101) If one were to trim the proposal of its purposive branch, one could thus conceive of the New Holy Trinity less as an alternative to the account offered here than as an early predecessor to it.

C. Constructed Constraint

In constitutional law, a popular claim as of late is that postenactment practice can render constitutional text clearer or--more controversially--less clear. (102) Sometimes termed the "historical gloss" approach to interpretation, a nod to Justice Frankfurter's famous opinion in Youngstown Sheet & Tube Co. v. Sawyer, (103) the motivating thought is that the "perceived clarity or ambiguity" of some text is attributable not just to considerations having to do with linguistic meaning (e.g., ordinary usage, apparent purpose), but also to nonlinguistic considerations such as congressional acquiescence to executive action or steady judicial enforcement. (104) So conceived, "textual clarity is not just some linguistic fact of the matter that exists apart from the overall process of constitutional implementation." (105) Rather, "the clarity and ambiguity of the constitutional text is," as Bradley and Siegel put it, "partially constructed" by subsequent practice. (106)

As one example, Bradley and Siegel cite the Supreme Court's recent decision in NLRB v. Noel Canning. (107) In that case, one question before the Court was whether the Recess Appointments Clause, which authorizes the president to Fill any vacancies that "may happen during" a Senate recess, encompasses vacancies that come into existence prior to the recess at issue. (108) As Justice Breyer, writing for the majority, conceded, the "most natural" reading of that language limits the recess appointment power to vacancies that come into existence during the relevant recess. (109) At the same time, Justice Breyer continued, the Clause's language at least "permits" a "broader interpretation" according to which vacancies that arise before but persist into the recess are included within the president's power (110)--an interpretation, Justice Breyer went on to argue, more consonant with the Clause's "purpose." (111) As evidence of the Clause's "ambigu[ity]," Justice Breyer appealed to the longstanding executive branch practice of reading that language expansively. (112) Concurring in the judgment, Justice Scalia contested the linguistic availability of the majority's interpretation, maintaining that "no reasonable reader" would have understood the Clause's language as the majority suggested. (113) According to Justice Scalia, to use those words to achieve that end would have been "surpassingly odd," in particular given the "read[y] availability]" of "alternative phrasings" that would have "convey[ed] that meaning clearly." (114) As to executive branch practice, Justice Scalia went on to observe that appointments for vacancies coming into existence before the recess at issue became common only around the mid-nineteenth century, and only after a great deal of contestation within the executive branch. (115)

Characterizing the dispute between Justice Breyer and Justice Scalia, Bradley and Siegel observe that both accept that "historical practice might be relevant" to constitutional interpretation, but only if "the constitutional text is ambiguous." (116) For Justice Scalia, the text of the Recess Appointments Clause was "clear," making postenactment practice largely irrelevant. (117) For Justice Breyer, by contrast, the text of the Clause was "ambiguous," thus giving historical practice "significant weight." (118) According to Bradley and Siegel, this difference in perception of textual clarity is at least partially attributable to a difference in which considerations each justice thought relevant to the question of whether text is "clear." For Justice Scalia, the clarity of the text was shown mostly if not entirely by appeal to linguistic considerations such as ordinary usage (119)--characteristic, as Bradley and Siegel observe, of originalist approaches to constitutional interpretation. (120) For Justice Breyer, on the other hand, the reason--indeed, the "only reason"--for "s[eeking] out a possible reading" beyond the one deemed "most natural" was, seemingly, the longstanding historical practice of acting contrary to that "most natural" reading. (121) Thus, the language of the Recess Appointments Clause was actually made ambiguous, claim Bradley and Siegel, by that historical practice. (122)

However it fares as an approach to constitutional interpretation, (123) "historical gloss" has ready appeal when it comes to statutes. If, for example, Adler is right that the Court has a bias in favor of the status quo, (124) one way to make sense of that "bias" is as an application of "historical gloss." Applying that framework, one could say that the Court perceives otherwise "clear" text as "ambiguous" in the cases Adler cites precisely because the "most natural reading" of the text in those cases would, if accepted, be disruptive of the existing implementation regime. So understood, it is not that the Court pretends to see "ambiguity" to avoid a disruptive result. Rather, it is that the specter of disruption "constructs" unclarity where none was perceived before. (125)

Despite its appeal, "historical gloss" also has serious limitations in terms of explaining the high-stakes-low-stakes disparity in statutory interpretation. (126) First and most obvious, even if some high-stakes cases are plausibly situated within the "historical gloss" framework, others are plainly not. As discussed below, cases involving nonconstitutional challenges to what can only be characterized as "major" statutes are uncontroversially "high stakes." (127) Some of those cases involve challenges to longstanding implementation regimes and thus fit squarely within the "historical gloss" approach. (128) Other such cases, however, such as King v. Burwell, (129) involve what are more or less preemptive strikes. (130) And while a nascent implementation regime is plausibly a "gloss" on the corresponding statutory text, a historical gloss it is not.

Second, much like Re's account, the "historical gloss" approach would seem to predict loose treatment of text if a text's "most natural" reading is contrary to settled postenactment practice, regardless of the practical stakes. Again, that prediction appears incorrect. In Milner v. Navy, (131) for instance, the Supreme Court considered whether the Navy could invoke a Freedom of Information Act (FOIA) exemption for an agency's "personnel rules and practices" to withhold internal maps pertaining to the storage of munitions. (132) Below, the Ninth Circuit held that the Navy could, reasoning that the maps in question related to "predominantly internal" matters the disclosure of which "presents a risk of circumvention of agency regulation." (133) In so holding, the Ninth Circuit applied a test articulated almost three decades earlier by the D.C. Circuit (134)--a test, according to Justice Breyer, "consistently followed, or favorably cited, by every Court of Appeals to have considered the matter during the past (30) years." (135) The Supreme Court reversed, holding that "the plain meaning of the term 'personnel rules and practices,' encompasses only records relating to issues of employee relations and human resources," and so excludes maps of explosives. (136) Rejecting the more expansive reading of the relevant exemption, the Court insisted that to accept the D.C. Circuit's reading would require it to "flout all usual rules of statutory interpretation." (137) The Court continued that it had "no warrant to ignore clear statutory language on the ground that other courts have done so." (138) The Court concluded by observing that "the Government has other tools at hand to shield national security information and other sensitive materials"--in particular, other exemptions within FOIA. (139) For that reason, the Court's decision was likely of limited practical significance despite its invalidating a (relatively) settled interpretive practice.

Third, the appeal of "historical gloss" is limited insofar as it is fundamentally at odds with familiar formalist approaches to interpretation. Bradley and Siegel rightly contrast the "historical gloss" approach with "originalist" and "texualist" methods of interpretation. (140) Pursuant to those methods, what a text means turns exclusively on linguistic considerations. (141) By contrast, "historical gloss" grounds textual meaning in decidedly nonlinguistic considerations, in particular political branch practice not at all proximate to the enactment of the corresponding text. (142) For that reason, "historical gloss" also involves a rejection of the "fixation thesis," a basic premise of most any originalist or textualist theory. (143) If, for example, historical practice contrary to a text's "most natural" reading becomes sufficiently settled, the meaning of that text might flip, such that the "unnatural" reading becomes "clear." (144) According to Bradley and Siegel, this surprising implication is a feature of the "historical gloss" approach, not a bug. (145) The reason is that it facilitates textual "updating," permitting legal texts "to evolve in response to ... changing needs." (146) Needless to say, such "updating" is anathema to constitutional or statutory formalists, for whom Article V (147) and Article I, Section 7 (148) are, respectively, of the utmost importance. (149)


This Article's working hypothesis is that to say that it is "clear" (or "plain" or "unambiguous") that something is the case is, roughly speaking, to claim that one is in an epistemic position to "know" it. (150) Various linguistic data support this hypothesis. Like claims about "knowledge," claims about "clarity" appear to be factive (i.e., truth entailing). (151) In addition, claims about "clarity" and claims about "knowledge" seem to be mutually warranting--that is, if one is warranted in claiming that something is "clear," one is warranted in claiming to "know" it, and vice versa. (152)

Building on the above hypothesis, this Part articulates an alternate, epistemological explanation or justification of courts' disparate treatment of text in high- and low-stakes cases. Drawing on contemporary work in philosophy of language and epistemology, it argues that courts' hesitancy to identify "clear" or "plain" statutory meaning in high-stakes cases is plausibly an instantiation of the more general reluctance on the part of ordinary speakers to claim to "know" things when the practical stakes are raised. As this Part goes onto explain, this general reluctance on the part of speakers reflects a basic insight concerning the relationship between epistemic and practical reason, specifically that the epistemological justification required to act on some premise increases as do the practical stakes.

A. Motivating Examples

In recent years, philosophers have offered various technical explanations of the relationship between the practical stakes of a situation and the use of certain epistemological predicates. In particular, a great deal of work has gone into making sense of how changing the practical stakes affects speakers' willingness to claim to "know" things. This work has been motivated largely by a handful of intuitive, everyday examples that suggest the appropriateness of "knowledge" attributions vary according to the practical circumstances. (153) Perhaps the best-known of these examples are the so-called Bank Cases, imagined by Keith DeRose:
   [low stakes]: My wife and I are driving home on a Friday afternoon.
   We plan to stop at the bank on the way home to deposit our
   paychecks. But as we drive past the bank, we notice that the lines
   inside are very long, as they often are on Friday afternoons.
   Although we generally like to deposit our paychecks as soon as
   possible, it is not especially important in this case that they be
   deposited right away, so I suggest that we drive straight home and
   deposit our paychecks on Saturday morning. My wife says, "Maybe the
   bank won't be open tomorrow. Lots of banks are closed on
   Saturdays." I reply, "No, I know it'll be open. I was just there
   two weeks ago on Saturday. It's open until noon."

   [high stakes]: My wife and I drive past the bank on a Friday
   afternoon, as in [low stakes], and notice the long lines. I again
   suggest that we deposit our paychecks on Saturday morning,
   explaining that I was at the bank on Saturday morning only two
   weeks ago and discovered that it was open until noon. But in this
   case, we have just written a very large and very important check.
   If our paychecks are not deposited into our checking account before
   Monday morning, the important check we wrote will bounce, leaving
   us in a very bad situation. And, of course, the bank is not open on
   Sunday. My wife reminds me of these facts. She then says, "Banks do
   change their hours. Do you know the bank will be open tomorrow?"
   Remaining as confident as I was before that the bank will be open
   then, still, I reply, "Well, no. I'd better go in and make
   sure." (154)

The Bank Cases are noteworthy in that they suggest that the appropriateness of claiming to "know" something can vary with the practical stakes, holding constant the considerations that bear on the truth or falsity of the proposition at issue (e.g., the available evidence). (155) In low stakes, it seems appropriate for the speaker to claim to "know" that the bank will be open on Saturday on the basis of his recent experience. In high stakes, by contrast, it seems appropriate for the speaker to refrain from claiming to "know" that the bank will be open. The speaker's evidence concerning whether the bank will be open on Saturday is the same in each case. The only difference, seemingly, is that the practical consequences of mistakenly acting as if the bank will be open on Saturday are much greater in high stakes (e.g., a bounced check) than in low stakes.

Because of the seeming connection between claiming that something is "clear" and claiming to "know" that thing, analogous examples can, unsurprisingly, be constructed for ascriptions of "clarity." (156) For instance:
   ORDINARY: My wife and I are driving home on the evening of the New
   Hampshire primary. We plan to stop at the polling place on the way
   home to cast our votes. But as we drive past the place, we notice
   that the lines inside are very long, as they often are on the day
   of the primary. Although we generally like to vote, it is not
   especially important in this case that we do so. All of the
   candidates seem unremarkable, and all have similar policy
   platforms. For these reasons, I suggest that we drive straight
   home, trusting that our preferred candidate will prevail. My wife
   says, "Maybe she won't win. Lots of times turnout is the deciding
   factor." I reply, "No, it's clear she'll win. I checked the polls
   just a few days ago. She's comfortably ahead."

   TRANSFORMATIVE: My wife and I drive past the polling place, as in
   ordinary, and notice the long lines. I again suggest that we drive
   straight home, explaining that I checked the polls a few days ago
   and that our preferred candidate is ahead. But in this case, the
   candidate we prefer is markedly different from the other
   candidates, potentially a transformative figure. If she wins, her
   candidacy will gain immediate legitimacy, opening the door to a new
   era of American politics. My wife reminds me of these facts. She
   then says, "As you know, polls are wrong sometimes. Is it clear
   that she will win?" Remaining as confident as I was before that our
   preferred candidate will win, still, I reply, "Well, no. We'd
   better go in and vote just to be safe."

Like the Bank Cases, these examples suggest that the appropriateness of claiming that something is "clear" can vary with the practical stakes. In each case, the speaker's evidence concerning whether the preferred candidate will win regardless is the same. Be that as it may, because the practical consequences of mistakenly acting as if the preferred candidate will win are much lesser in ordinary than in transformative, it seems appropriate for the speaker to claim that the election outcome is "clear" in the former case but not in the latter.

B. Technical Explanations

Technical explanations of the examples discussed in Section II.A vary. DeRose and other so-called contextualists argue that such examples show that attributions of "knowledge" are context sensitive in that utterances of the form "X knows that p" express different propositions in different contexts of use. (157)

So characterized, attributions of "knowledge" are much like attributions of "tallness" or "flatness." (158) The same person, for instance, might be fairly characterized as "tall" in a conversation about gymnastics, but not in a conversation about basketball. The standard explanation is that to characterize someone as "tall" is to say that that person is tall relative to some comparison class determined by the context of use. In a conversation about gymnastics, to utter the sentence "Karen is tall" is to say, very roughly, that Karen is tall in relation to other gymnasts. By contrast, to utter the same sentence in a conversation about basketball is to claim, again roughly, that Karen is tall in relation to other basketball players. (159)

Turning back to "know," contextualists claim that the proposition expressed by uttering a sentence of the form "X knows that p" depends upon the practical stake the conversational participants have in the truth of p. (160) In a low-stakes situation, one in which the truth of p matters not very much to the participants, to say that one "knows that p" is to claim something like that one knows that p in a weak sense--that is, in relation to a moderately demanding epistemic standard. (161) By contrast, to utter that same sentence in a high-stakes situation (i.e., a situation in which the truth of p matters a great deal to the participants, is to claim that one knows that p in a strong sense (i.e., in relation to a very demanding epistemic standard). Hence, in low stakes, the speaker is in a position to claim to "know" that the bank will be open on Saturday because the evidence available to him satisfies a moderately demanding epistemic standard. By contrast, because that same evidence does not satisfy a very demanding epistemic standard, the speaker in high stakes reasonably refrains from claiming to "know" that the bank will be open.

A second explanation, defended by so-called interest-relative invariantists such as Jason Stanley, is that utterances of the form "X knows that p" express the same proposition regardless of the practical stakes: that X knows that p. (162) At the same time, whether X knows that p depends, according to such authors, not just upon familiar considerations such as X's evidence concerning p or whether p is true, but also, quite unconventionally, on X's practical stake in the truth of p. As Stanley puts it, "The basic idea is that, the greater the practical investment one has in a belief, the stronger one's evidence must be in order to know it." (163) In other words, for interest- relative invariantists, knowledge is fundamentally a practical concept. (164)

Interest-relative invariantism thus provides a straightforward explanation for speakers' reluctance to claim to "know" things in high-stakes situations-- namely, that it is more difficult to know something if the practical stakes are raised. In low stakes, for example, the speaker can claim to "know" that the bank is open because, given the conversational participants' limited practical interest, knowledge requires only moderate epistemic justification. In high stakes, by contrast, to know that the bank will be open would require very strong epistemic justification, a degree of justification the speaker apparently lacks. Given the conversational participants' heightened practical interest, merely moderate justification will not do. (165)

A third, pragmatic explanation suggested by Jessica Brown and others is that what varies according to the practical stakes is not the proposition expressed by an attribution of "knowledge," but rather whether that attribution is conversationally appropriate. (166) Here, a helpful analogy can be drawn to Paul Grice's classic examples of conversational implicature. (167) Suppose, for instance, that A is standing by an obviously immobilized car and is approached by B. A says to B, "I am out of petrol." B responds, "There is a garage round the corner." As Grice observed, by uttering the sentence "There is a garage round the corner" in this context, B implies that the garage is (at least possibly) open and sells petrol. The reason, Grice explains, is that otherwise B's response to A would be conversationally irrelevant and so inappropriate. (168) What matters to B (and hence to A), practically speaking, is that she is able to secure some petrol.

With "know," the suggestion by Brown and others is that the proposition expressed by uttering a sentence of the form "X knows that p" is just that X stands in a specific epistemic relation--namely, knowing--with p. At the same time, what a speaker communicates indirectly by uttering such a sentence can vary with the practical stakes. Thus, in the Bank Cases, the suggestion is that the proposition the speaker expresses is the same in each case: that he knows that the bank will be open on Saturday. At the same time, what the speaker communicates indirectly in each case is different. In low stakes, what matters to the conversational participants, practically speaking, is whether the speaker satisfies a moderately demanding epistemic standard as to the bank's being open. Thus, it is conversationally appropriate for him to say that he "knows" that the bank will be open so long as he satisfies that moderately demanding standard. By contrast, in high stakes, what matters to the conversational participants is whether the speaker satisfies a very demanding epistemic standard with respect to the bank being open. Hence, in that case, it is inappropriate for him to claim to "know" that the bank will be open if, as it seems, he fails to satisfy that more demanding standard.

