A Textualist Response to Two Texts: Positive-Law Codification and Interpreting Section 1983
abstract. What is a textualist to do when there are two texts, each with a colorable claim to legitimacy? This question arises every time the text of a codified statute that has been enacted into positive law differs from the language originally drafted by Congress. Such changes between the original text and positive-law codification are the work of unelected revisers (today, the Office of the Law Revision Counsel), and though these changes are technically voted upon by Congress, Congress’s review is cursory.
This Note proposes a path forward in the form of a novel canon the Note labels the “two texts canon.” As a textualist, formalist rule, the two texts canon begins at step one with the codified positive-law text and ends there where that text is clear, including where changes during codification have contributed to that clear meaning. However, when the codified positive-law text is ambiguous or silent on the interpretive question at issue, the two texts canon insists at step two that interpretations of such ambiguity or silence in the codified positive-law text may not contradict the original text.
The Note concludes with the case study of Section 1983 and qualified immunity. Because the positive-law text of what is now Section 1983 is silent as to immunities, step two of the two texts canon requires that the original text should be applied. Because the original text explicitly abrogates qualified immunity, the Note urges that qualified immunity should be abolished.
author. J.D. 2024, Yale Law School; B.A. 2021, Harvard College. I am deeply grateful to Abbe Gluck and the students of the Spring 2023 New Developments in Legislation and Statutory Interpretation class, as well as to Paul W. Hughes, Andrew A. Lyonsberg, Michael Kimberly, Sarah P. Hogarth, Charles A. Rothfeld, and the students of the 2023-2024 Supreme Court Advocacy Clinic for inspiring this work. For their comments and support, sincere thanks to Katherine Lempres and Judith Mae. I am also thankful for the brilliant insights and editing of Riley Gordon, Chase Hinman, Fiona Furnari, and the other members of the Yale Law Journal.
Introduction
What is a textualist to do when there are two texts, each with a colorable claim to legitimacy? This question arises every time the text of a codified statute that has been enacted into positive law differs from the language originally drafted by Congress. As one highly consequential example, scholars have recently rediscovered that the original text of 42 U.S.C. § 1983 (hereinafter Section 1983) specifically abrogated qualified immunity, despite the fact that the modern positive-law text, which is silent on the issue, has been read by the Supreme Court since 1967 to assume such immunity.1 Solving the puzzle of two texts thus has the potential to determine the correct interpretation of Section 1983, thereby shaping courts’ ability to remedy deprivations of rights.
Textualists have yet to offer an approach to the quandary of two texts. To do so, textualists will need to harmonize the seemingly conflicting results dictated by two core textualist values. Textualists espouse formalism,2 and codified positive-law text is formally “legal evidence of the law[]”—meaning the codified positive-law text is conclusive.3 So, should the codified positive-law text not govern?4 Yet textualists also espouse democratic values,5 like nondelegation of the legislative function,6 and codified text is the work of unelected bureaucrats.7 Then, is it not imperative that the original text be defended? Although formalism and democracy are not inherently in tension within the textualist framework, they seem at first glance to push for diverging resolutions of the dilemma of two texts. This Note is a first attempt to propose a path forward in the form of a novel canon of statutory interpretation that I label the “two texts canon.”
