Kevin Tobia††
Associate Professor of Law, Georgetown University Law Center. This essay originated from a conference on Corpus Linguistics and the Second Amendment, at the Duke Center for Firearms Law. Thanks to Joseph Blocher, Jacob Charles, and Darrell Miller for the invitation and to the co-panelists and participants for their comments, especially Dennis Baron, William Baude, Anya Bernstein, and Stephen Mouritsen. Great thanks to John Macy and the Duke Law Journal for outstanding editorial assistance.

PDFPDF

Introduction

Textualism has broad support—at the Supreme Court, [1][1]. See Victoria Nourse & William N. Eskridge, Textual Gerrymandering: The Eclipse of Republican Government in an Era of Statutory Popularism, 96 N.Y.U. L. Rev. 1718, 1722 (2021) (“Should interpreters focus on the readers and consumers of statutes (We the People) or the authors and producers of statutes (Congress)? . . . On its face, the now-dominant Supreme Court approach elevates the consumer perspective and belittles or ignores that of the producers. This is an alarming development.”); Kevin Tobia, Brian Slocum, & Victoria Nourse, Statutory Interpretation from the Outside, 122 Colum. L. Rev. 213, 216 (2022) (“[O]rdinary meaning is regularly deployed by all members of the current Supreme Court.”). within the lower federal courts’ new cohort of young “Trump judges,” [2][2]. Jason Zengerle, How the Trump Administration Is Remaking the Courts, N.Y. Times Mag. (Aug. 22, 2018), https://www.nytimes.com/2018/08/22/magazine/trump-remaking-courts-judiciary.html [https://perma.cc/UG99-J2QZ] (President Trump was committed to “nominating and appointing judges that are committed originalists and textualists.”). within many state courts, [3][3]. Abbe R. Gluck, The States as Laboratories of Statutory Interpretation: Methodological Consensus and the New Modified Textualism, 119 Yale L.J. 1750, 1758 (2010) (“[I]n the states studied, textualism is more than merely alive and well; it is the controlling interpretive approach—the consensus methodology chosen by the courts.”). and even within the legal academy. [4][4]. Eric Martínez & Kevin Tobia, The Legal Academy and Theory Survey, (unpublished manuscript) (on file with author). Textualism comes in several variations, [5][5]. See, e.g., Gluck, supra note 3; Tara Leigh Grove, Which Textualism?, 134 Harv. L. Rev. 265, 265 (2020) (comparing “formalistic” and “flexible” forms of textualism). and a new “populist” version is taking hold. [6][6]. See Nourse & Eskridge, supra note 1, at 1723; see also generally Anya Bernstein & Glen Staszewski, Judicial Populism, 106 Minn. L. Rev. 283 (2021) (commenting on judicial populism). Modern textualists claim to interpret law from the perspective of an ordinary person, [7][7]. Anya Bernstein, Democratizing Interpretation, 60 Wm. & Mary L. Rev. 435, 440 (2018) (“Textualism instructs judges to interpret a statute as its addressees would understand it.”); Amy Coney Barrett, Congressional Insiders and Outsiders, 84 U. Chi. L. Rev. 2193, 2195 (2017) (“[Textualists] view themselves as agents of the people rather than of Congress and as faithful to the law rather than to the lawgiver”). which includes giving terms in law their ordinary meanings. [8][8]. E.g., Amy Coney Barrett, Assorted Canards of Contemporary Legal Analysis: Redux, 70 Case W. Res. L. Rev. 855, 856 (2020) (noting the significance of “ordinary meaning”). This commitment is taken to promote rule of law values (e.g. publicity), fair notice, and a “democratic” mode of interpretation. [9][9]. See Bernstein, supra note 7, at 442; see also Kevin Tobia, Brian Slocum & Victoria Nourse, Progressive Textualism, 110 Geo. L.J. (forthcoming 2022) (documenting modern textualism’s motivations). Even non-textualist Justices have begun to appeal to the “ordinary speaker.” [10][10]. Consider Justice Roberts’s recent question in Facebook v. Duguid’s oral argument:
[O]ur objective is to settle upon the most natural meaning of the statutory language to an ordinary speaker of English, right? . . . So the most probably useful way of settling all these questions would be to take a poll of 100 ordinary – ordinary speakers of English and ask them what [the statute] means, right?
Transcript of Oral Argument at 51–52, Facebook, Inc. v. Duguid, 141 S. Ct. 1163 (2021) (No. 19-511). Justice Alito, in his concurring opinion in Duguid, noted that
[t]he strength and validity of an interpretive canon is an empirical question, and perhaps someday it will be possible to evaluate these canons by conducting what is called a corpus linguistics analysis, that is, an analysis of how particular combinations of words are used in a vast database of English prose.
Facebook, Inc. v. Duguid, 141 S. Ct. at 1174 (Alito, J., concurring).