Needless to say, adjudicating between these various technical explanations of how speakers use "know" goes beyond the scope of this Article. (169) Worth mentioning, however, is that explaining how speakers use "clear" might be a bit more straightforward. A central objection to contextualist explanations of "know" is that "know" does not appear to behave like other, uncontroversially context-sensitive terms. Stanley, for example, observes that unlike "tall" or "flat," "know" is not obviously gradable--that is, whereas it makes sense to describe someone as "very tall" or "taller" than someone else, "know" has no obvious analogues. (170) "Clear," by contrast, is plainly gradable--one can say, for example, that it is "very clear" that one's preferred candidate will win, or that the outcome of the gubernatorial election is "clearer" than the outcome of the Senate race. What this suggests is that, however one makes sense of "know," "clear" probably admits of a contextualist explanation. Indeed, such an explanation seems largely unavoidable: insofar as something can be more or less clear, context must determine how clear something must be to count as "clear" for purposes of a given conversation. So construed, to claim that something is "clear" in a low-stakes situation is to say that one satisfies a moderately demanding epistemic standard in relation to the thing at issue. By contrast, to say that something is "clear" in a high-stakes situation is to claim that one satisfies a very demanding epistemic standard with respect to that thing.

The reasoning above is similar to Chief Justice Marshall's analysis of "necessary" in McCulloch v. Maryland. (171) In that case, Marshall observed that "[a] thing may be necessary, very necessary, absolutely or indispensably necessary." (172) Because the term "admits of all degrees of comparison," Marshall continued, attention to "context" is necessary to determine whether the term is best read in a "rigorous" or "more mitigated sense." (173) A difference worth mentioning, however, is that Marshall's reasoning probably supports the more modest thesis that appeal to context is necessary to determine whether "necessary" is used literally or, as he put it, "figuratively]" (174)-- insofar it is not obviously felicitous to say that something is "more necessary" than something else, that "necessary" is a gradable adjective is at least controversial. By contrast, insofar as it is plainly felicitous to say that something "clearer" than something else," it seems comparably plain that "clear" is a gradable adjective and, in turn, that appeal to context is necessary to determine the threshold for "clarity" even if that term is used literally.

C. Basic Insight

Whichever technical explanation one prefers, a straightforward connection between epistemic justification and practical interest comes through. On any of the above explanations, it is appropriate to claim to "know" something only if one has adequate epistemic justification as to that thing. And, on any of those explanations, what counts as adequate justification depends upon the practical interests of those involved. In a low-stakes situation, the truth of the thing at issue (e.g., that the bank will be open on Saturday, that the preferred candidate will win the election) matters not very much to the participants in the conversation. Thus, what matters to those participants, practically speaking, is just that someone claiming to "know" that thing has moderate epistemic justification for it. In a high-stakes situation, by contrast, the truth of the thing at issue matters a great deal to the conversational participants. In those situations, what matters, practically speaking, is that claims to "know" that thing be supported by very strong epistemic justification.

A premise of the examples considered above is that the parties involved have considered all available evidence pertaining to their decision. A further way in which raising the practical stakes can affect our epistemic burdens is by increasing the amount of evidence it is reasonable to consider. This is just a corollary of the basic insight discussed above. In a low-stakes situation, it will often be reasonable to act after considering only limited evidence, the reason being that the epistemic justification required to act on some premise is relatively low. By contrast, in a high-stakes situation, reason will often require that one seek out additional evidence, assuming that time permits.

Because this Article focuses on Supreme Court cases, it seems reasonable to assume that, as in the above examples, the Court has considered all available evidence concerning statutory meaning in both high- and low-stakes statutory cases--as a rule, the quality of advocacy before the Supreme Court is excellent, such that even in low-stakes cases, the Court has all of the arguments in front of it. That said, one might argue that, reasonably, the Court considers arguments more or less carefully depending on the practical stakes of the case. If that's right, a further albeit related reason why the Court might find "plain" meaning less often in high-stakes cases is that, upon considering the arguments more carefully, the Court realizes that the issue is more complicated than it might have seemed given just a cursory glance. Again, because the connection between practical stakes and duty of inquiry is just a corollary of the basic insight discussed above, this variation of the story about why the Court behaves as it does is not really a competing explanation. Still, it is a variation worth noting.

The coupling of epistemic justification and practical interest in our linguistic practice lends support to an already intuitive connection between epistemic and practical rationality. Specifically, it bolsters the principle that the epistemological justification required to act on some premise increases as do the practical stakes. This principle of rationality is seemingly reflected in all sorts of everyday conduct. In the Bank Cases, for example, it is not just that it is reasonable for the speaker in low stakes to say that he "knows" that the bank will be open on Saturday; it is also reasonable for him to act on that premise--that is, it is okay for him to drive past the bank. By contrast, in high stakes, it seems reasonable, even mandatory, for the speaker to refrain from acting on that premise. Similar cases abound. The epistemic burden for acting on the premise that one turned off the stove plausibly varies according to whether one is leaving for the store or a weeklong vacation. So too the burden for acting on the premise that the holding of a particular Supreme Court decision was such and such, depending on whether one is at trivia night or arguing before the Supreme Court.

As the Bank Cases show, how to act in the absence of "knowledge" depends upon whether one course of action constitutes "playing it safe." In high stakes, for example, it would be equally unreasonable for the speaker to claim to "know" that the bank will be closed on Saturday. Be that as it may, the speaker stops and waits in the line. In stopping, however, he acts not on the premise that the bank will be closed on Saturday. Rather, he acts on the premise that whether the bank will be closed is uncertain. And under conditions of uncertainty, stopping is the safe thing to do for the reason that the cost of stopping if the bank turns out to be open on Saturday is much lower than the cost of driving by if it turns out to be closed. (175)

In addition to everyday life, the identified connection between epistemic justification and practical interest seems manifest in our law--in particular, our criminal law. (176) The increased burden of proof for criminal conviction, for example, suggests that acting on the premise that a defendant is guilty of a criminal offense is higher stakes than acting on the premise that she committed a civil violation. Less formally, the common sentiment that the beyond-a- reasonable-doubt standard is or should be more demanding in capital cases appears to reflect a practical understanding of "reasonable doubt." (177)

This intuitive principle of rationality--again, that the epistemological justification required to act on some premise increases as do the practical stakes--has ready application in the area of statutory interpretation. The contemporary consensus is that courts must adhere to statutory meaning when "clear" (or "plain" or "unambiguous"). (178) Per this Article's working hypothesis, for a court to say that statutory meaning is "clear" is for it to claim to have epistemic justification to say that it "knows" what that statute means. Thus, for a court to declare statutory meaning "clear" is for it to claim to have epistemic justification to act on the premise that the statute means what the court takes it to mean. As the discussion above suggests, however, what counts as epistemic justification to act on the premise that a statute has a certain meaning depends on the practical stakes of the case. If reading a statute in a particular way would raise no significant concerns, then moderate epistemic justification would seem to suffice for a court to declare that meaning "clear." If, by contrast, reading a statute that way would have "grave consequences," (179) then rationality would seem to require that a court have very strong epistemic justification before acting on the premise that it should be so read.

As in the everyday life, what to do in the absence of "knowledge" of statutory meaning depends upon whether there is a course of action that constitutes playing it safe. By instructing courts to resolve cases on nonlinguistic grounds if statutory meaning is not "clear," the various doctrines discussed in Part III seem to reflect the judgment that, in the relevant cases, nonlinguistic resolution is the safe course of action under conditions of uncertainty about meaning. (180) In other words, what the various doctrines instruct is that, under conditions of uncertainty, courts should resolve statutory cases on the basis of some value other than interpretive accuracy. (181) And while the relative cost assessments contained in those doctrines are open to question (e.g., "Is erroneous invalidation really that bad?",182) those doctrines do at least provide a positive law justification for courts resolving cases on nonlinguistic grounds when statutory meaning is uncertain. (183)


This Part applies the epistemological insight identified in Part II to different statutory interpretive settings. Section III.A considers statutory cases involving constitutional challenges--cases the Court expressly regards as high stakes. Section III.B looks at cases involving nonconstitutional challenges to major statutes, paradigmatic high-stakes cases from the perspective of citizens. As Sections III.A and III.B argue, the connection between practical stakes and epistemic justification analyzed in Part II helps to explain (or at least justify) courts' seemingly loose treatment of text in these two areas.

Section III.C discusses the potential application of the above insight to criminal cases. As this Section observes, current treatment of text in criminal cases suggests that courts regard such cases as relatively low stakes. To the extent this attitude is normatively unjustifiable, however, this Section contends that, as a purely epistemic matter, much more aggressive application of the rule of lenity is called for.

Section III.D looks at implications for judicial deference to agency interpretations of agency-administered statutes. As this Section explains, whether the epistemological insight identified above recommends more or less deference depends upon which aspect of deferring to agencies one thinks is more practically or constitutionally significant: (1) courts deferring to agencies concerning "what the law is," or (2) courts failing to defer to agencies concerning matters of policy.

A. Constitutional Avoidance

According to Justice Holmes, "[T]o declare an Act of Congress unconstitutional ... is the gravest and most delicate duty that this Court is called on to perform." (184) The reasons courts articulate in support of this claim vary. Sometimes they cite "respect for Congress, which we assume legislates in the light of constitutional limitations." (185) Other times courts cite the "prudential concern that constitutional issues not be needlessly confronted." (186) Whatever their reasons, however, courts agree that cases involving constitutional challenges are unambiguously high stakes. (187)

Starting from that premise, this Section suggests that courts' nonstandard treatment of statutory text in constitutional cases is partially attributable to (or at least justified by) the raised stakes of those cases. Again, pursuant to the canon of constitutional avoidance, a court will not read a statute in a way that raises serious constitutional doubts if an alternate reading is "fairly possible." (188) What this Section tries to show is that courts' assessment of what is "fairly possible" in such cases is plausibly (and reasonably) affected by the perceived practical stakes.

1. Bond v. United States

The Chemical Weapons Convention Implementation Act of 1998-(189)--which implements the near-identically worded international Convention on Chemical Weapons, ratified by the Senate in 1997 (190)--prohibits the knowing "possess[ion]" or "use" of "any chemical weapon." (191) "Chemical weapon" is defined to include any "toxic chemical" not used for a "peaceful purpose." (192) "Toxic chemical," in turn, is defined as "any chemical" that "can cause death, temporary incapacitation or permanent harm to humans or animals." (193)

In Bond, federal prosecutors charged a Pennsylvania woman with two counts of possessing and using chemical weapons in violation of the Act. (194) Upon discovering that her "closest friend" had engaged in an extramarital affair with her husband, the defendant acquired two chemicals conceded to be "toxic to humans and, in high enough doses, potentially lethal." (195) She proceeded to spread those chemicals on the friend's "car door, mailbox, and door knob," hoping that the friend would "develop an uncomfortable rash." (196) Though mostly unsuccessful, the defendant did cause the friend to suffer "a minor chemical burn on her thumb, which [the friend] treated by rinsing with water." (197)

At first blush, application of the statute in Bond is straightforward. The defendant's "purpose," after all, was evidently not "peaceful." (198) And the chemicals she possessed and used were "potentially lethal" and so seemingly "toxic." (199) Under the "plain" meaning of the statute, the defendant thus acted in violation. (200) Or so it would seem.

On appeal, the defendant argued that her conviction should be set aside on the ground that the Act as applied to her conduct exceeded Congress's enumerated powers and invaded powers reserved to the states by the Tenth Amendment. (201) The Constitution, the defendant maintained, does not permit Congress to police "local crime," a "bedrock principle" unaffected by the implementation of a valid, non-self-executing treaty. (202) In so arguing, the defendant appeared to call into question the Court's century-old precedent Missouri v. Holland, (203) which stated that " [i]f the treaty is valid there can be no dispute about the validity of the statute" that implements it "as a necessary and proper means to execute the powers of the Government." (204)

Rather than address that constitutional question, the Court sided with the defendant on statutory grounds, holding that the Act, interpreted correctly, did not cover her conduct. (205) Anticipating skepticism, the Court gestured at the epistemological insight discussed in Part II, remarking that otherwise "dear" text can be made "ambigu[ous]" by the "deeply serious consequences of adopting" its otherwise most natural reading. (206) As the Court observed, our "constitutional structure" leaves the prosecution of "purely local crimes" to the states. (207) Therefore, it continued, one should hesitate to infer that Congress intended to "upset the Constitution's balance between national and local power" by "defining] as a federal crime conduct readily denounced as criminal by the States." (208) The Court emphasized further the "ordinary meaning" of the phrase "chemical weapon" calls to mind "chemical warfare," not "spreading irritating chemicals on [a] doorknob." (209) For all of these reasons, the Court concluded that, to whatever conduct the statute extends, (210) it did not extend to this "unremarkable local offense." (211) Concurring in the judgment, Justice Scalia ridiculed the majority for its reliance upon, in his view, a laughable if also dangerous epistemological principle. (212) He likewise dismissed the majority's specific reasoning, insisting that, for example, the "ordinary meaning" of a phrase is "irrelevant" when that phrase is specifically and clearly defined. (213)

For the reasons articulated in Part II, Justice Scalia was wrong to mock the majority's epistemological principle. He was right, nonetheless, to criticize the majority's application thereof. As Part II argued, raising the practical stakes makes it more difficult to "know" what a statute means. It does not, however, make it impossible to do so. Nor does it make all readings equally or even minimally plausible. In Bond, the majority rightly observed that, given the constitutional stakes, it was more difficult to "know" what Congress meant by "chemical weapon" than in the ordinary case. (214) Be that as it may, the majority put forth only minimal effort to explain why, constitutional stakes notwithstanding, the series of seemingly precise and applicable statutory definitions did not control the case. (215) In terms of linguistic analysis, really all the Court had to offer was that sometimes it "go[es] without saying" that a class of cases is implicitly excluded from a superficially general prescription. (216) That's true: sometimes, for example, "person" refers just to natural persons; (217) other times "conviction" means domestic conviction. (218) Be that as it may, the majority in Bond gave no indication which class of cases the statute excluded implicitly (compare, for example, corporate persons, foreign convictions). Nor did it explain to which specific class of cases the Act was implicitly limited. (219) Such specification of Congress's meaning is part and parcel of a plausible linguistic story of implicit exclusion. (220) Here instead the majority reasoned, in effect, that this case was surely excluded, (221) harkening back to the earlier era of ad hoc judicial carve-outs. (222)

2. Northwest Austin

Bond is helpful in that it supplies a judicial gloss on the epistemological insight discussed in Part II. Northwest Austin, by contrast, says nothing of interest. It does, however, better show that insight in action in the sense that it provides a more plausible example of a case in which the raised practical stakes justify (and possibly explain) the Court's deviation from a text's "most natural reading."

Like Bond, commentators conceived Northwest Austin principally as a constitutional case. (223) In Northwest Austin, the plaintiff was a small Texas utility district seeking relief from the "preclearance" requirements set forth by section 5 of the Voting Rights Act of 1965. (224) Under section (5), covered "States" and "political subdivisions" are required to obtain federal approval before changing their election laws in any way. (225) Section 4(b), in turn, contains a "coverage formula" used to determine which "States" or "political subdivisions" are subject to preclearance requirements. (226) Any covered "State" or "political subdivision" may seek relief from preclearance requirements pursuant to the "bailout" provision contained in section 4(a). (227)

In Northwest Austin, the plaintiff argued that it was a "political subdivision" within the meaning of section 4(a) and so entitled to seek bailout relief. (228) As the Court observed, "There is no dispute that the district is a political subdivision of [a covered State] in the ordinary sense of the term." (229) It is, after all, "[a] division of a state that exists primarily to discharge some function of local government." (230) The problem for the plaintiff, as the district court panel observed below, (231) was that section 14(c)(2) of the Act specifically defines "political subdivision" as "any county or parish, except that where registration for voting is not conducted under the supervision of a county or parish, the term shall include any other subdivision of a State which conducts registration for voting." (232) Because the plaintiff did not conduct voting registration, it plainly fell outside that definition. (233) Therefore, it seemed straightforwardly ineligible for bailout relief under section 4(a).