The problem of two texts is not just an academic puzzle. There is genuine disarray among the lower courts as to which text governs. The Fourth Circuit has reached something of a compromise that attempts to balance respect for the original text with respect for the codified positive-law text by searching for clarity in both:
Even if there is a conflict between the original Congressional enactment contained in the Statutes at Large and a codification that has been enacted into positive law, the Statutes at Large control when (1) the meaning of the original enactment was “clear and quite different from the meaning . . . ascribe[d] to the codified law,” and (2) “the revisers expressly stated that changes in language resulting from the codification were to have no substantive effect.”8
The Ninth and Second Circuits share the Fourth Circuit’s intuition that clear meaning in the original text or clear statements of changed meaning in the codified text are particularly significant, though neither circuit has stated a generalizable approach to the problem of two texts.9
By contrast, while claiming to apply the Fourth Circuit’s rule, the District Court for the District of Columbia has embraced a less nuanced position: “Where there is a discrepancy between the language in the United States Code and the Statutes at Large, the language in the Statutes at Large controls.”10 Such an unwavering preference for the Statutes at Large should be untenable. Codification statutes not only declare enacted titles of the Code to be positive law, but also repeal and replace the corresponding provisions of the Statutes at Large.11 But even when they do not go so far as to assert that the Statutes at Large control over positive law, courts frequently side with the original text where conflicts are presented, claiming that changes in substantive meaning through codification are unlikely.12
At the other extreme, lower courts across multiple circuits routinely adhere to the codified text, either intentionally ignoring the original text or interpreting changes as substantively meaningful.13 Several courts cite the rule from Sutherland’s treatise on statutory construction: “[T]he text of a code section in an enacted title can be taken as authoritative and need not be checked or verified with the corresponding section in the original Statutes at Large . . . .”14 This refusal to crack open the Statutes at Large in deference to the Code’s superior formal legal status is also problematic, as it deprives the reader of powerful and potentially clear evidence of statutory meaning in cases where codified text is ambiguous and must be interpreted using other tools of construction.
The range of unsatisfactory approaches to the problem of two texts is particularly concerning given the frequency with which courts are called upon to interpret statutes contained in positive-law titles of the U.S. Code.15 Nearly half of all federal statutory law is codified positive law, and the text of a statute in the Code almost always differs in some way from the text originally enacted by Congress and recorded in the Statutes at Large.16 Thus, even as courts frequently confront the problem of two texts directly, reaching contradictory results,17 the issue more frequently is presented but left unaddressed.18 And although the problem is significant enough to have featured in dozens of Supreme Court opinions since 1884,19 the Supreme Court has not squarely confronted the problem of two texts since 198920—well before textualism became the dominant mode of statutory interpretation.21 Existing Supreme Court decisions have stated a rule that is incompatible with textualism and with statutes governing codification because it favors the original text, unless Congress has clearly expressed an intent to change the meaning of that text.22 While some lower courts continue to cite these cases (notably, one Eleventh Circuit opinion concluded the Court’s precedents establish a “recodification canon”),23 these cases are often ignored, particularly where lower courts follow the codified positive-law text.24 Updated guidance is urgently needed.
Resolving the problem of two texts could also provide the Supreme Court with the most promising legal hook for abolishing qualified immunity. Despite a slew of academic refutations of the atextual and ahistorical common-law and policy justifications for qualified immunity,25 the Court has not responded, notably denying certiorari in Section 1983 cases proposing to abolish the doctrine like Baxter v. Bracey in 2020.26 This may very well be because, notwithstanding the possibility of an unlikely-bedfellows coalition of conservative textualists and liberal supporters of civil-rights litigation,27 the Justices are bound by stare decisis.28 The original text of Section 1983 could burst through this barrier for reasons that will be explained in Part III. However, scholars and practitioners persist in framing the original text merely as evidence of congressional intent29—a strategy unlikely to persuade the textualist Court. Only by understanding the proper textualist response to two texts will it finally become clear that not only was there no satisfying textual justification for qualified immunity in 1967, but also the original text requires that the doctrine be abolished today.