How do today’s textualists go about finding ordinary meaning? They regularly appeal to sources including “dictionaries, corpus linguistics, and canons of construction.” [11][11]. Nourse & Eskridge, supra note 1, at 1727. The flexibility of dictionaries and canons of construction is well-documented. [12][12]. On canons, see Karl Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons about How Statutes Are to Be Construed, 3 Vand. L. Rev. 395, 401 (1950) (“[T]here are two opposing canons on almost every point.”); see also generally Anita Krishnakumar, Dueling Canons, 65 Duke L.J. 909 (2016); Anita Krishnakumar & Victoria Nourse, The Canon Wars, 97 Tex. L. Rev. 163 (2018); Ryan Doerfler, Late-Stage Textualism, 2022 Sup. Ct. Rev. (forthcoming 2022). On dictionaries, see generally Samuel A. Thumma & Jeffrey L. Kirschmeier, The Lexicon Has Become a Fortress: The United States Supreme Court’s Use of Dictionaries, 47 Buff. L. Rev. 227 (1999); Stephen C. Mouritsen, The Dictionary is Not a Fortress: Definitional Fallacies and a Corpus-Based Approach to Plain Meaning, 2010 B.Y.U. L. Rev. 1915 (2010); Ellen P. Aprill, The Law of the Word: Dictionary Shopping in the Supreme Court, 30 Ariz. St. L.J. 227 (1998); James J. Brudney & Lawrence Baum, Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras, 55 Wm. & Mary L. Rev 483 (2013). Judges can cherry-pick helpful dictionary definitions, [13][13]. pan> . Aprill, supra note 12, at 300 (“[O]pinions often cite or rely on only one definition in only one dictionary . . . . For the most part, opinions fail to explain or justify the basis for their choice.”); Brudney & Baum, supra note 12, at 491 (arguing that the Supreme Court has a “tendency to cherry-pick definitions that support results reached on other grounds”); Kevin Tobia, Brian Slocum & Victoria Nourse, Ordinary Meaning and Ordinary People, 171 U. Pa. L. Rev. (forthcoming 2023) (documenting the Supreme Court’s citation of dozens of ordinary and legal dictionaries). and for many canons of interpretation, there is an opposing canon that could support the opposite result.

This essay explores textualism’s newest tool: corpus linguistics. Over the past five years, the tool has been increasingly employed by U.S. courts. [14][14]. Kevin Tobia, The Corpus and the Courts, U. Chi. L. Rev. Online (2021) (documenting judge’s appeals to corpus linguistics, rising sharply over the past five years). Legal corpus linguistics has also caught the attention of the U.S. Supreme Court. Justice Thomas mentioned corpus linguistics in his 2018 Carpenter dissent, and Justice Alito noted it again in his 2021 Duguid concurrence. [15][15]. Carpenter v. United States, 138 S. Ct. 2206, 2238 n.4 (2018) (Thomas, J., dissenting); Facebook, Inc. v. Duguid, 141 S. Ct. 1163, 1175 (2021) (Alito, J., concurring). Most recently, Justices Roberts and Barrett discussed corpus linguistics in a 2022 oral argument. [16][16]. Transcript of Oral Argument at 9–11, ZF Automotive U.S., Inc. v. Luxshare, Ltd. (2022) (No. 21-401).