In addition to its statutory argument, the plaintiff in Northwest Austin argued in the alternative that the preclearance requirements imposed by section 5 exceeded Congress's powers under the Enforcement Clause of Fifteenth Amendment and so were unconstitutional. (234) According to the plaintiff, the Act's preclearance requirements "differentiate[] between the States, despite our historic tradition that all the States enjoy 'equal sovereignty.'" (235) Therefore, deviation from equal sovereignty "requires a showing that a statute's disparate geographic coverage is sufficiently related to the problem that it targets." (236) The concern raised by the plaintiff was that the problem that section 5 was intended to address "may no longer be concentrated in the jurisdictions singled out." (237) As the Court observed, "The statute's coverage formula is based on data that is now more than 35 years old, and there is considerable evidence that it fails to account for current political conditions." (238)

Again, rather than address that constitutional question, the Court sided with the plaintiff on statutory grounds, holding that the plaintiff was a "political subdivision" for purposes of section 4(a) even if not for purposes of section 14(c)(2). According to the Court, the definition of "political subdivision" articulated in [section] 14(c)(2) applied to some but not all uses of that phrase within the Act. (239) Specifically, the Court held that the definition applied to uses of the phrase in section 4(b), but not to uses in section 4(a) or, for that matter, section 5. (240) To substantiate this claim, the Court referred to its earlier decision in United States v. Board of Commissioners, (241) in which it remarked that section 14(c)(2)'s definition applied to uses of "political subdivision" in section 4(b), but not to those in section 5. (242) Reasoning that it had already established that section 14(c)(2)'s definition was of limited application within the Act, the Court concluded that it was at least plausible that that definition did not apply to section 4(a), and, in turn, that "political subdivision" as used in section 4(a) retained its ordinary, more expansive meaning.

As the district court panel observed, the Court's reading of section 4(a) is not without textual difficulty. (243) Specifically, section 4(a)'s bailout provision applies not to "political subdivisions" in general, but rather to
   any political subdivision of such State (as such subdivision
   existed on the date such determinations were made with respect to
   such State), though such determinations were not made with respect
   to such subdivision as a separate unit, or in any political
   subdivision with respect to which such determinations have been
   made as a separate unit. (244)

The function of the italicized language is to make clear that a "political subdivision" of a covered state is permitted to seek bailout relief even if that subdivision has not been deemed subject to preclearance requirements specifically. (245) In other words, even if a subdivision is "covered" only because the subdivision's state was determined to be "covered" pursuant to an application of the coverage formula articulated in section 4(b), that subdivision can request relief apart from the state as a whole. (246) The problem that the italicized language creates for the Court's analysis is that it strongly suggests that the "political subdivisions" referred to in section 4(a) were at least eligible for coverage determination pursuant to an application of the coverage formula in section 4(b). (247) But, of course, the only "political subdivisions" eligible for a section 4(b) determination are those that fall within section 14(c)(2)'s definition.

Under normal circumstances, the textual difficulty just identified would seemingly be enough to render the Court's reading of section 4(a) unavailable. What this Section suggests, however, is that given the heightened practical stakes of the case--again, this is a case involving a constitutional challenge, to a key provision of a landmark civil rights statute, no less--and, correspondingly, the increased epistemological justification required for "knowledge" of statutory meaning, (248) it is at least plausible that the Court's preferred reading is "fairly possible." To see why, consider first that the Court's reading is not, strictly speaking, foreclosed by the structure of the sentence in section 4(a). So long as some subset of the "political subdivisions" referenced in section 4(a) are eligible for coverage determination pursuant to section 4(b), one can read "political subdivision" in section 4(a) broadly without violating what Grice called the "maxim of quantity." (249) By analogy, suppose that Congress were to enact a public-financing system for congressional races, providing matching funds for any candidate who has raised in excess of $10,000 in her congressional district, though she has not been approved for presidential-primary matching funds, or any candidate who has been approved for presidential-primary matching funds. (250) The eligibility criteria for president are more restrictive than those for Congress. (251) Thus, not all candidates for Congress are eligible to run for president simultaneously, or, in turn, to be considered for presidential-primary matching funds. Be that as it may, it would be bizarre to suggest that "candidate," as used above, referred only to candidates eligible to run for president simultaneously.

A related but distinct worry is that the Court's reading of section 4(a) renders the italicized language superfluous. (252) If, after all, in-state "political subdivisions" ineligible for a section 4(b) determination are eligible for bailout relief, what is the purpose of specifying that a subdivision need not have received such a determination? Again, under normal conditions, this argument would have real force. Consider, though, that there is a conceivable response. Just before the addition of the italicized language, the Court had held in City of Rome v. United States that only "political subdivisions" specifically designated for coverage under section 4(b) were eligible for bailout relief under section 4(a). (253) The effect of the subsequent amendment was thus to eliminate specific designation as a prerequisite for bailout relief. So even if the italicized language is, in terms of content, redundant on the Court's reading of section 4(b), that language could conceivably serve to emphasize the change in the law brought about by that amendment. (254)

More generally, note that, in contrast to Bond, the Court in Northwest Austin is able to articulate a distinctly linguistic story in support of its seemingly unusual reading. (255) As in Bond, the Court's reading appears to be undermined by a clear, applicable statutory definition. Unlike in Bond, however, the Court was able to offer a specific explanation as to why that definition is inapposite--namely, that the definition applies to certain subsections but not to others. Schematically, that explanation is familiar to ordinary language and, for that matter, the law. (256)

In sum, although the Court's reading of the statute in Northwest Austin is not without difficulty, it is, upon closer inspection, at least not ridiculous. And, given the high stakes of the case--the prospect of invalidating an act of Congress, and a significant civil rights statute at that--perhaps that is enough to render its reading "fairly possible."

B. Nonconstitutional Challenges

Following the Court's textualist turn, litigants have taken to "challenging" statutes on nonconstitutional grounds. Such challenges consist of a litigant advancing an interpretation that, if accepted, would radically curtail the implementation regime of the statute at issue. In recent years, litigants have mounted nonconstitutional challenges to a number of major statutes. (257) These cases are, from the public's perspective, the very definition of "high stakes": What could be higher stakes than litigation threatening to upset the implementation of hugely important, politically salient legislation? Up to now, such challenges have been mostly unsuccessful. The reason, according to critics, is that courts have been willing to stretch or even disregard statutory text to preserve the status quo. (258)

This Section argues instead that courts' repeated rejection of nonconstitutionai challenges to major statutes is at least partially attributable to (or justified by) the epistemic insight identified in Part II. In such cases, courts have indeed rejected readings of statutes that would have been "clear" under ordinary circumstances. Because of the incredibly high stakes of these cases, however, this Section suggests that courts are epistemically rational in exhibiting extraordinary caution before accepting readings that would have such unsettling effects.

1. King v. Burwell

In King v. Burwell, the plaintiffs challenged the implementation regime of the Patient Protection and Affordable Care Act (PPACA), (259) the most significant health reform legislation since the passage of Medicare and Medicaid. (260) As widely discussed, the PPACA consists primarily of three interdependent reforms. (261) First, the Act prohibits insurers from denying coverage or increasing premiums on the basis of preexisting conditions (community rating). (262) Second, it imposes a tax penalty on nonexempt individuals who fail to maintain coverage (individual mandate). (263) Third, it provides subsidies in the form of tax credits for the purchase of insurance by low-income persons (subsidies). (264) Together, community rating and subsidies help make insurance affordable for customers by ensuring a price not in excess of a reasonable percentage of income. At the same time, the individual mandate helps make the provision of affordable insurance financially feasible for insurers by ensuring a broad risk pool.

As King illustrated, a concern with the above characterization is that the specific language of the PPACA's subsidies provision appears to limit the availability of tax credits in a way that is largely incompatible with the Act functioning as a "three-legged stool." (265) The worry results from the Act's permitting different types of health insurance "exchanges," state-specific marketplaces on which customers can compare and purchase insurance policies. (266) Under section 1311 of the Act, "[e]ach State shall ... establish an [exchange] for the State." (267) In turn, section 36B of the Internal Revenue Code, enacted as part of the PPACA, instructs that tax credits shall be available for persons who purchase health insurance "through an Exchange established by the State under [section] 1311 of the [Act]." (268) Because, however, Congress cannot require states to implement federal laws, (269) if a state refuses or is unable to set up an exchange, section 1321 of the Act provides that the federal government, through the secretary of Health and Human Services, "shall ... establish and operate such Exchange within the State." (270) As Jonathan Adler and others observed, these provisions read together seem to suggest that the PPACA authorizes tax credits only for insurance purchased on a state-run exchange--that is, an exchange "established" by a state "under [section] 1311." (271)

In King, the plaintiffs challenged an Internal Revenue Service (IRS) rule interpreting the PPACA's subsidy provision as authorizing the agency to grant tax credits to persons who purchased insurance through either a state-run or a federally facilitated exchange. (272) The challengers argued that the IRS rule was inconsistent with the plain language of section 36B, which, according to them, authorizes subsidies for insurance purchased through state-run exchanges alone. (273) At the time of the challenge, more than half of states utilized federally facilitated exchanges. (274) Within two years of the rule's adoption, millions of individuals had purchased insurance through such an exchange, with the vast majority relying upon subsidies. (275) If the challenge were to have succeeded, health insurance would thus have been rendered unaffordable for a huge number of would-be customers absent state or congressional action. In turn, most of those individuals would have become exempt from the individual mandate on grounds of financial hardship. Under these conditions, the individual mandate would plausibly have failed to produce a broad enough risk pool to avoid adverse selection, causing premiums to increase precipitously. (276)

For doctrinal reasons, the government conceded in King that section 36B was precisely worded--that is, that the language at issue was not attributable to a drafting mistake. (277) Given that concession, the Government was left to argue that section 36B's reference to exchanges "established by the State under [section 1311]" is a "term of art," encompassing state-run and federally facilitated exchanges alike. (278) Famously, the Supreme Court agreed, reasoning that, "when read in context," the language of section 36B was "ambiguous," and that although the "most natural reading" of the pertinent phrase was limited to state-run exchanges, it was "also possible" that the phrase referred to federally facilitated exchanges as well. (279) As evidence, the Court cited various anomalies the plaintiff's reading would produce elsewhere in the statute, requiring, for example, the creation of federally facilitated exchanges on which there would be no "qualified individuals" eligible to shop, (280) as well as the reporting of information for a reconciliation of tax credits that could never occur. (281) As to the apparent contrast between exchanges "established "under [section] 1311" and section 1321, respectively, the Court observed that section 1321 instructs the secretary to "establish and operate such Exchange within the State." (282) "By using the phrase 'such Exchange,"' the Court explained, section 1321 "instructs the Secretary to establish and operate the same Exchange that the State was directed to establish" under section 1311. (283)

So was a broad reading of section 36B at least "possible"? Under normal circumstances, the answer would seem to be no. Again, given the plain contrast between section 1311 and section 1321 together with section 36B's reference to exchanges "established under [section] 1311," it would be "natural" to infer that section 36B excludes exchanges established under section 1321. More still, that section 1321 instructs the secretary to establish "such Exchange within the State," would suggest ordinarily that an exchange established under section 1321 is qualitatively similar to but numerically distinct from an exchange "established by the State under [section] 1311." (284) And while the various anomalies to which the Court pointed are real--and indicative of a drafting mistake (285)--those anomalies do not cause section 36B's specific language to vanish. (286)

As the Court noted, however, these were not normal circumstances. To accept the plaintiffs' reading of section 36B would have been to undermine substantially the PPACA's implementation regime, potentially "destabilizing] the individual insurance market in any State" with a federally facilitated exchange. (287) Under such circumstances, it would make sense for a court to require increased epistemic justification before regarding the "destabilizing]" reading as "clear." Whether King is best understood as a display of reasonable epistemic caution is, of course, open to question. (288) Also unclear is whether the Court's preferred reading was "fairly available," even considering the heightened practical stakes. (289) Regardless, what King represents is a type of case in which it would be entirely reasonable, as an epistemic matter, for a court to look at a text with more hesitation than it would in a run-of-the-mill case. And, if nothing else, the Court in King did look at section 36B with greater-than-usual skepticism--and it did so with an eye to the practical stakes of the case. (290)

2. Inclusive Communities

Whereas King was about preempting the implementation of a new statute, the challenge in Inclusive Communities was about curtailing the implementation of an old one.

In Inclusive Communities, the defendant was a state agency responsible for the distribution of federal low-income housing tax credits. (291) A local nonprofit sued the agency, alleging that the agency's "disproportionate" allocation of those tax credits--skewed toward projects in "predominantly black inner-city areas" and away from those in "predominantly white suburban neighborhoods" (292)--caused segregated housing patterns to persist in violation of the Fair Housing Act (FHA), (293) a landmark 1968 civil rights law. (294) Importantly, the nonprofit's claim was a so-called disparate-impact claim, a claim predicated upon disproportionate adverse effects on minorities not justified by a legitimate rationale. (295) A disparate-impact claim contrasts with a claim of intentional discrimination, a claim that a defendant acted with discriminatory intent or motive. (296)

The question presented in Inclusive Communities was "whether disparate-impact claims are cognizable under the FHA." (297) Arguing no, the defendant relied significantly upon the "plain text" of the statute. (298) In relevant part, the FHA makes it unlawful to "refuse to sell or rent after the making of a bona fide offer, or to refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex, familial status, or national origin." (299) The Act similarly provides that
   [i]t shall be unlawful for any person or other entity whose
   business includes engaging in residential real estate-related
   transactions to discriminate against any person in making available
   such a transaction, or in the terms or conditions of such a
   transaction, because of race, color, religion, sex, handicap,
   familial status, or national origin. (300)

According to the defendant, because the FHA is limited to actions taken "because of' a protected trait, it precludes disparate-impact liability. In the words of Justice Alito, "A person acts 'because of something else ... if that something else 'was the "reason" that the [person] decided to act.'" (301) And a protected trait is only one's "reason" for action if discriminatory treatment is intentional. As Justice Alito went on to observe, "many other federal statutes use the phrase 'because of to signify what that phrase means in ordinary speech," (302) including, for example, the federal hate crime statute, which authorizes enhanced sentences for certain crimes only if the defendant acted with discriminatory intent. (303)

As in King, the Court rejected the restrictive reading of the statute proposed, apparent textual strength notwithstanding. Holding that disparate-impact claims are, indeed, cognizable under the FHA, the Court emphasized that the Act had long been so understood. By 1988, the Court noted, all nine courts of appeals to have addressed the question had so held. (304) Furthermore, when Congress amended the FHA that year, it left unaltered the portions of the statute at issue--evidence, according to the Court, that Congress had "accepted and ratified" those "unanimous" holdings. (305)

The Court relied, in addition, on two prior decisions in which it held that similar "because of' language created disparate-impact liability in similar statutes. First, in 1971, the Court held in Griggs v. Duke Power Co. (306) that provisions in Title VII making it unlawful for an employer to make an employment decision "because of' a protected trait of an employee or applicant encompassed disparate-impact claims. (307) In Griggs, the Court based its decision almost entirely upon Congress's apparent purpose of "achiev[ing] equality of employment opportunities" and "removing] barriers" that have operated to the advantage of white employees. (308) In a sign of an earlier methodological era, the Court made no effort to explain how its holding could be reconciled with the specific wording of the provisions at issue. (309) More than thirty years after Griggs, in Smith v. City of Jackson, (310) a plurality held that sections of the Age Discrimination in Employment Act of 1967 (ADEA) that render unlawful employment decisions made "because of' an employee's or applicant's age similarly create disparate-impact liability. (311) Recognizing the parallel with Griggs, the plurality insisted that, despite its reliance on purpose, that decision also "represented the better reading of the statutory text." (312) According to the plurality, because Title VII makes unlawful actions that "deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of' a protected trait, the statute evinces a concern not just with employer motivation but with "the effects of the action on the employee." (313) So too, the plurality reasoned, with the ADEA, which prohibits action that "would deprive or tend to deprive any individual of employment opportunities or otherwise adversely affect his status as an employee, because of' that individual's age. (314)

Returning to Inclusive Communities, the Court reasoned that, as with Title VII and the ADEA, the language of the FHA indicates a concern not just with motive but with consequence. Specifically, the FHA makes it unlawful to "refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of' a protected trait. (315)

This, according to the Court, is the same sort of "results-oriented language" that supported recognition of disparate-impact liability in both Griggs and Smith. (316)

The Court's reasoning on the above point is shaky. Even if the FHA (and Title VII and the ADEA) make unlawful the bringing about of certain consequences, unlawfulness is nonetheless contingent upon bringing about those consequences "because of' a protected trait. In other words, even if liability can result despite one's having some nondiscriminatory goal, it still seems that, to be liable, one must have a discriminatory motivation. To illustrate the contrast, suppose that a landlord were to tell racially stereotypical jokes to minority rental applicants, intending not to dissuade those applicants from renting, but rather to cause them to laugh. In that case, the landlord might still be liable under the FHA for "mak[ing]" his rental property "unavailable" to minority applicants insofar as her jokes make minority applicants uncomfortable and so discourage them from renting. Although the landlord's goal in joking is innocent (e.g., to cause laughter), the consequence of his joking is not (e.g., to discourage minority applicants). More still, the landlord's motivation for joking is partially discriminatory in that one of his reasons for joking with minority applicants is that those applicants are minorities--in part, the landlord told the jokes he did because of the applicant's minority status.