In Part I, this Note begins by explaining the phenomenon of codification. Codification emerged to organize and clarify the law as the legal system came to be dominated by statutes instead of judge-made law. Perhaps ironically given the significance of this shift in favor of the people’s representatives, codification itself became the task of bureaucrats—today, in the age of the U.S. Code, of codifiers at the Office of the Law Revision Counsel (OLRC). Along with codification came statutes—today, 1 U.S.C. § 204(a)—providing that codified text could be enacted by Congress into positive law, thus becoming “legal evidence” of the law, as opposed to “prima facie” evidence of the law.30 However, statutes this Note calls “meaning-conformity statutes,” which Congress passes each time it enacts a positive-law codification, state that the enacted codified text has a meaning identical to that of the original text.31 Worryingly, Congress does not meaningfully check OLRC’s work to guard against changes in meaning.32 This leaves to courts the task of effectuating both the command that codified text is positive law once enacted and the command that the codified text retains the meaning of the original text. Thus far, courts have failed to adequately harmonize these commands,33 and the Supreme Court’s pre-textualist approach instead subverts the positive-law status of enacted codified text in the Court’s effort to prevent changes in meaning introduced through codification. A textualist update is urgently required.
In Part II, this Note derives the two texts canon by harmonizing 1 U.S.C. § 204(a) and meaning-conformity statutes. These commands also correspond to textualism’s guiding ideals of formalism and democracy. Textualism’s commitment to formalism requires textualists to begin and end the statutory-interpretation inquiry with the text whenever that text answers the question at issue. When there are two texts, one original text and one codified positive-law text, the codified positive-law text alone is formally the text, per Article I of the Constitution as well as 1 U.S.C. § 204(a).34 Accordingly, step one of the two texts canon holds that if the meaning of the codified positive-law text is clear, then that meaning is conclusive, and there is no need to look to the original text or any other interpretive tool.
However, textualism recognizes that formalism does not prevent recourse to aids for construction under certain, carefully delineated circumstances when the text itself cannot resolve the interpretive question. Specifically, contemporary textualism has come to endorse canons serving constitutional norms as legitimate tools for discerning the meaning of ambiguous text.35 This is where textualism’s commitment to democracy enters the picture. As used in this Note in relation to textualism, the admittedly capacious term “democracy” denotes textualism’s focus on preserving the Constitution’s design, which ostensibly assigns the legislative power exclusively to the people’s representatives, not to unelected bureaucrats (or judges).36 This focus on democracy produces step two of the two texts canon, which requires that interpretations of ambiguity or silence in the codified positive-law text may not contradict the original text. Maintaining the original text’s status as inferior to formal text but superior to all other tools of construction protects the original work of the people’s representatives against later distortions by unelected bureaucrats and judges.
Finally, in Part III, this Note applies the two texts canon to Section 1983 as perhaps the most significant case study for interpreting changes between original text and codified positive-law text. In proposing a broader textualist approach to the problem of two texts, this Note contributes to the ongoing effort led by scholars like Alexander A. Reinert, William Baude, and Joanna C. Schwartz to challenge the Court’s qualified-immunity jurisprudence,37 providing fresh grounds for revisiting qualified immunity when an appropriate vehicle presents itself to the Court.
Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in 18 The Tanner Lectures on Human Values 79, 99-100 (Grethe B. Peterson ed., 1997) (addressing the critique that textualism amounts to formalism by proclaiming, “Long live formalism”).
See Kevin Tobia, Brian G. Slocum & Victoria Nourse, Progressive Textualism, 110 Geo. L.J. 1437, 1440-41 (2022) (noting a turn toward democracy as the guiding principle of textualism); William N. Eskridge, Jr. & Victoria F. Nourse, Textual Gerrymandering: The Eclipse of Republican Government in an Era of Statutory Populism, 96 N.Y.U. L. Rev. 1718, 1736-38 (2021) (characterizing the democratic turn in textualism as populist).
See infra notes 148-151 and accompanying text (discussing the textualist Supreme Court’s reliance on nondelegation principles in justifying the major-questions doctrine as a way to protect the Constitution’s democratic structure); John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 675 (1997) (“[T]extualism should be understood as a means of implementing a central and increasingly well-settled element of the separation of powers—the prohibition against legislative self-delegation.”).