Broadly speaking, legal corpus linguistics treats collections of texts (“corpora”) as data. [17][17]. Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning, 127 Yale L.J. 788, 795 (2018) (“Corpus linguists study language through data derived from large bodies—corpora—of naturally occurring language.”). To learn about the ordinary meaning of a statutory or constitutional term (e.g. “commerce”), a textualist would evaluate how that term is commonly used in different written sources (e.g. books) and what other words tend to appear near it in those sources. For example, the interpreter might consider different senses of a term (e.g. “commerce” in the narrow sense of “the trading . . . and selling of goods,” versus “commerce” in the broader sense of “all forms of social and economic intercourse”). [18][18]. Thomas R. Lee & James C. Phillips, Data-Driven Originalism, 167 U. Pa. L. Rev. 261, 300 (2019). Next, the interpreter could evaluate how often each of those senses appears in the corpus. Perhaps, for example, a scholar may find that the narrower trade sense appears more frequently than the broader sense. Some scholars suggest that these data evince the constitutional or statutory meaning of the term. For example, a recent article suggests that corpus linguistics data about “commerce” “at least arguably, tells us that the original [constitutional] meaning of commerce is the trade sense of the term.” [19][19]. Id. at 323.

The essay argues that corpus linguistics—an important and useful method in linguistics—is unlikely to achieve textualist’s theoretical aims. Many have criticized legal corpus linguistics’ prospects, cautioning that judges do not have the training or expertise to employ the tool or pointing to fundamental flaws of the current method, as applied to legal debates. [20][20]. See, e.g., Bernstein, supra note 7; Anya Bernstein, What Counts as Data?, 86 Brook. L. Rev. 435 (2021); Anya Bernstein, Legal Corpus Linguistics and the Half-Empirical Attitude, 106 Cornell L. Rev. 1397 (2021); John S. Ehrett, Against Corpus Linguistics, 108 Geo. L.J. Online 50 (2019); Ethan J. Herenstein, The Faulty Frequency Hypothesis: Difficulties in Operationalizing Ordinary Meaning Through Corpus Linguistics, 70 Stan. L. Rev. Online 112 (2017); Donald L. Drakeman, Is Corpus Linguistics Better than Flipping a Coin?, 109 Geo. L.J. Online 81 (2020); Stanley Fish, The Interpretive Poverty of Data, Balkinization (Mar. 2, 2018) https://balkin.blogspot.com/2018/03/the-interpretive-poverty-of-data.html [https://perma.cc/4X4S-7QZ8]; Carissa Byrne Hessick, Corpus Linguistics and the Criminal Law, 2017 B.Y.U. L. Rev. 1503 (2018); Brian G. Slocum & Stefan Th. Gries, Judging Corpus Linguistics, 94 S. Cal. L. Rev. Postscript 13 (2020); Kevin Tobia, Testing Ordinary Meaning, 134 Harv. L. Rev. 726 (2020); Evan C. Zoldan, Corpus Linguistics and the Dream of Objectivity, 50 Seton Hall L. Rev. 401 (2019). But see Thomas R. Lee & Stephen C. Mouritsen, The Corpus and the Critics, 88 U. Chi. L. Rev. 275 (2021) (defending legal corpus linguistics). This essay starts from a different perspective, noting that judges are already employing and referencing corpus linguistics in legal interpretation, [21][21]. Tobia, supra note 14 (documenting judge’s appeals to corpus linguistics, rising sharply over the past five years). as scholars advance provocative corpus linguistics arguments about statutory and constitutional language. [22][22]. E.g., Lee & Phillips, supra note 18, at 300–11 (providing a corpus linguistic analysis of “commerce”).

Corpus linguistics has been offered as a preferred interpretive tool, avoiding the pitfalls of dueling canons or cherry-picked dictionary definitions. [23][23]. E.g., Lee & Mouritsen, supra note 17, at 877 (suggesting that corpus linguistics offers better evidence of ordinary meaning than dictionaries). However, this essay proposes, there is one important commonality among textualists’ current use of dictionaries, canons, and legal corpus linguistics: Flexibility. The essay articulates ten emerging “arguments” and “counterarguments” of legal corpus linguistics. Alongside the pitfalls of dueling canons and dueling dictionaries, legal interpreters should be aware of the similar possibility of “clashing corpora.” Corpus linguistics can greatly enrich our understanding of language and cognition, but—at least in the form employed by textualist judges and commentators—it does not provide inexorable determinations of how ordinary people understand legal language in contested cases of legal interpretation.