So, in an ordinary case, a court would probably read the "because of' language at issue to prohibit only intentional discrimination. That seems to be the better reading of the statute. Indeed, under typical circumstances, one could fairly characterize that language as "clear." Be that as it may, the circumstances surrounding Inclusive Communities were far from typical: adopting the "most natural" reading of the statute would have dramatically curtailed the implementation of a major civil rights statute. (317) Given those unusually high stakes, one can see how members of the Court might reasonably have become less than certain that the "most natural" reading was, in fact, correct. Particularly so, given the longstanding judicial practice of reading the FHA and similar statutes "unnaturally"--under the circumstances, considerations of epistemic humility alone might have been enough to shake a justice's confidence in her own linguistic assessment. (318)

The previous two Sections dealt with types of cases that are uncontroversially "high stakes," either from the perspective of courts (constitutional challenges) or from the standpoint of citizens, judges and justices included (nonconstitutional challenges to major statutes). Because courts take those cases to matter specially, one would expect those courts, for the reasons described in Part II, to be more hesitant to claim "knowledge" of statutory meaning in those cases and, hence, more reluctant to characterize statutory meaning in those cases as "clear." And this is, of course, exactly what we see.

The next two Sections address types of cases the status of which as "high stakes" is more controversial. Building on the previous two Sections, these Sections suggest that one can infer from courts' reluctance (or, alternatively, zeal) to declare statutory meaning "clear" whether, rightly or wrongly, courts regard those cases as "high" or "low" stakes. In turn, one can then ask whether courts' assessment of the stakes is reasonable.

C. Rule of Lenity

The rule of lenity requires courts to resolve unclarity in criminal statutes in favor of defendants. (319) The purpose of the rule is to "ensure that those subjected to criminal prosecution have adequate notice of the conduct that the law prohibits," (320) and to prevent courts, rather than legislatures from "defin[ing] criminal activity." (321) Structurally, the rule of lenity is identical to the canon of constitutional avoidance: if statutory meaning is less than clear, courts must select among available readings on the basis of nonlinguistic considerations.

Same as the avoidance canon in form, the rule of lenity could hardly be more different in operation. As discussed in Section III.A, courts have a high threshold for what counts as "clear" text in the face of constitutional questions. In criminal cases, by contrast, courts' threshold for clarity is remarkably low. Indeed, that threshold is so low as to approach its logical limit.

In the abstract, the current approach to lenity is summarized neatly as follows:
   The simple existence of some statutory ambiguity ... is not
   sufficient to warrant application of th[e] rule, for most statutes
   are ambiguous to some degree. "The rule of lenity applies only if,
   after seizing everything from which aid can be derived, we can make
   no more than a guess as to what Congress intended." To invoke the
   rule, we must conclude that there is a "grievous ambiguity or
   uncertainty in the statute." (322)

In other words, so long as courts have some inclination about statutory meaning, lenity has no role to play. Functionally, then, statutory text counts as "clear" in criminal cases so long as courts have an "opinion as to the best reading." (323) Put differently, so long as some reading seems to courts more plausible than the others, operationally speaking, that reading is "plainly" correct.

In the concrete, courts' handling of criminal statutes is exemplified by the Court's recent decision in Lockhart v. United States. (324) In that case, the question before the Court was whether an enhanced sentencing provision applicable to defendants with a "prior conviction ... under the laws of any State relating to aggravated sexual abuse, sexual abuse, or abusive sexual conduct involving a minor or ward" (325) applies to a defendant whose only prior state law conviction is for sexual abuse of an adult. (326) Arguing no, the defendant in Lockhart--previously convicted under New York state law for sexual abuse of his then-fifty-three-year-old girlfriend--contended that the participial phrase "involving a minor or ward" binds each item in the list of predicate crimes ("aggravated sexual abuse," "sexual abuse," and "abusive sexual conduct"). (327) Because his prior conviction was not for "sexual abuse ... of a minor or ward," the defendant continued, the enhancement provision thus did not apply. In response, the government argued that "involving a minor or ward" binds only the last item in the list of predicate offenses ("abusive sexual conduct"), and so the defendant's prior state law conviction for "sexual abuse" of an adult was enough to trigger the enhancement provision. (328)

Siding with the government, the Court assured that "the provision's text and context together reveal a straightforward reading." (329) Writing for the majority, Justice Sotomayor invoked what it called the "rule of the last antecedent," according to which a "limiting clause or phrase ... should ordinarily be read as modifying only the noun or phrase that it immediately follows." (330) Acknowledging that this "rule" reflected at best a presumption, Justice Sotomayor went on to note the seeming parallel between the language at issue and the chapter of the federal criminal code pertaining to "sexual abuse," a "prior conviction" under which triggered the same enhancement provision. (331) As Justice Sotomayor observed, the first three sections of that chapter are titled "Sexual Abuse," "Aggravated Sexual Abuse," and "Sexual Abuse of a Ward or Minor," respectively. (332) According to Justice Sotomayor, that "similarity appears to be more than a coincidence." (333) Swatting aside the defendant's rule-of-lenity argument, Justice Sotomayor thus concluded that the "text and structure of [the provision] confirm [ed] " the government's reading. (334)

In dissent, Justice Kagan offered various ordinary conversation and legal examples in which, intuitively, the "modifying phrase ... applies to each term in the preceding list, not just the last" (e.g., "the laws, the treaties, and the constitution of the United States"). (335) In an effort to match canon with canon, Justice Kagan also invoked what she called the "series-qualifier" canon, (336) the principle that "'[w]hen there is a straightforward, parallel construction that involves all nouns or verbs in a series,' a modifier at the end of the list 'normally applies to the entire series.'" (337) Despite its "fancy name," Justice Kagan insisted, that principle "reflects the completely ordinary way that people speak and listen, write and read." (338)

Whether Justice Sotomayor or Justice Kagan has the better reading in Lockhart is hard to say. Commenting on the exchange, Judge Frank Easterbrook remarked that he has no idea who is right. (339) But if even archtextualist Judge Easterbrook can't make heads or tails of a text, how can the rule of lenity fail to apply? The only explanation, it seems, is that, in criminal cases, whatever reading a court deems "best" is, almost by definition, "clear" enough to render that rule inapplicable. What this suggests is that courts regard criminal cases as having remarkably low stakes. Indeed, if courts deem statutory meaning "clear" in criminal cases just in virtue of having an "opinion as to the best reading," the attributed stakes could not be any lower.

That courts implicitly regard criminal cases as low stakes--indeed, the lowest of stakes--is somewhat surprising, or at least disappointing, given the special concern with criminal conviction embedded in our constitutional and broader legal system. (340) If, for example, "criminal conviction constitutes a formal judgment of [moral] condemnation by the community," imposing "stigma" and causing "damage to the defendant's reputation," (341) then surely greater epistemic caution is called for than when adjudicating a generic civil dispute. Perhaps in an era of mass incarceration, criminal conviction truly is run of the mill. (342) Still, insofar as courts at least purport to take criminal conviction very seriously, it is disconcerting that, in practice, courts attribute to criminal cases minimal practical significance. What this suggests, then, is that, as a normative matter, the rule of lenity ought to look much like the avoidance canon, not just in form, but also in operation. If criminal conviction carries with it such significant consequences, then it seems that, as an epistemic matter, courts should be much more hesitant to declare criminal statutes "clear." This is not to claim that all criminal cases are high stakes--whether, for example, some class of criminals is subject to a ten-year, as opposed to a five-year mandatory minimum sentence is perhaps not the most pressing of issues. Nor is it to suggest that all criminal cases are equally significant. (343) The claim here instead is just that, contrary to current practice, the typical criminal case ought not to be regarded as having the lowest stakes imaginable. (344)

D. Chevron

Like the rule of lenity, the so-called Chevron doctrine is a rule for resolving statutory unclarity. Pursuant to that doctrine, a court is, famously, to defer to an agency's reasonable construction of a statute it administers unless, as to the "precise question at issue," statutory meaning is "clear." (345) The standard justification for so deferring is equal parts democratic and technocratic. Unclear statutory language, the Court has reasoned, leaves "gaps" in the law that need to be filled, and filling those gaps inevitably "involves difficult policy choices." (346) And because agencies are both more democratically accountable (347) and more technically expert than courts, (348) it makes sense for courts to defer to them on "how best to construe" indeterminate language "in light of competing policy interests." (349) Based on these considerations, courts attribute to Congress the--concededly fictional (350)--intention to delegate to agencies rather than to courts the authority to resolve such unclarity. (351)

In recent years, various judges and justices, principally on the right side of the political spectrum, have hinted or outright declared that Chevron should be reconsidered. Justice Thomas, for example, has argued that relying upon agencies to resolve statutory unclarity violates the separation of powers. (352) Drawing on the work of Phillip Hamburger, Justice Thomas has asserted that "the judicial power, as originally understood, requires a court to exercise its independent judgment in interpreting and expounding upon the laws." (353) And because "[t]hose who ratified the Constitution knew that legal texts would often contain ambiguities," a court's obligation includes " resolv[ing] these ambiguities over time." (354) Similarly, then-judge Neil Gorsuch declared that deferring to agency interpretations amounts to "[transferring the job of saying what the law is from the judiciary to the executive," and that doing so "invites the very sort of due process (fair notice) and equal protection concerns the framers knew would arise if the political branches intruded on judicial functions." (355) Also citing Professor Hamburger, then-judge Gorsuch went on to insist that permitting "agencies rather than courts to declare the law's meaning" leads inevitably to the "exploitation of) ambiguous laws as license for [the executive's] own prerogative," a prospect of which the founders were exceedingly "wary." (356) Deploying less heavy artillery, (357) Justice Scalia, in some of his later administrative law opinions, appeared to suggest that deferring to an agency construction of an agency-administered statute is inconsistent with the instruction by the Administrative Procedure Act (APA) that a "reviewing court shall decide all relevant questions of law" (358) and "interpret ... statutory provisions." (359) Though Justice Scalia went on to assure that Chevron "at least was in conformity with the long history of judicial review of executive action," (360) others such as Aditya Bamzai have called that historical claim into question. (361)

Assessing the merits of these critiques goes beyond the scope of this Article. (362) Here, instead, the thing to observe is that what a court makes of these critiques--as well as the doctrine's justification--is likely to affect, for purely epistemic reasons, how freely it defers to agencies under the current regime. Start with the determination whether the Chevron framework applies to the dispute in question, so-called Chevron step zero. (363) As first articulated, Chevron appeared to be a generic framework for disputes involving agency-administered statutes. (364) Over time, however, courts indicated a willingness to sort among administrative law disputes by type, (365) considering whether, for example, the Chevron framework applied to disputes concerning an agency's jurisdiction (366) or ones involving questions of "deep 'economic and political significance.'" (367) In each instance, the question presented was whether Congress would really intend to delegate to agencies rather than to courts the authority to make decisions of this type. (368) So characterized, one can see how a court's (or an individual judge's or justice's) receptiveness to step-zero arguments might be affected, for purely epistemic reasons, by its attitude toward Chevron in general: to the extent that a court (or individual judge or justice) regards deferring to agencies as constitutionally dubious, that court would reasonably demand heightened epistemic justification before acting on the premise that Congress intends to delegate such decisionmaking authority to agencies over some class of disputes. By contrast, if a court (or individual judge or justice) does not share that worry, it would require only ordinary epistemic justification before acting on that premise.

Turn next to the determination whether Congress has spoken "clearly" in an individual case. Here, the epistemic situation reverses. For Chevron proponents--unmoved by Hamburgerian critiques and persuaded by supporting democratic and technocratic arguments (369)--the risky premise upon which to act is that Congress has answered the "precise question at issue." (370) To explain: it is common ground among proponents and detractors of deference that "law" gives out at some point, after which there are questions of "policy." (371) For proponents of deference, or at least strong proponents of deference, the worrisome prospect is less of agencies exercising judicial authority than of courts making policy in complex areas absent technical competence or democratic mandate. (372) Given that concern, it makes sense for strong proponents to require increased epistemic justification before acting on the premise that Congress has spoken "clearly" on a particular question. By contrast, for those not disturbed by prospect of judicial policymaking, ordinary epistemic justification should suffice before acting on the premise that "law" settles the dispute at issue.

Whether this is how things play out in practice is tough to assess. Chief Justice Roberts, a consistent critic of Chevron, has also been a consistent vote in favor of restrictions on Chevron's domain. (373) On the other hand, Justice Thomas, the Court's most vocal critic of Chevron as of late, (374) has been less reliable on that front. (375) As to the so-called step-one determination, whether statutory meaning is "clear," (376) the suggestion above might seem to run contrary to Justice Scalia's observation that those "who find[] more often ... that the meaning of a statute is apparent" are more likely to think Chevron "desirable." (377) That remark, however, had to do with propensity to see "apparent" meaning in general, something Justice Scalia attributed a judge's interpretive methodology. (378) Holding interpretive methodology constant, it is hard to tell whether a distaste for Chevron correlates with finding "clear" meaning at step one. (379)

Data aside, whether the epistemological insight discussed in this Article recommends more or less deference to agencies thus turns on what, in a court's view, makes a decision whether to defer "high stakes." If the major worry is exercise by the executive branch of judicial authority, deference will seem appropriate even less often. If, on the other hand, the more pressing concern is judge-made communications or environmental policy, deference will appear proper all the more.


As David Strauss and others have noted, courts adhere to constitutional text much less closely than they do to the text of statutes. (380) As with statutory interpretation, courts insist that where constitutional text is "clear" or "unambiguous," the text controls. (381) At the same time, pursuant to Chief Justice Marshall's famous reminder that "it is a constitution we are expounding," (382) courts frequently construe constitutional text in ways that would be unimaginable in the modal statutory case. To give two examples, courts agree that the First Amendment applies to all branches of government despite the text instructing that "Congress shall make no law." (383) Similarly, courts accept that the Constitution bars suits in federal court by private citizens against their home state (384)--this notwithstanding the Eleventh Amendment's reference to suits "commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State." (385)

Given the inherently high-stakes nature of constitutional interpretation, one might think that the epistemological insight discussed in this Article helps to explain (or at least justify) the disparate treatment of constitutional text just described. (386) For reasons that Strauss articulates, however, it is hard to believe that courts are really trying to read the constitutional text. First, some claims--for instance, that "Congress," as used in the First Amendment, is a synecdoche referring to all branches of the federal government (387)--are so at odds with ordinary norms of interpretation that, if true, they would seem to entail interpretive skepticism. (388) To borrow Justice Scalia's phrasing, if "Congress" can mean all branches of government, the "[w]ords no longer have meaning." (389)) Second, as Strauss observes, in the modal constitutional decision, constitutional text serves at most a "ceremonial role," with "the serious analysis focus[ing] on the precedents." (390) In Strauss's view, courts' inattention to constitutional text suggests that constitutional "interpretation" is really something like common law adjudication, with constitutional text "treated in more or less the same way as precedents in a common law system." (391)

Whatever courts are doing with constitutional text, it seems plain that what they are not doing is attending to linguistic nuance. But if constitutional "interpretation" is, as Strauss suggests, a misnomer, (392) it is highly doubtful that the connection between epistemological and practical reason discussed here sheds any light on courts' treatment of constitutional text vis- a-vis statutory text. Perhaps if courts did care what the Constitution says, their ability to identify constitutional meaning definitively would be limited by the heightened practical stakes of that task. (393) On the other hand, perhaps discerning constitutional meaning would be easier in some cases rather than others, owing not just to differences in textual clarity, (394) but also to differences in the practical stakes. (395) Regardless, because it is so speculative to imagine what our constitutional order would look like if constitutional adjudication consisted of careful reading, speculating about the implications of contemporary epistemology for that would-be practice is not especially useful at this stage.


This Article is not so naive as to suggest that all loose treatment of statutory text is attributable to epistemological hurdles. Sometimes politics trumps text, right or wrong. What this Article claims, more modestly, is that, as an epistemological matter, there are limits to what one can reasonably expect of courts in high-stakes cases. (396) Happily, courts seem committed what comes afterward"); see also David A. Strauss, The Living Constitution (2010) (articulating a common law theory of constitutional adjudication). For an alternative, more eclectic nonoriginalist account, see, for example, Mitchell N. Berman & Kevin Toh, Pluralistic Nonoriginalism and the Combinability Problem, 91 Tex. L. Rev. 1739, 1741 (2013), which suggests that an array of nonlinguistic factors help to determine constitutional law. for the most part to following statutory text where "clear." Courts can only do so, however, if doing so is epistemologically feasible. And, for the reasons this Article explains, it is simply harder to "know" what a statute means when the practical stakes are raised. Again, this is not to say that it is impossible to discern statutory meaning in high-stakes cases, just that to do so is more difficult. The purpose of this Article is thus not to excuse all deviations from text in big cases. (397) It is, instead, to recommend a bit more sympathy-- and hence, a bit less cynicism--as the Court goes about its work at the end of the Term.

Ryan D. Doerfler, Assistant Professor of Law, University of Pennsylvania Law School. Thanks to Hrafn Asgeirsson, William Baude, Mitchell Berman, Jonah Gelbach, Sophia Lee, Jeffrey Pojanowski, Richard Re, and participants at the Law, Language, and Normativity Workshop at the Surrey Centre for Law and Philosophy for helpful comments and suggestions.

(1.) Justice Elena Kagan, The Scalia Lecture: A Dialogue with Justice Kagan on the Reading of Statutes, at 8:28 (Nov. 17, 2015), scalia-lecturekagan-discusses- statutoryinterpretation/ [].

(2.) Hardt v. Reliance Standard Life Ins. Co., 560 U.S. 242, 251 (2010).