For an additional decision following clear meaning in the original text over ambiguous meaning in the codified text, see United States v. Sforza, 326 F.3d 107, 114 (2d Cir. 2003), which relies on “express[]” meaning conveyed in the Statutes at Large to fill in ambiguous meaning in positive-law provisions of the Code. For a decision engaging the issue of two texts and requiring a clear statement in the codified text for changed meaning to be found, see Redmond-Issaquah Railroad Preservation Ass’n v. Surface Transportation Board, 223 F.3d 1057, 1062 (9th Cir. 2000), which finds it unlikely that Congress would change the substantive meaning of original text in a positive-law codification without speaking clearly, particularly in light of other express substantive changes in the relevant statute.
For an assortment of lower-court decisions actively discussing the problem of two texts and siding with the original text, see, for example, United States v. Rivera, 131 F.3d 222, 234-36 (1st Cir. 1997) (Torruella, C.J., concurring), which assumes, in light of no legislative history indicating that Congress intended to change the meaning of the text, that a federal positive-law criminal statute retained the same meaning as the statute originally passed by Congress; American Cyanamid Co. v. Hammond Lead Products, Inc., 495 F.2d 1183, 1186 (3d Cir. 1974), which relies on Reviser’s Notes and the lack of legislative-history evidence to conclude that the revision of the positive-law Judicial Code should not be read to have substantively changed the statutory meaning; A.E.A. ex rel. Angelopoulos v. Volvo Penta of the Americas, LLC, 77 F. Supp. 3d 481, 491 (E.D. Va. 2015), which relies on legislative history to conclude that additional language in positive-law codification was not intended as a change in meaning; Aberdeen & Rockfish Railroad Co. v. United States, 682 F.2d 1092, 1103 (5th Cir. 1982), which relies on legislative history suggesting positive-law codification was not intended to change meaning to conclude that an explicit requirement contained in the original text should still apply despite its omission in the revised text, vacated, 467 U.S. 1237 (mem.) (1984); Carter v. Welles-Bowen Realty, Inc., 493 F. Supp. 2d 921, 927 (N.D. Ohio 2007) (quoting Finley v. United States, 490 U.S. 545, 554 (1989)), which refuses to conclude that by removing a phrase in an amendment consolidating statutory text, Congress could have changed its meaning “unless such intention is clearly expressed,” rev’d sub. nom. In re Carter, 553 F.3d 979 (6th Cir. 2009); Gonzalez v. Village of West Milwaukee, 671 F.3d 649, 661-62 (7th Cir. 2012), which determines that a change in the arrangement of statutory provisions in positive-law codification cannot alter the meaning of the law; Newton v. Federal Aviation Administration, 457 F.3d 1133, 1143 (10th Cir. 2006), which reads general language in codified positive-law text (“certificates issued under this chapter”) to apply only to enumerated lists contained in repealed-and-replaced original text (listing certificate types) based on the title of the revising statute, which asserted that codified text would not substantively change original text; In re Bayou Shores SNF, LLC, 828 F.3d 1297, 1319 (11th Cir. 2016), which applies the so-called “recodification canon of statutory construction,” concludes that the Office of the Law Revision Counsel must have erred in changing the meaning of positive-law codified text and that this changed text should be ignored absent a clear indication of an intentional change in meaning, and does “not find it significant . . . that Congress enacted the error into positive law”; and Gannett Satellite Information Network, LLC v. U.S. Department of Justice, No. 22-cv-475, 2023 WL 2682121, at *6 (D.D.C. Mar. 29, 2023), which refuses to read the substitution of “chapter” for “title” in codification as changing the substantive effect of the statute because the change occurred during positive-law codification.