I. Popular Textualism

Today’s textualism is popular in two different senses. First, it has significant support from judges and scholars. [24][24]. See supra notes 1–4 and accompanying text. (That said, it is not universally approved; modern critics lambast textualist practice as flawed [25][25]. E.g., Victoria Nourse, Textualism 3.0: Statutory Interpretation After Justice Scalia, 70 Ala. L. Rev. 667 (2019). and even “bogus.” [26][26]. Mitchell N. Berman & Guha Krishnamurthi, Bostock was Bogus: Textualism, Pluralism, and Title VII, 97 Notre Dame L. Rev. 67 (2021).). Today’s textualism is also “popular” in a second sense: its interpretive inquiry is focused on the public. [27][27]. Bernstein & Staszewski, supra note 6, at 287 (“[T]he brand of populism we address here . . . makes claims justifying action in the name of ‘the people.’”). As Justice Barrett puts it, modern textualists “view themselves as agents of the people rather than of Congress and as faithful to the law rather than to the lawgiver.” [28][28]. Barrett, supra note 7, at 2195. Textualists thus “approach language from the perspective of an ordinary English speaker.” [29][29]. Id. at 2194. Judges increasingly adopt this popular stance, committing to interpret statutory and constitutional language empirically, in line with its “ordinary” or “public” meaning. [30][30]. E.g., Lawrence B. Solum, Triangulating Public Meaning: Corpus Linguistics, Immersion, and the Constitutional Record, 2017 B.Y.U. L. Rev. 1621 (2018).

How does one find “ordinary public meaning”? [31][31]. This phrase, appearing in Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1738 (2020), reflects the synthesis of textualist’s ordinary meaning and originalist’s public meaning. As Victoria Nourse documents, “new” textualists are statutory originalists. Nourse, supra note 25, at 669. Textualists appeal to linguistic evidence, like dictionary definitions. [32][32]. Thumma & Kirschmeier, supra note 12, at 260–62 (documenting dictionary usage by Justices of the U.S. Supreme Court); Mouritsen, supra note 12, at 1918 (noting the “overarching trend to rely upon dictionaries to resolve lexical ambiguity”). Commentators question that approach. Differing definitions allow judges to go “dictionary-shopping,” [33][33]. Aprill, supra note 12, at 318 (arguing that Justice Scalia sometimes treats dictionary definitions as authoritative, but other times rejects dictionary definitions). and empirical studies suggest that judges’ dictionary use is often “ad hoc and subjective.” [34][34]. Brudney & Baum, supra note 12, at 483 (“[T]he Court’s patterns of dictionary usage reflect a casual form of opportunistic conduct.”).

“Picking and choosing” is a broader issue for textualists. Traditionally, textualism has aimed to constrain legal interpretation and limit judicial discretion. [35][35]. John F. Manning, What Divides Textualists from Purposivists?, 106 Colum. L. Rev. 70, 74–75 (2006). For example, textualists offer the limitation of judicial discretion as one reason for courts to avoid evaluating legislative intent. [36][36]. See John F. Manning, Second-Generation Textualism, 98 Cal. L. Rev. 1287, 1289 (2010). With respect to legislative history, many textualists seek to avoid judge’s freedom to pick and choose their friends among the crowd. But recent critics note that textualist themselves can pick and choose what text to analyze, [37][37]. Victoria Nourse, Picking and Choosing Text: Lessons for Statutory Interpretation from the Philosophy of Language, 69 Fla. L. Rev. 1409, 1423–29 (2017); Nourse & Eskridge, supra note 1, at 1747–51. pick and choose dictionaries, [38][38]. Brudney & Baum, supra note 12, at 529–31. and pick and choose definitions. [39][39]. Brudney & Baum, supra note 12, at 529–31. Textualists also choose among canons, “hundreds of interpretive presumptions that have no hierarchy among them.” [40][40]. Abbe R. Gluck, Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62, 62 (2015); see also Llewellyn, supra note 12, at 401 (“[T]here are two opposing canons on almost every point.”). The dueling canons and “dueling dictionaries” show that a mere commitment to “text” does not guarantee limited judicial discretion.