(3.) See, e.g., Universal Health Servs., Inc. v. United States, 136 S. Ct. 1989, 2002 (2016) ("[P]olicy arguments cannot supersede the clear statutory text."); Ross v. Blake, 136 S. Ct. 1850, 1856 (2016) ("Statutory interpretation, as we always say, begins with the text.... [T]he statute speaks in unambiguous terms...."); Nichols v. United States, 136 S. Ct. 1113, 1119 (2016) ("[E]ven the most formidable argument concerning the statute's purposes could not overcome the clarity we find in the statute's text." (quoting Kloeckner v. Solis, 568 U.S. 41, 55 n.4 (2012))); Sebelius v. Cloer, 133 S. Ct. 1886, 1895 (2013) ("Our 'inquiry ceases [in a statutory construction case] if the statutory language is unambiguous and the statutory scheme is coherent and consistent.' " (alteration in original) (quoting Barnhart v. Sigmon Coal Co., 534 U.S. 438, 450 (2002))).

(4.) John F. Manning, The New Purposivism, 2011 Sup. Ct. Rev. 113, 114 ("[T]he Court in the last two decades has mostly treated as uncontroversial its duty to adhere strictly to the terms of a dear statutory text, even when doing so produces results that fit poorly with the apparent purposes that inspired the enactment."); see also Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 12 (2012) ("Holy Trinity is a decision that the Supreme Court stopped relying on more than two decades ago.").

(5.) See, e.g., Cloer, 133 S. Ct. at 1886; Nichols, 136 S. Ct. at 1113; Freeman v. Quicken Loans, Inc., 566 U.S. 624 (2012); Milner v. Dep't of the Navy, 562 U.S. 562 (2011).

(6.) Manning, supra note 4, at 115 (emphasis added); accord Ryan D. Doerfler, The Scrivener's Error, 110 Nw. U. L. Rev. 811, 827-30 (2016).

(7.) This is, to be clear, a claim about our positive law of statutory interpretation. See William Baude Sc Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079 (2017). In principle, a statute's contribution to the law could diverge sharply from its communicative content. See Mark Greenberg, The Standard Picture and Its Discontents, in 1 Oxford Studies in Philosophy of Law 39, 48 (Leslie Green Sc Brian Leiter eds., 2011) (observing that the phrase "legal interpretation" is ambiguous between ascertaining the communicative content of a legal text and determining its legal significance). This claim assumes also that the communicative content of the statute is not superseded by some other source of law (e.g., the Constitution). See Hrafn Asgeirsson, Can Legal Practice Adjudicate Between Theories of Vagueness?, in Vagueness and Law: Philosophical and Legal Perspectives 95, 103-04 (Geert Keil & Ralf Poscher eds., 2016) (arguing that the communicative content of a statute is coextensive with its legal content absent some "rebutting" or "undercutting" source of law).

(8.) See Clark v. Martinez, 543 U.S. 371, 381 (2005); accord Rust v. Sullivan, 500 U.S. 173, 191 (1991); Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575 (1988).

(9.) Neal Kumar Katyal & Thomas P. Schmidt, Active Avoidance: The Modern Supreme Court and Legal Change, 128 Harv. L. Rev. 2109, 2112 (2015).

(10.) That is, cases in which a litigant advances an interpretation the acceptance of which would dramatically limit the practical effect of a statute relative to the existing implementation regime.

(11.) George F. Will, On Obamacare, John Roberts Helps Overthrow the Constitution, Wash. Post (June 25, 2015), helps-overthrow- theconstitution/2015/06/25/47d9ffdelb67-11 e5-93b7-5eddc056ad8a_story.html []; see also Ilya Shapiro, How John Roberts Begat Donald Trump, Federalist (May 5, 2016), begat-don ald-trump [] (attributing to the chief justice "contempt for [the] rule of law").

(12.) Cf Frederick Schauer, Statutory Construction and the Coordinating Function of Plain Meaning, 1990 Sup. Ct. Rev. 231, 232 (arguing that ordinary meaning serves a coordinating function in low-stakes cases).

(13.) Adrian Vermeule (@avermeule), Twitter (June 27, 2016, 12:42 PM), https://twitter. com/avermeule/status/747515425764810752 [].

(14.) Richard M. Re, The New Holy Trinity, 18 Green Bag 2d 407, 417 (2015).

(15.) Id. at 421.

(16.) See Curtis A. Bradley 8; Neil S. Siegel, After Recess: Historical Practice, Textual Ambiguity, and Constitutional Adverse Possession, 2014 Sup. Ct. Rev. 1, 1-2 [hereinafter Bradley & Siegel, After Recess]', Curtis A. Bradley & Neil S. Siegel, Constructed Constraint and the Constitutional Text, 64 Duke L.J. 1213, 1216-17 (2015) [hereinafter Bradley Sc. Siegel, Constructed Constraint].

(17.) NLRB v. Noel Canning, 134 S. Ct. 2550, 2559 (2014) (alteration in original) (quoting The Pocket Veto Case, 279 U.S. 655, 689 (1929)).

(18.) See Bradley & Siegel, Constructed Constraint, supra note 16, at 1264-87 (allowing that "the meaning of an otherwise clear numerical provision could become unclear" through subsequent practice).

(19.) E.g., Bradley & Siegel, After Recess, supra note 16, at 61-63.

(20.) That is, reasons reflecting Congress's apparent policy aims (e.g., to expand health insurance coverage, to curtail insider trading).

(21.) Re, supra note 14, at 417.

(22.) Re cites Yates v. United States, 135 S. Ct. 1074 (2015), as an example of a low-stakes case in which purposive reasons overcome an otherwise "plain" textual meaning. See Re, supra note 14, at 411-13. As I have argued elsewhere, however, that assessment of Yates rests on a confused understanding of "plain" meaning. See William Baude & Ryan D. Doerfler, The (Not So) Plain Meaning Rule, 84 U. Chi. L. Rev. 539 (2017).

(23.) But see infra notes 131-139 and accompanying text (discussing Milner v. Dep't of the Navy, 562 U.S. 562, 580-81 (2011)).

(24.) That is, incomplete as an explanation of the disparate treatment of text in highstakes--low-stakes cases. To the extent that Bradley and Siegel set out to explain a subset of that phenomenon (or a distinct but overlapping phenomenon), it is no discredit to their account that it fails to explain in full the phenomenon under consideration here.

(25.) Re, supra note 14, at 421; cf. King v. Burwell, 135 S. Ct. 2480, 2497 (2015) (Scalia, J., dissenting) ("Under all the usual rules of interpretation, in short, the Government should lose this case. But normal rules of interpretation seem always to yield to the overriding principle of the present Court: The Affordable Care Act must be saved.").

(26.) Lawrence B. Solum, The Fixation Thesis: The Role of Historical Fact in Original Meaning, 91 Notre Dame L. Rev. 1, 1 (2015).

(27.) Or, possibly, convey.

(28.) Baude & Doerfler, supra note 22, at 546-47.

(29.) See, e.g., Keith DeRose, The Case for Contextualism (2009); Jason Stanley, Knowledge and Practical Interests (2005); J. Brown, Contextualism and Warranted Assertibility Manoeuvres, 130 Phil. Stud. 407 (2006).

(30.) See Stewart Cohen, Contextualism, Skepticism, and the Structure of Reasons, 13 Phil. Persp. 57, 58 (1999) (offering an analogous example); see also infra Section II.A.

(31.) See generally Stanley, supra note 29.

(32.) See generally DeRose, supra note 29.

(33.) See generally Brown, supra note 29.

(34.) See, e.g., Stanley, supra note 29, at 9.

(35.) See, e.g., Baude & Sachs, supra note 7, at 1083 ("Yet we still have to decide the case. We don't keep fruitlessly hunting for a hidden meaning; but neither do we tell judges to fill the gap with whatever they think best. Instead, we use law to displace our ordinary inquiries about meaning."); Tun-Jen Chiang & Lawrence B. Solum, The Interpretation-Construction Distinction in Patent Law, 123 Yale L.J. 530, 537 (2013) ("[I]f a court chooses to follow the linguistic meaning of text, it must decide how to fill in the gaps when the linguistic meaning does not fully answer a legal dispute....").

(36.) Ashwander v. Tenn. Valley Auth., 297 U.S. 288, 345 (1936) (Brandeis, J., concurring).

(37.) See id. at 345-48.

(38.) Katyal & Schmidt, supra note 9, at 2112.

(39.) 134 S. Ct. 2077 (2014).

(40.) 557 U.S. 193 (2009).

(41.) See Bond, 134 S. Ct. at 2090.

(42.) Nw. Austin, 557 U.S. at 206-11.

(43.) Because judicial opinions are written for an audience, the implicit assessment of a case's relative importance contained in an opinion must be presumed by deciding court to be shared by that audience, on pains of being an uncooperative interlocutor. See Stefano Predelli, Painted Leaves, Context, and Semantic Analysis, 28 Linguistics & Phil. 351, 365 (2005) (observing that successful communication depends upon agreement among conversational participants as to what "matters"). Thus, it is not enough to render a case "high stakes" that it matters specially to the deciding court or to the litigants (as is almost always the case). Cf. Richard Primus, Debate, The Cost of the Text, 103 Cornell L. Rev. 1651, 1665 (2017) ("The parties to particular cases often have a great deal at stake, even if nothing about the case is unusually salient from the perspective of the decisionmaker.").

(44.) See, e.g., David A. Strauss, The Supreme Court: 2014 Term--Foreword: Does the Constitution Mean What It Says?, 129 Harv. L. Rev. 1 (2015) (cataloging consensus "readings" of constitutional provisions that are utterly at odds with the constitutional text).

(45.) See supra notes 10-11 and accompanying text.

(46.) See Gavin Bade, Why Insiders Think the EPA Got the Best of the Clean Power Plan Hearing Last Week, Util. Dive (Oct. 5, 2016),

powerplan-hearing/427657/ [] (quoting an attorney for petitioners challenging the Clean Power Plan: "I tend to agree that if you take all of the passion of climate change and the Paris agreement out, 1 think EPA loses 10-0 ... [b]ut there certainly seem to be some judges who were looking for a way to allow EPA to do something like this, and 1 think that's hard to take out of the case.").

(47.) See Katyal & Schmidt, supra note 9.

(48.) King v. Burwell, 135 S. Ct. 2480, 2495 (2015); accord Nat'l Fed'n of Indep. Bus. v. Sebelius, 567 U.S. 519, 574-75 (2012).

(49.) Nw. Austin Mun. Util. Dis't. No. One v. Holder, 557 U.S. 193, 204 (2009).

(50.) United States v. Locke, 471 U.S. 84, 92 (1985) ("[T]his Court will not pass on the constitutionality of an Act of Congress if a construction of the Act is fairly possible ... by which the constitutional question can be avoided.").

(51.) See Peretz v. United States, 501 U.S. 923, 929-30 (1991).

(52.) Katyal & Schmidt, supra note 9, at 2111-12.

(53.) Id. at 2112.

(54.) Id. at 2116.

(55.) Id. at 2118.

(56.) Id. at 2119-20.

(57.) Id. at 2114.

(58.) Id. at 2128.

(59.) Adler discusses Chief Justice Roberts's individual approach; here, I extrapolate from his analysis and apply it to the Roberts Court as a whole.

(60.) Jonathan H. Adler, Anti-Disruption Statutory Construction, 38 Cardozo L. Rev. 509, 511-12 (2016).

(61.) Id. at 511; see also id. at 528 ("There is often room for reasonable people to differ on the best interpretation of a complex statute, but some of the Chief Justice's opinions seem to stretch interpretive choices to their breaking point.").

(62.) Id. at 511.

(63.) Id. at 513-14.

(64.) Id. at 514.

(65.) Id. at 528 ("The analysis presented here is descriptive, not normative.").

(66.) Id.

(67.) Id. at 105.

(68.) See Katyal & Schmidt, supra note 9, at 2163 (observing that, by merely hinting at answers to constitutional questions via application of the avoidance canon, higher courts provide lower courts little guidance concerning how to handle similar challenges).

(69.) Cf. Adler, supra note 60, at 511 ("Whatever the merits of this approach, it [is] not textualism as we have come to know it.").

(70.) See id. at 511-12 (remarking that the Roberts Court's "preference for limiting the disruptive impact of the Court's decisions takes priority over any commitment to a particular interpretive technique").

(71.) Re, supra note 14, at 407-09 (discussing the evolution of purposivism in particular); see also id. at 421 (suggesting that his account make sense of the difference between "banal" and "unusual but pivotal cases").

(72.) Id. at 408.

(73.) Id. at 417.

(74.) Id.

(75.) Id. at 421

(76.) Id. at 417.

(77.) Id. at 408.

(78.) 143 U.S. 457, 459 (1892) ("It is a familiar rule that a thing may be within the letter of the statute and yet not within the statute, because not within its spirit, nor within the intention of its makers.").

(79.) Re, supra note 14, at 421.

(80.) Id. at 418.

(81.) John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387, 2388 (2003).

(82.) Id. at 2403.

(83.) Id. at 2389.

(84.) Id. at 2389-90.

(85.) See, e.g., Wyeth v. Levine, 555 U.S. 555, 601 (2009) (Thomas, J., concurring in the judgment) ("[A] statute's text might reflect a compromise between parties who wanted to pursue a particular goal to different extents."); Landgraf v. USI Film Prods., 511 U.S. 244, 286 (1994) (dicta) ("Statutes are seldom crafted to pursue a single goal, and compromises necessary to their enactment may require adopting means other than those that would most effectively pursue the main goal.").

(86.) See Manning, supra note 4, at 116.

(87.) See id.

(88.) See supra note 3 and accompanying text.

(89.) Manning, supra note 81, at 2392 ("Following such a course would more often require judges to accept seemingly odd or awkward results in particular cases, but doing so would serve important systemic values implicit in the constitutional structure-- legislative supremacy, the evenhanded application of statutes, respect for legislative compromise, and the conception of limited judicial power implicit in rationality review." (emphasis omitted)).

(90.) See Doerfler, supra note 6, at 833.

(91.) That is, a substantive mistake, as opposed to a linguistic mistake. See id. at 830-34.

(92.) Re, supra note 14, at 417.

(93.) Id. at 418.

(94.) Id. at 417.

(95.) See Manning, supra note 81, at 2456. At times, Re sounds as if he regards textualist judges as belonging to the old "plain meaning" school. See, e.g., Re, supra note 14, at 421 (attributing to textualists the principle that "legal ambiguity must be discoverable in text alone"). So understood, the double-counting problem articulated above disappears. It does so, however, only at the cost of rendering Re's opponent an anachronism.

(96.) To illustrate, Manning considers Puffendorfs classic example of a statute imposing criminal penalties on any person who "drew blood in the streets." Manning, supra note 81, at 2461 (citing United States v. Kirby, 74 U.S. (7 Wall.) 482, 487 (1868)). Puffendorf concluded, quite sensibly, that the statute should not extend to a surgeon who opened the vein of a person in the street felled by a seizure. As Manning observes, for the old "plain meaning" textualist, Puffendorfs conclusion is awkward since, "[r]ead literally," that statute seemingly extends to the surgeon. Id. For the contextualist, by contrast, that sensible conclusion is straightforwardly available. See id. at 2461-62 (observing that, for example, as used "in the criminal code, one might expect the term 'drew blood' to describe a violent act"). For extensive discussions of the ways in which context informs textual interpretation within a textualist framework, see also Ryan D. Doerfler, Who Cares How Congress Really Works?, 66 Duke L.J. 979, 986-98 (2017), which catalogues examples.

(97.) Doerfler, supra note 96, at 981.

(98.) See supra notes 21-22 and accompanying text.

(99.) See Re, supra note 14, at 409.

(100.) See supra note 5 (collecting cases).

(101.) See infra Part II.

(102.) See, e.g., Bradley & Siegel, After Recess, supra note 16, at 20-25, 44- 49; Bradley & Siegel, Constructed Constraint, supra note 16, at 1262-67; William Baude, Constitutional Liquidation 27-29 (Jan. 29, 2017) (unpublished manuscript) (on file with author).

(103.) 343 U.S. 579, 610 (1952) (Frankfurter, J., concurring) ("It is an inadmissibly narrow conception of American constitutional law to confine it to the words of the Constitution and to disregard the gloss which life has written upon them.").

(104.) Bradley & Siegel, After Recess, supra note 16, at 44.

(105.) Id. at 45.

(106.) Id. Bradley and Siegel contrast the "historical gloss" approach to the idea of "constitutional liquidation," according to which uncertainties about constitutional meaning would be "worked out, or 'liquidated,' through decisions and practices." Id. at 30. "Once liquidated," according to Bradley and Siegel, "the meaning of the Constitution on those questions would become 'fixed' and so not subject to change." Id. at 30-31. But see Baude, supra note 102 (manuscript at 34-48) (arguing that liquidation correctly understood does not require that "liquidated" meanings be permanently "fixed").

(107.) 134 S. Ct. 2550 (2014).

(108.) U.S. Const, art. II, [section] 2, cl. 3 ("The President shall have Power to fill up all Vacancies that may happen during the Recess of the Senate, by granting Commissions which shall expire at the End of their next Session."); Noel Canning, 134 S. Ct. at 2556.