For decisions observing the problem of two texts and siding with the modern, codified positive-law text over original text, see, for example, Stoianoff v. Commissioner of Motor Vehicles, 107 F. Supp. 2d 439, 445 (S.D.N.Y. 2000), which refuses to consider a provision that appears in the Statutes at Large version of the statute at issue because the provision was omitted from the main positive-law codified text and relegated to a note, aff’d, 12 F. App’x 33 (2d Cir. 2001); Kahrer v. Ameriquest Mortgage Co., 418 F. Supp. 2d 748, 754 (W.D. Pa. 2006), which concludes that because Congress removed a phrase in consolidating an amendment, a practice analogous to positive-law codification, it must have intended to expand the scope of the provision; Schmitt v. City of Detroit, 395 F.3d 327, 329-31 (6th Cir. 2005), which explains that “[g]iven that Title 5 [of the U.S. Code] has the force of positive law, the viability of [a provision contained in the original text in the Statutes at Large] is premised upon whether it was codified”; Borders v. United States, No. 09-CV-616, 2010 WL 5093427, at *3 (S.D. Ohio Dec. 8, 2010), which asserts that positive law controls over the Statutes at Large; Nichols v. Rysavy, 809 F.2d 1317, 1327-28 (8th Cir. 1987), which rejects the argument that positive-law codification should not be read to change original text’s meaning and instead relies on “clear and definitive” codified positive-law text; and United States v. Romig, Nos. 00CR355, 03-CV-2640, 2003 WL 22143730, at *1 n.2 (D. Minn. Aug. 18, 2003), which refuses to consider the Statutes at Large where a provision has been enacted as positive law.
See Washington-Dulles Transp., Ltd. v. Metro. Wash. Airports Auth., 263 F.3d 371, 378 (4th Cir. 2001) (noting that “slightly less than half” of the titles of the U.S. Code have been enacted into positive law). For a demonstration of the extent to which the text in a positive-law title of Code differs from the original text, consider a House Judiciary Committee report summarizing the line-by-line changes contained in a codification bill. See H.R. Rep. No. 113-44, at 6-7 (2013).
For example, qualified-immunity precedents have consistently failed to consider the omission of a provision in the original text that specifically abrogated state common-law immunities. See Reinert, supra note 1, at 244 (noting that the Supreme Court’s qualified-immunity jurisprudence “has entirely failed to grapple with the Civil Rights Act’s enacted text”); see also Pierson v. Ray, 386 U.S. 547, 553-57 (1967) (creating qualified immunity and not considering the original text contained in the Civil Rights Act of 1871).
The first case to raise the issue after the first major codification effort was United States v. Ryder, 110 U.S. 729, 739-40 (1884). See also In re Bayou Shores SNF, LLC, 828 F.3d 1297, 1315-16 (11th Cir. 2016) (cataloguing a nonexhaustive list of twenty-two Supreme Court decisions engaging with the problem of two texts, omitting all cases concerning Section 1983 discussed in Part III, infra).
See Finley v. United States, 490 U.S. 545, 554 (1989). The Court continues to cite the rule from Finley in interpreting changes arising out of the 1948 recodification of the Judicial Code specifically, but the Court has not acknowledged the broader problem of two texts in recent years. See Ankenbrandt v. Richards, 504 U.S. 689, 700 (1992); Keene Corp. v. United States, 508 U.S. 200, 209 (1993); Scheidler v. Nat’l Org. for Women, Inc., 547 U.S. 9, 20 (2006); John R. Sand & Gravel Co. v. United States, 552 U.S. 130, 136 (2008). One year after Finley, the Court considered a change in the meaning of codified text in Ngiraingas v. Sanchez, and Justice Brennan discussed Finley in dissent, but the majority did not explicitly take a generalizable approach to the problem. 495 U.S. 182, 187-92 (1990); see id. at 200 (Brennan, J., dissenting). The current status of Finley and its predecessor precedents addressing the problem of two texts is therefore unclear.
In re Bayou Shores SNF, LLC, 828 F.3d at 1319 (summarizing Supreme Court and Eleventh Circuit cases to conclude that the so-called “recodification canon” required following the original text where the court surmised that if Congress had intended to make a fundamental change during positive-law codification, “it would [have] merit[ted] some mention”); see also Washington-Dulles Transp., Ltd. v. Metro. Wash. Airports Auth., 263 F.3d 371, 378-79 (4th Cir. 2001) (quoting the Supreme Court’s rule as stated in Cass v. United States, 417 U.S. 72, 82 (1974)).