II. Legal Corpus Linguistics

Enter “legal corpus linguistics,” an exciting new tool for textualists and other theorists committed to ordinary meaning. [41][41]. Lee & Mouritsen, supra note 17, at 795. Corpora (the plural of “corpus”) are samples of language-usage. [42][42]. Jesse Egbert, The Corpus—A Sample By Another Name, Linguistics with a corpus (May 27, 2021), https://linguisticswithacorpus.wordpress.com/2021/05/27/the-corpus-a-sample-by-another-name/ [https://perma.cc/W3ND-YKQK]. To learn about the ordinary or public meaning of a term like “commerce” or “bear arms,” interpreters might look beyond just a few dictionaries, to hundreds of uses of the phrase in a corpus. For example, a search of the Corpus of Founding Era American English revealed 281 instances of the phrase “bear arms.” “[O]nly a handful don’t refer to war, soldiering or organized armed action,” suggesting to some that “the natural meaning of ‘bear arms’ in the framers’ day was military.” [43][43]. Dennis Baron, Opinion: Antonin Scalia Was Wrong About the Meaning of “Bear Arms.” Wash. Post. (May 21, 2018), https://www.washingtonpost.com/opinions/antonin-scalia-was-wrong-about-the-meaning-of-bear-arms/2018/05/21/9243ac66-5d11-11e8-b2b8-08a538d9dbd6_story.html [https://perma.cc/E2FA-QMFW]; see also Dennis Baron, Corpus Evidence Illuminates the Meaning of Bear Arms, 46 Hastings Const. L.Q. 509, 510 (2019); Alison L. LaCroix, Historical Semantics and the Meaning of the Second Amendment, Panorama (Aug. 3, 2018), http://thepanorama.shear.org/2018/08/03/historical-semantics-and-the-meaning-of-the-second-amendment/ [https://perma.cc/5WKC-S4AY]; Josh Jones, Note, The “Weaponization” of Corpus Linguistics: Testing Heller’s Linguistic Claims, 34 B.Y.U. J. Pub. L. 135, 135 (2020). But see James C. Phillips & Josh Blackman, Corpus Linguistics and Heller, 56 Wake Forest L. Rev. 609 (2021). With an appeal to “big data,” legal corpus linguistics offers a new, and perhaps more objective and empirically robust, basis to seek “ordinary meaning.”

Despite its promise, legal corpus linguistics approach has faced criticism. [44][44]. See supra note 20. Yet, over the past decade, scholars and judges have adopted corpus linguistic tools to address questions about public meaning. This trend is sharp in the past five years, with citation to corpus linguistics from several state and federal courts. [45][45]. Tobia, supra note 14. Scholars have advanced new corpus linguistics arguments about constitutional language including “commerce,” [46][46]. Lee & Phillips, supra note 18, at 300–11. and in 2018, Justice Thomas cited corpus linguistics evidence about the meaning of “search” at the U.S. Supreme Court. [47][47]. Carpenter v. United States, 138 S. Ct. 2206, 2238 n.4 (2018) (Thomas, J., dissenting).

As legal corpus linguistics gains prominence, one of its primary concerns should be its political (non-)neutrality and potential for abuse. Some critics accuse textualist theory to be more motivated by conservatism than fidelity to democracy or separation of powers, [48][48]. See, e.g., Andrei Marmor, The Immorality of Textualism, 38 Loy. L.A. L. Rev. 2063, 2065 (2005)(“I believe that the underlying motivation of textualism derives from a neoconservative conception of the regulatory state, much more so, anyway, than from a concern with principles of democracy and separation of powers.”). suggesting textualism is a mere “smokescreen by conservative judges to reach ideologically acceptable outcomes.” [49][49]. Grove, supra note 5, at 266 (Grove does not endorse this idea, but cites others who do, including Neil H. Buchanan & Michael C. Dorf, A Tale of Two Formalisms: How Law and Economics Mirrors Originalism and Textualism, 106 Cornell L. Rev. 591, 640 (2020) (suggesting that textualism is “a rhetorical smokescreen for extremely conservative results”)); William N. Eskridge, Jr. & Philip P. Frickey, The Supreme Court, 1993 Term — Foreword: Law as Equilibrium, 108 Harv. L. Rev. 26, 77 (1994); Margaret H. Lemos, The Politics of Statutory Interpretation, 89 Notre Dame L. Rev. 849, 851 (2013)). A similar concern applies to textualist tools, including legal corpus linguistics. Thus far, legal corpus linguistics has been discussed more frequently (and favorably) by Republican-appointed judges than by Democratic-appointed ones. [50][50]. Tobia, supra note 14.