(109.) Noel Canning, 134 S. Ct. at 2567; see also Ryan D. Doerfler, Go Big or Go Home: The Constitutionality of Recess Appointments Following Pro Forma Sessions of the Senate, 65 Admin. L. Rev. 975, 980-88 (2013) (providing a linguistic analysis of this portion of the Recess Appointments Clause).

(110.) Noel Canning, 134 S. Ct. at 2567.

(111.) Id. at 2568 ("The Clause's purpose strongly supports the broader interpretation. That purpose is to permit the President to obtain the assistance of subordinate officers when the Senate, due to its recess, cannot confirm them.").

(112.) Id. (noting, for example, that "Attorney General William Wirt advised President Monroe to follow the broader interpretation" of the Recess Appointments Clause); see also id. at 2570 (contending that " [h]istorical practice over the past 200 years strongly favors the broader interpretation").

(113.) Id. at 2606 (Scalia, J., concurring in the judgment).

(114.) Id. at 2606-07 (remarking that "the reasonable reader might have wondered, why would any intelligent drafter intending the majority's reading" have used language contained in the Clause, thereby making that reading "awkward and unnatural").

(115.) Id. at 2610-13.

(116.) Bradley & Siegel, After Recess, supra note 16, at 18.

(117.) Noel Canning, 134 S. Ct. at 2600 (Scalia, J., concurring in the judgment). Irrelevant, that is, insofar as historical practice is evidence of something other than the original public meaning of the Clause's language. See, e.g., Vasan Kesavan & Michael Stokes Paulsen, The Interpretive Force of the Constitution's Secret Drafting History, 91 Geo. L.J. 1113, 1164-76 (2003) (discussing the--in their view, limited--probative value of early congressional, executive, and judicial precedents to the original public meaning of the constitutional text).

(118.) See Noel Canning, 134 S. Ct. at 2559, 2568.

(119.) See id. at 2606-10 (Scalia, J., concurring in the judgment) (discussing the "plain meaning" of the language at issue). As Bradley and Siegel observe, Justice Scalia also appears moved by what he perceives as the "purpose" of the Recess Appointments Clause. See Bradley & Siegel, After Recess, supra note 16, at 48-49. As I have argued elsewhere, however, a drafter's apparent purpose is a quintessentially linguistic consideration. See Doerfler, supra note 96, at 995-98.

(120.) See Bradley & Siegel, Constructed Constraint, supra note 16, at 1241 (observing that for most originalists, non-linguistic considerations such as historical practice are relevant only if, on the basis of linguistic considerations, the constitutional text is unclear); see also id. (observing that originalists "are likely to accept [pre-ratification practice] but not [post-ratification practice] as relevant to textual interpretation, especially if the postratification practice occurs long after the Founding").

(121.) Bradley 8r Siegel, After Recess, supra note 16, at 48.

(122.) Id. at 47 ("[T]he Court's finding of ambiguity for the phrase 'vacancies that may happen' suggests substantial extratextual construction"); see also Bradley & Siegel, Constructed Constraint, supra note 16, at 1266.

(123.) See infra Part IV.

(124.) See supra notes 60-66 and accompanying text.

(125.) See Strauss, supra note 44, at 28 ("If an agency has consistently adhered to a view that seems to be at odds with the text of the statute, that might persuade a court to find a degree of vagueness or ambiguity in the text that the court would otherwise not perceive and to defer to the agency's longstanding view.").

(126.) Again, this is no discredit to the proponents of the "historical gloss" approach in the area of constitutional interpretation. The purpose of this Section is to explore a potentially helpful analogy, not to impute to the authors discussed the intention to explain areas of interpretation other than the one they address explicitly.

(127.) See infra Section III.B.

(128.) See infra Section III.B.2 (discussing Tex. Dep't of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., 135 S. Ct. 2507 (2015)).

(129.) 135 S. Ct. 2480 (2015).

(130.) See infra Section III.B. 1.

(131.) Milner v. Dep't of the Navy, 562 U.S. 562 (2011).

(132.) 5 U.S.C. [section] 552(b)(2) (2012).

(133.) Milner v. U.S. Dep't of the Navy, 575 F.3d 959, 968 (9th Cir. 2009), rev'd, 562 U.S. 562.

(134.) See Crooker v. Bureau of Alcohol, Tobacco 8e Firearms, 670 F.2d 1051 (D.C. Cir. 1981) (en banc), abrogated by Milner, 562 U.S. 562. The D.C. Circuit grounded its broad reading of the exemption in FOIA's "secondary purpose" of "preserving the effective operation of governmental agencies." Id. at 1074.

(135.) Milner, 562 U.S. at 585 (Breyer, J., dissenting). But see id. at 576 (majority opinion) ( "Prior to Crooker, three Circuits adopted the reading of [the exemption] we think right, and they have not changed their minds.").

(136.) Id. at 581.

(137.) Id. at 577.

(138.) Id. at 576.

(139.) Id. at 580-81.

(140.) See Bradley & Siegel, Constructed Constraint, supra note 16, at 1216-17, 1241.

(141.) Such methods often do permit consideration of nonlinguistic considerations to fill gaps in textual meaning. See, e.g., Lawrence B. Solum, Originalism and Constitutional Construction, 82 Fordham L. Rev. 453, 453 (2013) (distinguishing constitutional "interpretation," "the activity that discovers the communicative content or linguistic meaning of the constitutional text," from constitutional "construction," "the activity that determines the legal effect given the text" and observing that "the actual text of the U.S. Constitution contains general, abstract, and vague provisions that require constitutional construction that goes beyond the meaning of the text for their application to concrete constitutional cases").

(142.) See e.g., Bradley & Siegel, Constructed Constraint, supra note 16, at 1.

(143.) See Solum, supra note 26, at 1.

(144.) Perhaps the best example Bradley and Siegel provide is the consensus that the First Amendment's instruction that "Congress shall make no law ... [for example,] abridging the freedom of speech," U.S. Const, amend. I, applies not just to Congress but to all branches of the federal government. See Bradley & Siegel, Constructed Constraint, supra note 16, at 1243-47. But see infra Part IV.

(145.) Bradley & Siegel, Constructed Constraint, supra note 16, at 1242.

(146.) Bradley & Siegel, After Recess, supra note 16, at 63.

(147.) U.S. Const, art. V (establishing formal procedures for amending the Constitution).

(148.) U.S. Const, art. I, [section] 7 (establishing formal procedures for enacting legislation).

(149.) See, e.g., John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 714-15 (1997) (discussing the importance of bicameralism and presentment); Antonin Scalia, Originalism: The Lesser Evil, 57 U. Cin. L. Rev. 849, 862 (1989) (emphasizing the value of the "long and hard consideration required for a constitutional amendment").

(150.) Unlike claims about "knowledge" (e.g., "X knows that p."), claims about "clarity" are typically impersonal (e.g., "It is clear that p."). Therefore, claims about "clarity" appear to be indexed to a particular body of evidence (e.g., "It is clear (given the available evidence) that p."). In turn, claims about clarity seem to imply that those in the conversational circle have access to that body of evidence and so are in a position to "know" the proposition at issue.

(151.) For example, if I "know" that Oswald killed Kennedy, then Oswald in fact killed Kennedy. The same is true if it is "clear" that Oswald killed Kennedy.

(152.) The best evidence of mutual warrant is, perhaps, the infelicity of claiming to "know" something while denying that it is "clear," or vice versa ("#" indicates infelicity):

# (A) I know that Oswald killed Kennedy, but it isn't clear that he did.

# (B) It's clear that Oswald killed Kennedy, but I don't know that he did.

For reasons below, this Article takes no position on whether such claims are mutually entailing.

(153.) A further motivation is that varying the epistemic standards required to "know" something according to the practical circumstances has seemed to many a promising approach to solving various skeptical puzzles. See, e.g., DeRose, supra note 29, at 41- 43.

(154.) Keith DeRose, Contextualism and Knowledge Attributions, 52 Phil. & Phenomenological Res. 913, 913 (1992). The other well-known motivating example is Stuart Cohen's so-called Airport Case:
   Mary and John are at the L.A. airport contemplating taking a
   certain flight to New York. They want to know whether the flight
   has a layover in Chicago. They overhear someone ask a passenger
   Smith if he knows whether the flight stops in Chicago. Smith looks
   at the flight itinerary he got from the travel agent and responds,
   "Yes I know--it does stop in Chicago." It turns out that Mary and
   John have a very important business contact they have to make at
   the Chicago airport. Mary says, "How reliable is that itinerary? It
   could contain a misprint. They could have changed the schedule at
   the last minute." Mary and John agree that Smith doesn't really
   know that the plane will stop in Chicago. They decide to check with
   the airline agent.

Cohen, supra note 30, at 58.

(155.) More generally, cases of this sort suggest varying appropriateness of knowledge attributions--that is, claims of the form "X knows that p."

(156.) See DeRose, supra note 29, at 180-84.

(157.) See generally DeRose, supra note 29.

(158.) Contextualists also analogize "know" to familiar indexicals such as "I," "here," and "now." See, e.g., Stewart Cohen, How to Be a Fallibilist, 2 Phil. Persp. 91, 97 & n.14 (1988).

(159.) See also DeRose, supra note 29, at 180-82.

(160.) See, e.g., id. at 1-3.

(161.) An alternative gloss is that practical circumstances determine which possibilities are salient and so which possibilities need to be ruled out in order to "know" something. See Stewart Cohen, Contextualist Solutions to Epistemological Problems: Scepticism, Gettier, and the Lottery, 76 Australasian J. Phil. 289, 290-91 (1998).

(162.) See generally Stanley, supra note 29. Theorists such as Stanley are "invariantists" in the sense that they argue that the proposition expressed by an utterance of the form "X knows that p" does not vary according to the practical stakes that accompany the utterance. At the same time, such theorists are "interest relativists" in the sense that they insist that the truth conditions of the proposition expressed by utterances of the form "X knows that p" depend in part upon the practical interests of the relevant parties.

(163.) Id. at 88.

(164.) Interest-relative invariantism thus involves a rejection of what Stanley calls "intellectualism"--that is, the thesis that the truth of a given "knowledge" attribution turns only on familiar, truth-conducive considerations. Id. at 6-7.

(165.) Put differently, to claim to "know" something is, for the interest- relative invariantist, to allege sufficient epistemic justification for practical purposes.

(166.) Brown, supra note 29; see also Patrick Rysiew, The Context-Sensitivity of Knowledge Attributions, 35 Nous 477 (2001).

(167.) H.P. Grice, Logic and Conversation, in 3 Syntax and Semantics 41 (Peter Cole & Jerry L. Morgan eds., 1975).

(168.) Id. at 51. Compare this with a situation in which A's car is plainly in working order, and A says to B, "I am supposed to pick up a friend at a nearby garage." In that case, if B were to respond, "There is a garage round the corner," B's utterance would have no such implication.

(169.) Nor is this list of candidate explanations exhaustive. See, e.g., John MacFarlane, The Assessment Sensitivity of Knowledge Attributions, in 1 Oxford Studies in Epistemology 197 (Tamar Sbazo Gendler & John Hawthorne eds., 2005) (defending an assessment- relativist account of "knowledge" attributions); Jonathan Schaffer, From Contextualism to Contrastivism, 119 Phil. Stud. 73 (2004) (defending a contrastivist account of "knowledge" attributions).

(170.) Stanley, supra note 29, at 37-46.

(171.) 17 U.S. (4 Wheat.) 316, 413-15 (1819).

(172.) McCulloch, 17 U.S. at 414.

(173.) Id. at 414-15.

(174.) Id. at 414.

(175.) Contrast this with a situation in which neither course of action constitutes playing it safe. Suppose, for example, that a bomb is located in one of two buildings and is set to go off in a short period of time. Suppose further that the available evidence, although far from uniform, suggests on the whole that the bomb is in building A rather than building B. Finally, suppose that, because of the time constraint, the bomb squad can only secure one building or the other. In that paradigmatically high-stakes situation, it would be implausible for a member of the bomb squad to claim to "know" that the bomb is in building A and not building B. Be that as it may, the thing to do is for the bomb squad to rush to building A, acting on its "best guess" as to where the bomb is hidden. The reason is that the cost associated with rushing erroneously to building A or building B is symmetrical. Thus, the thing to do under conditions of uncertainty is to act on one's probabilistic assessment, however weak. Thanks to Mitch Berman for suggesting this type of example.

(176.) But see infra Part III.

(177.) See Jon O. Newman, Beyond "Reasonable Doubt", 68 N.Y.U. L. Rev. 979, 999- 1000 (1993) ("Should not 'reasonable doubt' be taken more seriously when a defendant's life is at stake?"); see also id. at 1000 n.94 (citing state court cases formally imposing a heightened standard for sufficiency of evidence in capital cases).

(178.) Supra notes 2-3 and accompanying text.

(179.) Bridges v. Wixon, 326 U.S. 135, 165 (1945) (Murphy, J" concurring) (noting the "grave consequences" of deportation).

(180.) See infra Part III. For cases not governed by some such doctrine, courts are probably obliged to act on their best guess as to what the statute at issue means. See, e.g., Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 984 (2005) (contrasting cases in which courts hold that a statute's "best reading" is X with those in which they hold that a statute's "only permissible reading" is X).

(181.) See infra Part III. Otherwise the thing to do under conditions of uncertainty would be for a court to act on its "best guess" as to what the statute means--that is, to decide the case in accordance with the statute's most natural reading. See supra note 175. Presumably, this is what courts must do under conditions of uncertainty absent some doctrine that authorizes case resolution on some nonlinguistic basis--this assumes the object of interpretation (e.g., apparent communicative content) is the same in cases in which statutory meaning is less than "clear."

(182.) See Katyal 8< Schmidt, supra note 9, at 2119-22 (questioning the assumption that erroneous invalidation is always more costly than erroneous interpretation).

(183.) See Baude 8< Sachs, supra note 7, at 1083 ("[When statutory language is uncertain], we don't keep fruitlessly hunting for a hidden meaning; but neither do we tell judges to fill the gap with whatever they think best. Instead, we use law to displace our ordinary inquiries about meaning.").

(184.) Blodgett v. Holden, 275 U.S. 142, 147-48 (1927) (Holmes, J., concurring), modified, 276 U.S. 594 (1928) (per curiam).

(185.) Rust v. Sullivan, 500 U.S. 173, 191 (1991); see also Rostker v. Goldberg, 453 U.S. 57, 64 (1981) ("The Congress is a coequal branch of government whose Members take the same oath we do to uphold the Constitution of the United States.... [W]e must have 'due regard to

the fact that this Court is not exercising a primary judgment but is sitting in judgment upon those who also have taken the oath to observe the Constitution and who have the responsibility for carrying on government.'" (quoting Joint Anti-Fascist Refugee Comm. v. McGrath, 341 U.S. 123, 164 (1951) (Frankfurter, J" concurring))).

(186.) Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 575 (1988); see also Alexander M. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics 16 (Yale Univ. Press 2d ed. 1986) (1962) (discussing the "counter-majoritarian difficulty" involved in judicial review).

(187.) See supra notes 36-38 and accompanying text.

(188.) Ashwander v. Tenn. Valley Auth., 297 U.S. 288, 348 (1936) (Brandeis, J., concurring) (quoting Crowell v. Benson, 285 U.S. 22, 62 (1932)).

(189.) 22 U.S.C. [section][section] 6701-6771 (2012).

(190.) Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and on Their Destruction, Apr. 29, 1997, 1974 U.N.T.S. 317; S. Res. 75, 105th Cong. (1997).

(191.) 18 U.S.C. [section] 229(a)(1) (2012).

(192.) Id. [section] 229F(1)(A), (7)(A) (exempting the use of chemicals for "[a]ny peaceful purpose related to an industrial, agricultural, research, medical, or pharmaceutical activity or other activity"). The Act contains other exemptions not relevant here. See, e.g., id. [section] 229F(7)(B)-(D).

(193.) Id. [section] 229F(8)(A).

(194.) Bond v. United States, 134 S. Ct. 2077, 2085-86 (2014).

(195.) Id. at 2085.

(196.) Id. (noting that it was "undisputed" that the defendant "did not intend to kill" her friend).

(197.) Id.

(198.) Id. at 2086.

(199.) Id.

(200.) Id. at 2094 (Scalia, J., concurring in the judgment).

(201.) See id. at 2087 (majority opinion).

(202.) Brief for Petitioner at 2, 16-19, Bond, 134 S. Ct. 2077 (No. 12-158), 2013 WL 1963862, at *2, *16-19.

(203.) 252 U.S. 416 (1920).

(204.) Holland, 252 U.S. at 432.

(205.) Bond, 134 S. Ct. at 2090.

(206.) Id.

(207.) Id. at 2083, 2087.

(208.) Id. at 2093.

(209.) Id. at 2090-91.

(210.) See id. at 2097 (Scalia, J., concurring in the judgment) ("Thanks to the Court's revisions, the Act, which before was merely broad, is now broad and unintelligible. '[N]o standard of conduct is specified at all.' " (alteration in original) (quoting Coates v. Cincinnati, 402 U.S. 611, 614 (1971))); see also infra notes 219-222 and accompanying text.