See, e.g., Borders v. United States, No. 09-CV-616, 2010 WL 5093427, at *3 (S.D. Ohio Dec. 8, 2010) (stating that the original text need not be considered once the codified text is positive law and not citing Supreme Court cases concerning positive-law codification); United States v. Romig, Nos. 00CR355, 03-CV-2640, 2003 WL 22143730, at *1 n.2 (D. Minn. Aug. 18, 2003) (same); Nichols v. Rysavy, 809 F.2d 1317, 1327 (8th Cir. 1987) (noting the litigants’ reliance on Supreme Court authority stating that Congress does not intend to change the meaning of the law through codification but ignoring this authority and adhering to the codified positive-law text).
See generally William Baude, Is Qualified Immunity Unlawful?, 106 Calif. L. Rev. 45 (2018) (refuting the common-law justification for qualified immunity); Joanna C. Schwartz, The Case Against Qualified Immunity, 93 Notre Dame L. Rev. 1797 (2018) (refuting policy-based arguments in favor of qualified immunity with empirical evidence about government indemnification of officers facing Section 1983 liability).
See Mullenix v. Luna, 577 U.S. 7, 24-26 (2015) (Sotomayor, J., dissenting); Ziglar v. Abassi, 582 U.S. 120, 159 (2017) (Thomas, J., concurring) (arguing that the Court’s qualified-immunity “analysis is no longer grounded in the common-law backdrop against which Congress enacted the 1871 Act”); Jay R. Schweikert, Qualified Immunity: A Legal, Practical, and Moral Failure, Cato Inst. 15, 18 (Sept. 14, 2020), https://www.cato.org/sites/cato.org/files/2020-09/PA%20901_1.pdf [https://perma.cc/NH5J-J6NM].
See, e.g., Act of Dec. 18, 2010, Pub. L. No. 111-314, § 2, 124 Stat. 3328, 3328 (explaining that the enacted Title 51 of the U.S. Code conforms “to the understood policy, intent, and purpose of Congress in the original enactments”); Act of Dec. 19, 2014, Pub. L. No. 113-287, § 2, 128 Stat. 3094, 3094 (using the same language in enacting Title 54). The earliest enactments did not use this precise language, but they were still identified as acts to “revise, codify, and enact into law,” rather than substantively amend, the new title of the U.S. Code. See Act of Aug. 10, 1956, ch. 1041, pmbl., 70 Stat. 1126, 1126 (1956) (enacting Title 10, “Armed Forces,” and Title 32, “National Guard,” of the U.S. Code).
Rather than harmonizing applicable statutory commands, lower courts have tended to emphasize either the original text or the codified positive-law text, and even those that have attempted to strike a compromise have not achieved the harmony this Note seeks to advance with the two texts canon. See supra notes 8-14 and accompanying text.
See West Virginia v. EPA, 597 U.S. 697, 736 (2022) (Gorsuch, J., concurring) (explaining that “clear-statement rules help courts” fulfill their duty “to ensure that acts of Congress are applied in accordance with the Constitution”); Amy Coney Barrett, Substantive Canons and Faithful Agency, 90 B.U. L. Rev. 109, 169 (2010) (arguing that constitutionally inspired substantive canons are a legitimate exercise of judicial review where the judicial obligation to enforce the Constitution qualifies the judicial obligation to serve as a faithful agent of the legislature).
Although this Note typically uses the term democracy in this context, the Note also has occasion to discuss other senses in which textualism promotes democracy. Specifically, Section II.B considers how textualism’s formalism itself promotes a distinct ideal also understood by scholars and textualist judges as “democracy”: treating the ordinary meaning of the text as conclusive allows ordinary people to ascertain the meaning of the law so that they can rely on the law and hold elected representatives accountable for the laws they create.