But some legal corpus linguistics research appears to have the opposite valence. For example, some of the most robust legal corpus linguistics research, from a number of scholars, questions the conclusions of Heller, finding that, “the Supreme Court’s reasoning may be flawed.” [51][51]. Baron, supra note 43; see also Baron, supra note 43; LaCroix, supra note 43; Jones, supra note 43; Neal Goldfarb, A (Mostly Corpus-Based) Linguistic Reexamination of D.C. v. Heller and the Second Amendment (unpublished manuscript), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3481474 [https://perma.cc/8E4E-3SE3]. But see Phillips & Blackman, supra note 43. The Second Amendment is an alluring test case: Can legal corpus linguistics attain textualism’s promise of objectivity, and will commentators and judges persuaded by corpus linguistics evidence concerning “commerce” and “search” be similarly moved by evidence about “bear arms”?

As an example, consider a recent case. In Jones v. Becerra (concerning a Second Amendment challenge to California’s ban on firearm purchases by those between age 18 and 21), the Ninth Circuit ordered supplemental briefing concerning legal corpus linguistics and the Second Amendment. Specifically, the parties were instructed to address the “original public meaning” of the phrases: “A well regulated Militia”; “the right of the people”; and “shall not be infringed”—and to address “[h]ow does the tool of corpus linguistics help inform the determination of the original public meaning of those Second Amendment phrases?” [52][52]. Jones v. Becerra, Order, Case 20-56174, at 1 (9th Cir. Mar. 26, 2021).

The parties’ responses were striking. Both the plaintiff-appellants and defendant-appellees criticized the method of corpus linguistics. [53][53]. Supplemental Brief for Appellees at 2, Jones v. Bonta, 2022 WL 1485187 (9th Cir. 2022) (No. 20-56174), 2021 WL 1727661, at *2 (“[I]nitial results suggest that a corpus linguistics analysis would likely be of limited utility in answering [the] question.”); Supplemental Brief for Appellants at 2–3, Jones v. Bonta, 2022 WL 1485187 (9th Cir. 2022) (No. 20-56174), 2021 WL 1727665, at *2–*3 (“Because of the weaknesses inherent in the methodology of corpus linguistics, however, it ultimately sheds little light on the matter—and it certainly can do nothing to upset the interpretation of the Second Amendment adopted by binding Supreme Court precedent.”). At the same time, both the plaintiff-appellants and defendant-appellees conducted corpus linguistic analyses and managed to find corpus linguistics data to support opposing conclusions about the original public meaning of the Second Amendment. [54][54]. Supplemental Brief for Appellees, Jones v. Bonta, 2022 WL 1485187 (9th Cir. 2022) (No. 20-56174), 2021 WL 1727661, at *25–*26. In their supplemental brief, the Appellees noted that
preliminary searches in COHA and COFEA for the phrase ‘right of the people’ return a relatively manageable number of hits: approximately 200 in each database. They do not appear to provide clear evidence that this phrase, as used in the Second Amendment, was originally understood to protect an individual right for persons under 21 to keep or bear arms (much less to purchase or receive them from a commercial dealer), however.
Id.; Supplemental Brief for Appellants, Jones v. Bonta, 2022 WL 1485187 (9th Cir. 2022) (No. 20-56174), 2021 WL 1727665, at *2 (“We have conducted a corpus-linguistics analysis of the three phrases identified by the Court, and we set forth the results below—results that are fully consistent with the conventional evidence of the original public meaning of those phrases (and with the determinations in Heller).”).

This result is not surprising, but Jones v. Becerra portends a new time of “clashing corpora.” Like judicial use of dictionaries, judicial use of corpus linguistics admits of interpretive choice and flexibility. Judges and advocates have flexibility in terms of which selection from the legal text to analyze, which corpus or corpora to search, which search(es) to conduct, and what conclusions to draw from the results returned from the corpus.

The phenomenon of “dueling dictionaries” is well-known. But this essay concludes by sketching some of the emerging “moves” of legal corpus linguistic argumentation (inspired by the style of Llewellyn’s dueling canons). [55][55]. Llewellyn, supra note 12, at 401–06. Here, thrust and parry 1 is nearly identical to Llewellyn’s pair concerning ordinary versus legal meaning. Here, the thrust and parry arguments imply conflicting, although not necessarily opposite, conclusions.

To enumerate these clashing arguments is not to endorse or discredit any particular one. It is simply to question the legal corpus linguistics (and textualist) claim that the introduction of these new empirical methods will straightforwardly constrain legal interpreters or provide uncontroversial answers to hard interpretive questions.

To be sure, where textualists aim to uncover how ordinary people understand language, legal corpus linguistics seems no less promising than dictionaries, canons, intuition, and other textualist tools. But where textualists can freely leverage any of these arguments and counterarguments (and pick and choose which text to analyze and searches to run [71][71]. See, e.g., Nourse & Eskridge, supra note 1, at 1721 (“As this Article suggests, in any difficult case, the textualist judge starts with two potentially outcome-determinative decisions: a choice of text—the scope of text the judge decides to focus on when interpreting a statute—and a choice of context surrounding this text.”).), legal corpus linguistics is unlikely to provide much more predictability or constraint.

Conclusion

Legal corpus linguistics is not yet “popular” in the sense of receiving widespread approval. [72][72]. See supra note 20 and accompanying text. And it is not necessarily “popular” in the sense of relating to the ordinary public; the historical corpora that legal scholars have most often relied upon tend to overrepresent elites’ language. [73][73]. See Bernstein, supra note 69 (noting that a popular corpus of founding era language represents a “tiny minority” of the founding era population, consisting of the language of “political superstars, lawmakers and government agents, [and] a few legal scholars.”). But legal corpus linguistics is popular in the sense of being increasingly encountered—cited by commentators and judges as relevant to the public meaning of “commerce,” “search,” and maybe even “bear arms.” [74][74]. See also Tobia, supra note 14 (documenting judge’s appeals to corpus linguistics, rising sharply over the past five years).

This essay has argued that legal corpus linguistics is unlikely to provide easy answers in hard cases of interpretation. Will the fact that legal corpus linguistics often admits of “clashing” arguments sap its popularity? It is unlikely. For one, that an interpretive tool could support putatively clashing arguments does not imply that the tool is ultimately flawed—perhaps the clashes can be resolved. Some argue that legal corpus linguistics’ current problems are largely the result of proponents’ presentation of a “highly impoverished version of [corpus linguistics]” and caution that legal scholars and judges should avoid “reduc[ing it] to the point of caricature.” [75][75]. Stefan Th. Gries, Corpus Linguistics and the Law: Extending the Field from a Statistical Perspective, 86 Brook. L. Rev. 321, 324 (2021). As legal corpus linguistics develops, perhaps the judges and scholars who rely upon these tools will clarify the appropriate methodological moves.

But even if the clashing is more fundamental, I would bet that legal corpus linguistics is here to stay. Despite its critics, textualism is increasingly influential at the Supreme Court and lower courts. [76][76]. See supra notes 1–10 and accompanying text. And textualist argument requires apparent textualist evidence.Perhaps “clashing corpora” will share the fate of the “dueling canons” and “dueling dictionaries.” It’s been seventy years since Llewellyn noted the “dueling canons” and at least twenty years since the observation of “dueling dictionaries.” Today’s Supreme Court regularly relies on both tools. [77][77]. Anita S. Krishnakumar, Cracking the Whole Code Rule, 96 N.Y.U. L. Rev. 76, 97 (2021) (reporting that the Roberts court relies on language and grammar canons in 8.7% of statutory meaning cases, substantive canons in 14.9% of such cases, and dictionaries in 21.6% of such cases).