(211.) Bond, 134 S. Ct. at 2083.

(212.) Id. at 2096 (Scalia, J., concurring in the judgment) ("Imagine what future courts can do with that judge-empowering principle: Whatever has improbably broad, deeply serious, and apparently unnecessary consequences ... is ambiguous!").

(213.) Id.

(214.) Id. at 2090 (majority opinion).

(215.) Id. at 2087-93.

(216.) Id. at 2088.

(217.) Rowland v. Cal. Men's Colony, Unit II Men's Advisory Council, 506 U.S. 194, 196 (1993) (holding that only natural "person[s]" may qualify for treatment in forma pauperis under the federal statute).

(218.) Small v. United States, 544 U.S. 385, 387 (2005) (holding that a statute making it unlawful for a person "convicted in any court of a crime punishable by imprisonment for a term exceeding one year" to possess a firearm refers only to domestic convictions (quoting 18 U.S.C. [section] 922(g)(1)(a) (2012))).

(219.) Ironically, the defendant in Bond did suggest one such class when she argued that her conduct: fit within the Act's "peaceful purpose" exemption. According to the defendant, "peaceful" as used in the Act means something like not "warlike." Bond, 134 S. Ct. at 2094 n.2 (Scalia, J., concurring in the judgment). Because her conduct was clearly not "warlike," the defendant concluded, it thus fit within the "peaceful purpose" exception. Brief for Petitioner, supra note 202, at 50-57. As Justice Scalia observed in his concurrence, that reading fails for the reason that it renders superfluous the Act's specific exemption for the use of "self-defense devices" such as "pepper spray," 18 U.S.C. [section] 229C (2012), "the prosaic uses of which are surely nonwarlike." Bond, 134 S. Ct. at 2094 n.2 (Scalia, J., concurring in the judgment).

(220.) A distinguishing feature of statutory interpretation in the textualist era is that a court is expected to supply a plausible linguistic story in support of any reading it adopts. This contrasts with an earlier era in which providing a plausible policy story would suffice. See Doerfler, supra note 6, at 828-29. In recent years, Abbe Gluck in particular has praised courts for their renewed commitment to articulating policy stories in support of their readings. Abbe R. Gluck, The Supreme Court, 2014 Term--Comment: Imperfect Statutes, Imperfect Courts: Understanding Congress's Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62, 87-90 (2015) (praising the Court for its express attention to Congress's "legislative plans"). As I have argued elsewhere, any defensible version of statutory analysis is attendant to Congress's apparent practical ends. See Doerfler, supra note 96, at 995-98. Be that as it may, coming to accept readings of statutes on the basis of a plausible policy story alone (i.e., absent a plausible supporting linguistic story) would mark a dramatic, unfortunate shift in our interpretive practice. See supra notes 2-6 and accompanying text.

(221.) Along with an improbable hypothetical. See Bond, 134 S. Ct. at 2091 ("Any parent would be guilty of a serious federal offense--possession of a chemical weapon-- when, exasperated by the children's repeated failure to clean the goldfish tank, he considers poisoning the fish with a few drops of vinegar.").

(222.) See Doerfler, supra note 6, at 831-32 (analogizing Bond to the old Holy Trinity).

(223.) See Richard L. Hasen, Constitutional Avoidance and Anti-Avoidance by the Roberts Court, 2009 Sup. Ct. Rev. 181, 201-02 ("[V]oting rights experts believed that the statutory ... argument had no chance.... Instead, it seemed unavoidable that the Court would address the constitutionality of the [statute]."); see also Nw. Austin Mun. Util. Dist. No. One v. Holder, 557 U.S. 193, 197 (2009) ("Th[e] constitutional question has attracted ardent briefs from dozens of interested parties, but the importance of the question does not justify our rushing to decide it.").

(224.) Nw. Austin, 557 U.S. at 200-01.

(225.) 52 U.S.C. [section] 10304(a) (Supp. III 2016).

(226.) Id. [section] 10303(b), invalidated by Shelby County v. Holder, 133 S. Ct. 2612 (2013).

(227.) Id. [section] 10303(a).

(228.) Nw. Austin, 557 U.S. at 200-01.

(229.) Id. at 206.

(230.) Id. (quoting Political Subdivision, Black's Law Dictionary (8th ed. 2004)).

(231.) Nw. Austin Mun. Util. Dist. No. One v. Mukasey, 573 F. Supp. 2d 221, 230- 35 (D.D.C. 2008), rev'd sub nom. Nw. Austin, 557 U.S. 193.

(232.) 52 U.S.C. [section] 10310(c)(2).

(233.) See Nw. Austin, 557 U.S. at 200, 206.

(234.) Id. at 197.

(235.) Id. at 203 (quoting United States v. Louisiana, 363 U.S. 1, 16 (I960)).

(236.) Id.

(237.) Id.

(238.) Id.

(239.) Id. at 208.

(240.) Id. at 207-09; see also Shelby County v. Holder, 133 S. Ct. 2612 (2013) (invalidating section 4(b)'s coverage formula on equal-sovereignty grounds).

(241.) 435 U.S. 110 (1978).

(242.) Nw. Austin, 557 U.S. at 207 (citing United States v. Bd. of Comm'rs, 435 U.S. 110, 128-29, 130 n.18 (1978)).

(243.) See Nw. Austin Mun. Util. Dist. No. One v. Mukasey, 573 F. Supp. 2d 221, 230-35 (D.D.C. 2008), rev'd sub nom. Nw. Austin, 557 U.S. 193.

(244.) 52 U.S.C. [section] 10303(a)(1) (Supp. III 2016).

(245.) See Nw. Austin, 573 F. Supp. 2d at 231.

(246.) Prior to amendment in 1982, only "subdivisions" specifically determined to be "covered" were eligible to request bailout relief. Id.; see also City of Rome v. United States, 446 U.S. 156, 167 (1980) (holding that a city within a covered state was ineligible for bailout relief because section 4(b) had never been applied to it).

(247.) See Nw. Austin, 573 F. Supp. 2d at 232 ("Had Congress stopped at the comma, there might be some question as to whether it intended to use the term 'political subdivision' in its broadest sense. But Congress did not stop at the comma.").

(248.) See supra Part II.

(249.) Grice, supra note 167, at 45-46 ("[O]verinformativeness may be confusing in that it is liable to raise side issues; and there may also be an indirect effect, in that the hearers may be misled as a result of thinking that there is some particular point in the provision of the excess of information.").

(250.) To receive presidential-primary matching funds, a "candidate" must, among other things, raise in excess of "$5,000 in contributions from residents of each of at least 20 States." I.R.C. [section] 9033(b)(3) (2012).

(251.) See U.S. Const, art. II, [section] 1, cl. 4 ("No Person except a natural born Citizen, or a Citizen of the United States, at the time of the Adoption of this Constitution, shall be eligible to the Office of President; neither shall any person be eligible to that Office who shall not have attained to the Age of thirty five Years, and been fourteen Years a Resident within the United States.").

(252.) See Nw. Austin, 573 F. Supp. 2d at 232 (reasoning that "[u]nder the District's interpretation, this language would be surplusage").

(253.) 446 U.S. 156, 167 (1980).

(254.) See Nw. Austin Mun. Util. Dist. No. One v. Holder, 557 U.S. 193, 209 (2009) ("In 1982, however, Congress expressly repudiated City of Rome and instead embraced 'piecemeal' bailout."). Rick Hasen argues forcefully that the legislative history surrounding the 1982 amendment cuts against this reading. See Hasen, supra note 223, at 205 ("There was no 'express repudiation' of City of Rome in the text of the 1982 renewal. Indeed, City of Rome is not mentioned in the Senate Report as being repudiated."). But because the relevance of legislative history to statutory meaning is sharply contested it seems dubious to suggest that appeal to legislative history can render a reading not "fairly available." See, e.g., John F. Manning, Textualism and the Equity of the Statute, 101 Colum. L. Rev. 1, 124 (2001) ("[T]extualists tend to be stricter in their application of clear statement rules, because they require the expression of clear intent to be derived from the text of the statute, rather than in the legislative history.").

(255.) See supra notes 241-256 and accompanying text.

(256.) See Atl. Cleaners & Dyers, Inc. v. United States, 286 U.S. 427, 433 (1932) ("It is not unusual for the same word to be used with different meanings in the same act, and there is no rule of statutory construction which precludes the courts from giving to the word the meaning which the Legislature intended it should have in each instance.").

(257.) See, e.g., West Virginia v. EPA, 136 S. Ct. 1000 (2016) (involving a challenge to implementation of the Clean Air Act); King v. Burwell, 135 S. Ct. 2480 (2015) (involving a challenge to the implementation of the Patient Protection and Affordable Care Act).

(258.) See, e.g., sources cited supra note 11.

(259.) 135 S. Ct. at 2487-88.

(260.) See Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119 (2010) (codified as amended in scattered sections of 26 and 42 U.S.C.).

(261.) See King, 135 S. Ct. at 2485 (characterizing these reforms as "interlocking").

(262.) 42 U.S.C. [section] 300gg (2012).

(263.) I.R.C. [section] 5000A (2012).

(264.) Id. [section] 36B.

(265.) Jonathan Gruber, Health Care Reform Is a "Three-Legged Stool": The Costs of Partially Repealing the Affordable Care Act, Ctr. for Am. Progress (Aug. 5, 2010, 9:00 AM), https:// reform-is-athree-legged-stool/ [].

(266.) King, 135 S. Ct. at 2485.

(267.) 42 U.S.C. [section] 18031(b)(1) (2012); see also id. [section] 18024(d) (defining "State" to mean "each of the 50 States and the District of Columbia").

(268.) I.R.C. [section] 36B(b)(2)(A).

(269.) See Printz v. United States, 521 U.S. 898, 935 (1997).

(270.) 42 U.S.C. [section] 18041(c)(1).

(271.) See, e.g., Jonathan H. Adler & Michael F. Cannon, Taxation Without Representation: The Illegal IRS Rule to Expand Tax Credits Under the PPACA, 23 Health Matrix 119, 123 (2013).

(272.) Health Insurance Premium Tax Credit, 77 Fed. Reg. 30,377 (May 23, 2012) (codified at 26 C.F.R. [section] 1.36B-2; 45 C.F.R. [section] 155.20).

(273.) King v. Burwell, 135 S. Ct. 2480, 2488 (2015).

(274.) Id. at 2487.

(275.) See id. at 2489, 2493.

(276.) Id. ("The combination of no tax credits and an ineffective coverage requirement could well push a State's individual insurance market into a death spiral.").

(277.) See Doerfler, supra note 6, at 814 (explaining that section 36B likely contains a simple scrivener's error, but that the error is less than "absolutely clear," as required for recognition under current doctrine); Gluck, supra note 220, at 67 (observing that King "says nothing explicit about what the Court is to do when there is a statutory mistake--the enormous elephant that neither party dared mention throughout the litigation" (emphasis omitted)).

(278.) Brief for the Respondents at 13, King, 135 S. Ct. 2480 (No. 14-114), 2015 WL 349885, at *13.

(279.) King, 135 S. Ct. at 2490-91.

(280.) Id. at 2490 (quoting 42 U.S.C. [section][section] 18031(d)(2)(A), 18032(f)(1)(A) (2012)).

(281.) Id. at 2492.

(282.) Id. at 2489 (quoting 42 U.S.C. [section] 18041(c)(1) (2012)).

(283.) Id.

(284.) See Doerfler, supra note 6, at 848 ("Suppose ... that Ann instructs Beth to purchase a blueberry pie from Hi-Rise Bakery but instructs Carl to purchase 'such pie' from Petsi Pies if Beth fails. If Carl goes on to purchase 'such pie,' is that pie 'purchased by Beth'? Is it also 'from Hi-Rise'? Doubtful.").

(285.) See id. at 846-47 (arguing that these anomalies indicate that various provisions of the Act were drafted on the assumption that it would provide only for state-run exchanges, but that later it was amended, albeit incompletely, to provide for federally facilitated exchanges as a fallback).

(286.) Interestingly, the Court in King did, in some sense, try to make section 36B's reference to "the State" disappear, attributing that language to "inartful drafting." 135 S. Ct. at 2483. As Justice Scalia explained in dissent, however, unless the inartful drafting in question amounted to a scrivener's error--a claim rendered unavailable by current doctrine--it is unclear how that attribution supports the majority's reading. See id. at 2504 (Scalia, J., dissenting) ("Perhaps sensing the dismal failure of its efforts to show that 'established by the State' means 'established by the State or the Federal Government,' the Court tries to palm off the pertinent statutory phrase as 'inartful drafting.' This Court, however, has no free-floating power 'to rescue Congress from its drafting errors.'" (citation omitted) (quoting Lamie v. U.S. Tr., 540 U.S. 526, 542 (2004))). The government, for its part, addressed this language directly, arguing that the secretary acts as a state's "statutory surrogate" in establishing an exchange, and so that an exchange established by the secretary just is an "Exchange established by the State." Brief for the Respondents, supra note 278, at 13.

(287.) King, 135 S. Ct. at 2492-93.

(288.) As indicated above, King most likely involved a scrivener's error rendered unspeakable by current doctrine. See supra note 277. For that reason, it is likely that the opinion in King constituted an attempt to work around that doctrinal impediment. See Doerfler, supra note 6, at 849.

(289.) See Doerfler, supra note 6, at 847-49 (discussing the weaknesses of the government's argument).

(290.) See, e.g., King, 135 S. Ct. at 2495-96 (noting the "calamitous result" to which the challenger's interpretation would give rise).

(291.) Tex. Dep't of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., 135 S. Ct. 2507, 2513-14 (2015).

(292.) Id. at 2514.

(293.) 42 U.S.C. [section][section] 3601-3631 (2012).

(294.) See Inclusive Cmtys., 135 S. Ct. at 2516.

(295.) Id. at 2513.

(296.) Id.; see also Ricci v. DeStefano, 557 U.S. 557, 577 (2009).

(297.) Inclusive Cmtys., 135 S. Ct. at 2513.

(298.) See id. at 2537 (Alito, J., dissenting).

(299.) 42 U.S.C. [section] 3604(a) (2012) (emphasis added).

(300.) Id. [section] 3605(a) (emphasis added).

(301.) Inclusive Cmtys., 135 S. Ct. at 2533 (Alito, J., dissenting) (alteration in original) (quoting Univ. of Tex. Sw. Med. Ctr. v. Nassar, 133 S. Ct. 2517, 2527 (2013)).

(302.) Id. at 2535.

(303.) 18 U.S.C. [section] 249 (2012); see also Wisconsin v. Mitchell, 508 U.S. 476, 484-85 (1993).

(304.) Inclusive Cmtys., 135 S. Ct. at 2519.

(305.) Id. at 2520.

(306.) 401 U.S. 424 (1971).

(307.) See Griggs, 401 U.S. at 429-31.

(308.) Id.

(309.) See Inclusive Cmtys., 135 S. Ct. at 2544 (Alito, J., dissenting).

(310.) 544 U.S. 228 (2005).

(311.) Smith, 544 U.S. at 233-40.

(312.) Id. at 235.

(313.) Id. at 235-36 (quoting 42 U.S.C. [section] 2000e-2(a)(2) (2012)).

(314.) Id. at 236 n.6 (emphasis added) (quoting 42 U.S.C. [section] 2000e- 2(a)(2)).

(315.) 42 U.S.C. [section] 3604(a) (2012) (emphasis added).

(316.) Tex. Dep't of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., 135 S. Ct. 2507, 2518 (2015) ("[T]he logic of Griggs and Smith provides strong support for the conclusion that the FHA encompasses disparate-impact claims.").

(317.) Or, in more colloquial terms, to "gut" the FHA. See, e.g., Nikole Hannah- Jones, Supreme Court's Latest Race Case: Housing Discrimination, ProPublica (Jan. 21, 2015, 12:18 PM), discrimination [] (remarking that "[m]any fear [the] Texas case could gut the landmark Fair Housing Act").

(318.) See William Baude & Ryan D. Doerfler, Arguing with Friends 4 (Univ. of Pa. Law Sch. Pub. Law Working Paper No. 630, 2017), pers.cfm?abstract_id=2985032 []; see also Eric A. Posner & Adrian Vermeule, The Votes of Other Judges, 105 Geo. L.J. 159, 177-179 (2016).

(319.) E.g., Liparota v. United States, 471 U.S. 419, 427 (1985) (noting the Court's "longstanding recognition of the principle that 'ambiguity concerning the ambit of criminal statutes should be resolved in favor of lenity.'" (quoting Rewis v. United States, 401 U.S. 808, 812 (1971))); United States v. Bass, 404 U.S. 336, 347-48 (1971) ("In various ways over the years, we have stated that 'when choice has to be made between two readings of what conduct Congress has made a crime, it is appropriate, before we choose the harsher alternative, to require that Congress should have spoken in language that is clear and definite.'" (quoting United States v. Universal C.I.T. Credit Corp., 344 U.S. 218, 221-22 (1952))); Comm'r v. Acker, 361 U.S. 87, 91 (1959) ("The law is settled that 'penal statutes are to be construed strictly,' and that one 'is not to be subjected to a penalty unless the words of the statute plainly impose it.'" (citations omitted) (first quoting FCC v. Am. Broad. Co., 347 U.S. 284, 296 (1954); and then quoting Keppel v. Tiffing Sav. Bank, 197 U.S. 356, 365 (1905))).

(320.) Clark v. Martinez, 543 U.S. 371, 397 (2005) (Thomas, J., dissenting); accord United States v. Cardiff, 344 U.S. 174, 176 (1952); McBoyle v. United States, 283 U.S. 25, 27 (1931).

(321.) Bass, 404 U.S. at 348 ("[B]ecause of the seriousness of criminal penalties, and because criminal punishment usually represents the moral condemnation of the community, legislatures and not courts should define criminal activity."); accord Babbitt v. Sweet Home Chapter of Cmtys. for a Great Or., 515 U.S. 687, 704 n.18 (1995).

(322.) Muscarello v. United States, 524 U.S. 125, 138-39 (1998) (emphases added) (citations omitted) (first quoting United States v. Wells, 519 U.S. 482, 499 (1997); and then quoting Staples v. United States, 511 U.S. 600, 619 n.17 (1994)).

(323.) Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 983-84 (2005) (contrasting decisions that expresses a court's "opinion as to the best reading of an ambiguous statute" with those that identify a "statute's dear meaning" (quoting Maislin Indus., U.S., Inc. v. Primary Steel, Inc., 497 U.S. 116, 131 (1990))).

(324.) 136 S. Ct. 958 (2016).

(325.) 18 U.S.C. [section] 2252(b)(2) (2012) (emphasis added).

(326.) See Lockhart, 136 S. Ct. at 961-62.

(327.) Id.

(328.) Id. at 962.

(329.) Id.

(330.) Id. at 962-63 (quoting Barnhart v. Thomas, 540 U.S. 20, 26 (2003)); see also id. at 963 ("The rule reflects the basic intuition that when a modifier appears at the end of a list, it is easier to apply that modifier only to the item directly before it. That is particularly true where it takes more than a little mental energy to process the individual entries in the list, making it a heavy lift to carry the modifier across them all.").

(331.) Id. at 964.

(332.) Id.

(333.) Id. But see id. ("We cannot state with certainty that Congress used [the federal chapter] as a template for the list of state predicates set out in [the enhancement provision], but we cannot ignore the parallel...." (emphasis added)).

(334.) Id. at 968 ("We will not apply the rule of lenity to override a sensible grammatical principle buttressed by the statute's text and structure.").

(335.) Id. at 969, 972 n.2 (Kagan, J., dissenting) (quoting James v. City of Boise, 136 S. Ct. 685, 686-87 (2016) (per curiam)) (collecting cases); see also id. at 969 ("Suppose a real estate agent promised to find a client 'a house, condo, or apartment in New York.' Wouldn't the potential buyer be annoyed if the agent sent him information about condos in Maryland or California?").

(336.) Id. at 970.

(337.) Id. (quoting Scalia 8; Garner, supra note 4, at 147).

(338.) Id. Justice Kagan also questioned Justice Sotomayor's analogy between the language at issue and the chapter of the federal criminal code pertaining to "Sexual Abuse," observing that, unlike the state-law predicate offense language, the federal chapter divides "sexual abuse" into four categories: "(aggravated sexual abuse," "[s]exual abuse," "[s]exual abuse of a ward or minor," and "[a]busive sexual contact." Id. at 975-76 (quoting 18 U.S.C. [section][section] 2241-2244 (2012)).

(339.) Frank H. Easterbrook, The Absence of Method in Statutory Interpretation, 84 U. Chi. L. Rev. 81, 86-90 (2017).

(340.) See Hrafn Asgeirsson, On the Possibility of Non-Literal Legislative Speech, in Pragmatics and Law: Theoretical and Practical Perspectives (A. Capone & F. Poggi eds., forthcoming 2017) (manuscript at 1, 20-21) (on file with author) (arguing that, generally speaking, the practical stakes of criminal law cases are "fairly high").

(341.) Missouri v. Hunter, 459 U.S. 359, 373 (1983) (Marshall, J., dissenting).

(342.) See generally Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Colorblindness (2010); Elizabeth Hinton, From the War on Poverty to the War on Crime: The Making of Mass Incarceration in America (2016).

(343.) Surely some criminal cases are higher stakes than others (e.g., death- penalty cases). In turn, one would expect courts to apply the rule of lenity more or less aggressively depending upon the stakes of the particular case.

(344.) Cf Adam M. Samaha, Low Stakes and Constitutional Interpretation, 13 U. Pa. J. Const. L. 305, 320 (2010) (reminding that a "low-stakes decision" is not "zero- stakes decision").

(345.) Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842-43 (1984) ("[T]he court, as well as the agency, must give effect to the unambiguously expressed intent of Congress."); see also Matthew C. Stephenson & Adrian Vermeule, Chevron Has Only One Step, 95 Va. L. Rev. 597, 601 (2009) (explaining that an agency's interpretation is "reasonable" so long as it is among "the set of interpretations which the statute does not clearly prohibit").

(346.) Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 980 (2005).

(347.) See Chevron, 467 U.S. at 865-66 (observing that judges "have no constituency" and "have a duty to respect legitimate policy choices made by those who do").

(348.) See Michigan v. EPA, 135 S. Ct. 2699, 2726 (2015) (Kagan, J., dissenting) ("Far more than courts, agencies have the expertise and experience necessary to design regulatory processes suited to 'a technical and complex arena.'" (quoting Chevron, 467 U.S. at 863)).

(349.) City of Arlington v. FCC, 133 S. Ct. 1863, 1873 (2013).

(350.) See Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 516-17. But see Doerfler, supra note 96, at 1022-31 (arguing that all attributions of legislative intent are fictional in character, but that such attributions are valuable as apt or inapt nonetheless).

(351.) See Chevron, 467 U.S. at 843-44.

(352.) See Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2148 (2016) (Thomas, J., concurring); Michigan v. EPA, 135 S. Ct. at 2712 (Thomas, J., concurring); Dep't of Transp. v. Ass'n of Am. R.Rs., 135 S. Ct. 1225, 1241-45 (2015) (Thomas, J., concurring in the judgment); Perez v. Mortg. Bankers Ass'n, 135 S. Ct. 1199, 1217-20 (2015) (Thomas, J., concurring in the judgment) (discussing specifically deference to agency interpretations of their own regulations, but raising concerns with deference to agency interpretations generally).

(353.) Perez, 135 S. Ct. at 1217-18 (Thomas, J., concurring in the judgment) (citing Philip Hamburger, Law and Judicial Duty 507-08 (2008)); see also Philip Hamburger, Is Administrative Law Unlawful? (2014).

(354.) Perez, 135 S. Ct. at 1217 (Thomas, J., concurring in the judgment).

(355.) Gutierrez-Brizuela v. Lynch, 834 F.3d 1142, 1152 (10th Cir. 2016) (Gorsuch, J., concurring).

(356.) Id.

(357.) Cf. Cass R. Sunstein & Adrian Vermeule, The Unbearable Rightness of Auer, 84 U. Chi. L. Rev. 297, 298-99 (2017) (criticizing the deployment of "heavy constitutional artillery" against the doctrine that courts should defer to reasonable agency constructions of unclear regulations).

(358.) 5 U.S.C. [section] 706 (2012).

(359.) See Perez, 135 S. Ct. at 1211 (Scalia, J., concurring in the judgment) (quoting 5 U.S.C. [section] 706).

(360.) Id. at 1212.

(361.) Aditya Bamzai, The Origins of Judicial Deference to Executive Interpretation, 126 Yale L.J. 908 (2016). Bamzai urges that the "most natural reading" of the APA is that it restores what he terms the "traditional" approach to deference to agencies. Id. at 987. On that approach, courts "defer" to an agency's construction only if "contemporaneous" with the enactment of the statute at issue, or reflective of the "customary" interpretation thereof. Id. at 916. As Bamzai points out, neither of these grounds for deferring to executive agencies is "because they [a]re executive as such." Id. at 943.

(362.) For a forceful rejoinder, see, for example, Cass R. Sunstein & Adrian Vermeule, The New Coke: On the Plural Aims of Administrative Law, 2015 Sup. Ct. Rev. 41.

(363.) See Thomas W. Merrill & Kristin E. Hickman, Chevron's Domain, 89 Geo. L.J. 833, 836 (2001); Cass R. Sunstein, Chevron Step Zero, 92 Va. L. Rev. 187 (2006).

(364.) See, e.g., Scalia, supra note 350, at 511 (characterizing Chevron as "announc[ing] the principle that the courts will accept an agency's reasonable interpretation of the ambiguous terms of a statute that the agency administers").

(365.) Courts have never gone so far as to adopt Justice Breyer's recommended approach of determining Chevron s applicability case by case, at least not explicitly. See Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 373, 377, 381 (1986); Scalia, supra note 350, at 516 (characterizing pte-Chevron as adhering to a case-by-case approach to deference).

(366.) City of Arlington v. FCC, 133 S. Ct. 1863 (2013).

(367.) King v. Burwell, 135 S. Ct. 2480, 2489 (2015) (quoting Util. Air Regulatory Grp. v. EPA, 134 S. Ct. 2427, 2444 (2014)).

(368.) See id. ("[H]ad Congress wished to assign that question to an agency, it surely would have done so expressly.").

(369.) And, perhaps, by constitutional considerations of their own. See, e.g., Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 866 (1984) ("The responsibilities for assessing the wisdom of such policy choices and resolving the struggle between competing views of the public interest are not judicial ones: 'Our Constitution vests such responsibilities in the political branches.'" (quoting Tenn. Valley Auth. v. Hill, 437 U.S. 153, 195 (1978))).

(370.) Id. at 842.

(371.) See, e.g., Jeffrey A. Pojanowski, Neoclassical Administrative Common Law, New Rambler (Sept. 26, 2016), administrative-common-law [] (reviewing John Dickinson, Administrative Justice and the Supremacy op Law in the United States (2004)).

(372.) See, e.g., Adrian Vermeule, Judging Under Uncertainty: An Institutional Theory of Legal Interpretation 217 (2006) (arguing that the best "justification for Chevron is simply that judicial deference to agencies produces better consequences than nondeference"); Cass R. Sunstein, Beyond Marbury: The Executive's Power to Say What the Law Is, 115 Yale L.J. 2580, 2610 (2006) ("Chevron is best taken as a vindication of the realist claim that resolution of statutory ambiguities often calls for judgments of policy and principle.... The meaning of statutory enactments is no brooding omnipresence in the sky. Chevron is our Erie, and much of the time, it is emphatically the province and duty of the executive branch to say what the law is."). Indeed, one way to characterize the spectrum of support for deference to agencies is according to the relative concern assigned to those two prospects.

(373.) See King, 135 S. Ct. at 2489 (holding that the Chevron framework does not apply to questions of deep economic and political significance); City of Arlington v. FCC, 133 S. Ct. 1863, 1877-86 (2013) (Roberts, C.J., dissenting) (arguing that the Chevron framework should not apply to "jurisdictional" constructions).

(374.) Compare Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2148 (2016) (Thomas, J., concurring) (calling for Chevron to be reconsidered), with Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs., 545 U.S. 967, 982-86 (2005) (holding that a reasonable agency construction of a statute it administers supersedes any prior judicial construction).

(375.) See City of Arlington, 133 S. Ct. at 1874-75 (holding that the Chevron framework applies to "jurisdictional" constructions).

(376.) E.g., Brand X, 545 U.S. at 982-83; see also Merrill & Hickman, supra note 363, at 834, 836 (characterizing Chevron as establishing a "two-step deference regime"). But see Stephenson & Vermeule, supra note 345, at 597 (arguing that Chevron s two "steps" logically reduce to one).

(377.) Scalia, supra note 350, at 521.

(378.) Id. (contrasting "strict constructionalist[s]" with those who "abhor[] a 'plain meaning' rule, and [are] willing to permit the apparent meaning of a statute to be impeached by the legislative history"). In effect, Justice Scalia's claim was that support for Chevron is inversely proportional to the frequency with which one thinks the doctrine would make a difference. Empirical studies provide moderate support for Justice Scalia's observation. See Thomas J. Miles & Cass R. Sunstein, Do Judges Make Regulatory Policy? An Empirical Investigation of Chevron, 73 U. Chi. L. Rev. 823, 823, 828-29 (2006) (finding that "conservative" justices vote to validate agency decisions less often than "liberal" justices, noting that "as an empirical matter, the more conservative justices (Justices Antonin Scalia and Clarence Thomas) have embraced 'plain meaning' approaches and the more liberal justices have not").

(379.) As an empirical matter, this question is complicated by, for example, Justice Thomas's shifting assessment of Chevron. See supra notes 356-359 and accompanying text. Although Justice Scalia and Justice Thomas were consistently similar in terms of interpretive methodology, whether or at what point their respective attitudes towards Chevron diverged is unclear, making a meaningful side-by-side comparison difficult if not impossible.

(380.) See, e.g., John F. Manning, The Eleventh Amendment and the Reading of Precise Constitutional Texts, 113 Yale L.J. 1663, 1722-49 (2004) (discussing the disparity between the constitutional text and the Supreme Court's sovereign immunity jurisprudence); Strauss, supra note 44, at 3 (identifying various "anomalies"--that is, "outcomes that are inconsistent with established principles of constitutional Jaw[ ]that following the text of the Constitution would produce").

(381.) See Strauss, supra note 44, at 21 (observing that "it is not acceptable explicitly to ignore the text of the Constitution"); see also William Baude, Is Originalism Our Law?, 115 Colum. L. Rev. 2349 (2015) (arguing that courts rarely if ever "explicitly repudiate" the original meaning of constitutional provisions).

(382.) McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 407 (1819).

(383.) U.S. Const, amend. I (emphasis added).

(384.) See Hans v. Louisiana, 134 U.S. 1, 1 (1890).

(385.) U.S. Const, amend. XI (emphasis added).

(386.) Cf. Christopher R. Green, Constitutional Theory and the Activismometer: How to Think About Indeterminacy, Restraint, Vagueness, Executive Review, and Precedent, 54 Santa Clara L. Rev. 403, 430-32 (2014) (articulating a conception of "judicial activism" that reflects the "interest-sensitivity" of knowledge); see also Primus, supra note 43 (suggesting that the "textual gap" between statutory and constitutional interpretation is owed in part to the practical costs of adhering closely to text in "high stakes" cases).

(387.) See Akhil Reed Amar, America's Unwritten Constitution: The Precedents and Principles We Live By 34 (2012).

(388.) See Strauss, supra note 44, at 30-34 (observing, among other things, that "[t]he other provisions [of the Bill of Rights] refer to the 'right' of the people or of an individual, or they simply say that certain actions are not allowed" (footnotes omitted)).

(389.) King v. Burwell, 135 S. Ct. 2480, 2497 (2015) (Scalia, J., dissenting).

(390.) Id. at 8.

(391.) Id. at 4-5 (arguing that constitutional text is constitutional text is "expanded, limited, qualified, reconceived, relegated to the background, or all-but-ignored, depending on

(392.) To the extent that "interpretation" is an inherently linguistic task. But see Mitchell N. Berman & Kevin Toh, On What Distinguishes New Originalism from Old: A Jurisprudential Take, 82 Fordham L. Rev. 545, 546-50 (2013) (arguing that the phrase "constitutional interpretation" could refer either to the task of discerning the constitutional text's linguistic meaning or to that of determining the text's legal significance).

(393.) Seemingly, the heightened practical stakes of judicial review would push in favor of a Thayerian approach to constitutional interpretation. See generally James B. Thayer, The Origin and Scope of the American Doctrine of Constitutional Law, 7 Harv. L. Rev. 129 (1893) (recommending a highly deferential approach to judicial review).

(394.) See Bradley & Siegel, Constructed Constraint, supra note 16, at 1287 (observing the relative clarity of numeric constitutional provisions).

(395.) Daryl J. Levinson, Parchment and Politics: The Positive Puzzle of Constitutional Commitment, 124 Harv. L. Rev. 657, 709 (2011) ("The constitutional text is quite specific on many low-stakes issues, where agreement is more important to most political actors than achieving any particular outcome.").

(396.) As a corollary, this Article suggests we ought to have more modest expectations of Congress when drafting high-stakes legislation. Assuming that it is more "costly" for Congress to draft precise legislative text, see Colin S. Diver, The Optimal Precision of Administrative Rules, 93 Yale LJ. 65, 74-75, 103 (1983) (arguing that the costs associated with legislating precisely often leads legislatures to adopt "open-ended language"), an entailment of the observation here is that it is more difficult for Congress to speak "clearly" when drafting legislation that is inherently high stakes.

(397.) Nor is it to vindicate the various doctrines pursuant to which courts resolve cases in the absence of "clear" statutory meaning. See supra notes 182-185 and accompanying text.
COPYRIGHT 2018 Michigan Law Review Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Doerfler, Ryan D.
Publication:Michigan Law Review
Date:Feb 1, 2018
Previous Article:Opening the Gates of Cow Palace: Regulating Runoff Manure as a Hazardous Waste Under RCRA.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters