1.Why ordinary meaning?

Reliance on the plain or ordinary meaning of a legal text in the decisional law of the United States is a feature inherited from English courts, who referenced a similar formulation of the Ordinary Meaning Rule when interpreting the acts of parliament:

The only rule for construction of Acts of Parliament is that they should be construed according to the intent of the Parliament which passed the Act. If the words of the Statute are in themselves precise and unambiguous, then no more can be necessary than to expound those words in that natural and ordinary sense. The words themselves alone do, in such a case, best declare the intention of the law giver.11. The Sussex Peerage, 8 Eng. Rep. 1034, 1057 (H.L.) (appeal taken from Eng.) (1844) (Tindal, L.C.J.)

Above, as in the decisions of U.S. courts, judges make determination as to the clarity of the language used by the legislative body. If that language is perceived to be clear, then the judge is merely expected to apply the “ordinary” or “natural” sense of the words. These same features of the ordinary meaning rule can be observed in the decisions of the United States Supreme Court, which as early as 1917 stated that “[w]here the language is plain and admits and no more than one meaning the duty of interpretation does not arise and the rules which are to aid doubtful meanings need no discussion.”22. Caminetti v. United States, 242 U.S. 470, 485–86 (1917).

Appeals to the ordinary meaning of the legal texts have become extremely common in legal opinions in the United States. A search of the United States Case Law feature of Google Scholar reveals some 223,000 cases in which the phrase “ordinary meaning” occurs33.See https://​scholar​.google​.com​/scholar​_courts (last visited May 27, 2021). and the use of the phrase “ordinary meaning” has increased dramatically in recent decades (Mouritsen 2010Mouritsen, Stephen 2010 “The Dictionary Is Not a Fortress: Definitional Fallacies and a Corpus-Based Approach to Plain Meaning.” BYU Law Review 5:1915–1980., 1971).

There are a variety of reasons why courts turn to ordinary meaning. To begin with, there are rule-of-law interests in interpreting the words of a legal text according to their ordinary meaning. As Professor William S. Eskridge Jr. has observed, “[a] polity governed by the rule of law aspires to have legal directives that are known to the citizenry, that are predictable in their application, and that officials can neutrally and consistently apply based upon objective criteria” (Eskridge 2016Eskridge Jr., William S. 2016 Interpreting Law: A Primer on How to Read Statutes and the Constitution. St. Paul, MN: Foundation Press., 35). Judges endeavor to interpret the words of a legal text according to their ordinary meaning believing that doing so will render the application of the law neutral and consistent with the expectations of the citizens who are expected to abide by it.

Judges may also perceive that application of the ordinary meaning of the law to be consistent with their role as just and neutral arbiters of disputes. They may believe that “a sharp focus on the ordinary meaning of statutory language helps judges separate legal interpretation from their personal reaction to the situation being adjudicated” (Eskridge 2016Eskridge Jr., William S. 2016 Interpreting Law: A Primer on How to Read Statutes and the Constitution. St. Paul, MN: Foundation Press., 36). Judges may rely on ordinary meaning aspiring to be faithful agents of the drafters of legal texts, and to avoid imposing their own views on the subject matter of the dispute.

In the context of statutes and constitutions, an appeal to ordinary meaning may be perceived to serve the interests of democratic legitimacy. This is because applying the ordinary meaning of the enacted text of the statute may be perceived as inducing accountability of elected officials for the laws that they enact (Eskridge 2016Eskridge Jr., William S. 2016 Interpreting Law: A Primer on How to Read Statutes and the Constitution. St. Paul, MN: Foundation Press., 37). Under this view, if the judge is bound by the ordinary meaning of the text and reaches an unpopular result when applying that meaning, then citizens subject to the law will hold their representatives accountable and, in theory, vote for new representatives to change the law.

Finally, application of the ordinary meaning of legal texts has been argued to serve a coordinating function, allowing government institutions to coordinate plans that “advance the lives of citizens and the values of the community” (Eskridge 2016Eskridge Jr., William S. 2016 Interpreting Law: A Primer on How to Read Statutes and the Constitution. St. Paul, MN: Foundation Press., 49). Because of the limitations in both training and resources, “judges are not institutionally competent to make judgments about matters of economic policy, political philosophy, or even current affairs” (Eskridge 2016Eskridge Jr., William S. 2016 Interpreting Law: A Primer on How to Read Statutes and the Constitution. St. Paul, MN: Foundation Press., 40). In order for the institutions of government to coordinate governing efforts, statutes and regulations must have a shared understanding – an understanding that can be perceived by generalist judges.

With these justifications in mind, it is understandable the judges would turn to ordinary meaning when interpreting legal texts. Yet the frequency with which judges appeal to the ordinary meaning of legal texts raises a number of questions about the ordinary meaning canon, including: What does ordinary meaning actually mean? And how do judges determine what meanings are ordinary?

2.What is ordinary meaning?

Perhaps the most surprising thing about courts’ appeal to the “ordinary meaning” of legal language is that courts lack a shared, generally agreed upon definition of the phrase. “[I]ronically, we have no ordinary meaning for ‘ordinary meaning’” (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879., 798). Even the definition from Black’s Law Dictionary above claims that ordinary meaning requires “giving the words their ordinary sense” (Garner 2014Garner, Bryan A. ed. 2014 Black’s Law Dictionary. 10th ed. St. Paul, MN: West., 1128). There is an element of circularity in this definition. If courts are applying ordinary meaning to interpret legal texts, and doing so requires the use of the ordinary senses of words, then courts should be asking what does “ordinary” mean with respect to a legal text?

Courts use a variety of words and phrases as apparent synonyms to ordinary meaning, interpreting the words of a legal text according to their “most common” or “common” sense,44. Taniguchi v. Kan Pacific Saipan, 132 S.Ct. 1997, 2000–2003 (2012). or their “possible” or “natural” meaning.55. Mont v. United States, 139 S. Ct. 1826, 1832 (2019) (“the definition of ‘is imprisoned’ may well include pretrial detention.”); id. at 1838 (Sotomayor, J., dissenting) (imprisonment “is most naturally understood in context to mean post-conviction incarceration”). It is not at all clear that these notions of “ordinariness” carry the same meaning. Nor is it clear that when individual judges use the term ordinary meaning they use it consistently. Justice Antonin Scalia, for example, variously characterized “ordinary meaning” as (1) what an ordinary speaker of the English language would think the word means, (2) how an “ordinary Member of Congress” would read a legal text, or (3) as an “‘objectified’ intent–the intent that a reasonable person would gather from the text of the law, placed alongside the remainder of the corpus juris” (McGowan 2008McGowan, Miranda 2008 “Do as I Do, Not as I Say: An Empirical Investigation of Justice Scalia’s Ordinary Meaning Method of Statutory Interpretation.” Mississippi Law Journal 78:129–198. , 132–133). It is not at all clear that these characterizations of ordinary meaning have reference to the same concept.

2.1Context, speech community, and timeframe

Often (though not always) left unspoken when courts invoke the ordinary meaning of the words in a legal text are a collection of assumptions about the context in which the words of the text appear, the speech community that both creates and reads the words of the text, and the timeframe in which the text was drafted. By invoking the ordinary meaning of a word in a text, courts necessarily imply that a word ordinarily carries a given meaning in a context similar to that of the legal text, and when used by some unidentified group of speakers at an unspecified time.

Linguistic context

With respect to context, courts seem to agree that the meaning of the words of a legal text must be understood “in context.” 66. Lorillard Tobacco Co. v. Am. Legacy Found., 903 A.2d 728, 740 (Del. 2006) (“A court must accept and apply the plain meaning of an unambiguous term in the context of the contract language and circumstances, insofar as the parties themselves would have agreed ex ante.”). But courts rarely identify what is meant when context is invoked. In statutory interpretation, context sometimes refers to the historical events to the statutes’ enactment.77. United Steelworkers of America, AFL-CIO-CLC v. Weber, 443 U.S. 193, 201 (1979) (“The prohibition against racial discrimination in §§ 703(a) and (d) of Title VII must therefore be read against the background of the legislative history of Title VII and the historical context from which the Act arose.”). In constitutional interpretation, context sometimes refers to the historical circumstances surrounding the enactment of the Constitution.88. Central Virginia Community College v. Katz, 546 U.S. 356, 362 (2006) (“It is appropriate to presume that the Framers of the Constitution were familiar with the contemporary legal context when they adopted the Bankruptcy Clause”). And in contract interpretation, context may refer to extratextual background information regarding “the subject matter of the transaction, preliminary negotiations and statements made therein, usages of trade, and the course of dealing between the parties” (American Law Institute 1973American Law Institute 1973 Restatement of the Law, Second: Contracts 2d. § 212 cmt. b. St. Paul, MN: American Law Institute Publishers., § 212 cmt. b.)

In addition to these extratextual notions of context, courts often invoke context simply in reference to “[t]he surrounding text of a word or passage, used to determine the meaning of that word or passage.” (Garner 2014Garner, Bryan A. ed. 2014 Black’s Law Dictionary. 10th ed. St. Paul, MN: West., 386). What emerges, then, are at least two related notions of context–one textual and one extratextual (Duranti and Goodwin 1992Duranti, Alessandro and Charles Goodwin 1992 “Rethinking Context: An Introduction.” In Rethinking Context: Language as an Interactive Phenomenon, edited by Alessandro Duranti and Charles Goodwin, 1–42. Cambridge: Cambridge University Press.). The verbal context in which a given provision of a legal text occurs may include both its syntactic environment and may include certain semantic information. Syntax, of course, concerns itself with the way words are arranged in a sentence to convey meaning (Chomsky 1957Chomsky, Noam 1957 Syntactic Structures. The Hague: Mouton de Gruyter. , 11). To understand the meaning of a word or phrase, interpreters should look for examples of the word or phrase in a similar syntactic environment. Semantics is the study of meaning at the word or phrase level (Morris 2006Morris, Michael 2006 An Introduction to the Philosophy of Language. Cambridge: Cambridge University Press. , 152). Interpretation must also take into account the semantic features of a word that have bearing on meaning (Saeed 2015Saeed, John I. 2015 Semantics. 4th ed. Hoboken, NJ: Wiley-Blackwell., 260–265). A usage-based evaluation of contractual meaning may also endeavor to evaluate usage evidence that shares semantic features with the word or phrase in the contract.

Timeframe

Courts are sometimes called upon to interpret texts many years after they are drafted. Sometimes courts acknowledge that language change may have a bearing on the interpretation of these texts. Where the text at issue is an older instrument, a method that accounts for the temporal dimension of interpretation may be required. Consider the following four definitions of car, listed in chronological order in the Oxford English Dictionary:

1.a

A wheeled, usually horse-drawn conveyance; a carriage, cart, or wagon.

2.a

The passenger compartment of a balloon, airship, cableway, etc.; a gondola.

3.a

A railway carriage or wagon….

4.

motor car n. 2. Now the usual sense. 99.“car, n.1”. OED Online. May 2021. Oxford University Press, http://​www​.oed​.com​/view​/Entry​/27674​?rskey​=6D1tvb​&result​=1​&isAdvanced​=false (accessed May 28, 2021).

A legal text that addresses, for example, the importation of cars, but that was drafted in a distant point in history, would have a different meaning than one drafted today. Courts often attempt to take into account the possibility of language change by interpreting texts according to the meaning of the words in the text at the time the text was drafted.1010.This is true of statutes, Dellmuth v. Muth, 491 U.S. 223 (1989) (quoting Perrin v. United States, 44 U.S. 37, 42 (1979) (“As usual, our job is to interpret the words consistent with their ‘ordinary meaning… at the time Congress enacted the statute.’”)), of contracts, Winn-Dixie Stores, Inc. v. Dolgencorp, LLC, 746 F.3d 1008, 1022 (11th Cir. 2014) (quoting Moore v. Stevens, 106 So. 901, 903 (Fla. 1925) (“[W]ords used must be given their ordinary, obvious meaning as commonly understood at the time the instrument containing the covenants was executed ….”, emphasis added), and the Constitution, see Solum, discussing the Fixation Thesis–the notion that “[t]he meaning of the constitutional text is fixed when each provision is framed and ratified” (2015Solum, Lawrence B. 2015 “The Fixation Thesis: The Role of Historical Fact in Original Meaning.” Notre Dame Law Review 91(1): 1–78. , 1). Language is constantly, naturally changing (Lyons 1968Lyons, John 1968 Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press. , 43). Yet language change does not occur at a constant, predictable rate (Crowley and Bowern 2011Crowley, Terry and Claire Bowern 2011 An Introduction to Historical Linguistics. 4th ed. Oxford: Oxford University Press., 149–151).

Not everyone agrees that the interpretation of a legal text should focus on the time of drafting. As Professor William S. Eskridge Jr. has observed: “When Congress has basically dropped a problem into the collective judicial lap, with imprecise and only general directions, then it makes sense for courts to develop that statute in accordance with contemporary, rather than purely historical, policy” (Eskridge 1987Eskridge Jr., William S. 1987 “Dynamic Statutory Interpretation.” University of Pennsylvania Law Review 135:1479–1555. , 1517). However, as discussed below, some features of current language use cannot predictably be intuited. Even a present-day interpretation of a legal text may be better informed by evidence of historical usage.

Ordinary meaning and speech community

When judges make claims about ordinary meaning, they necessary imply that there is some community of language users for which a given meaning would be regarded as ordinary. There must be some speech community–some group that “share[s] values and attitudes about language use, varieties and practices”–that ordinarily uses or understands the words in question in the way the court insists is ordinary (Morgan 2014Morgan, Marcyliena H. 2014 Speech Communities: Key Topics in Linguistic Anthropology. Cambridge University Press: Cambridge. , 1). Stated another way, a speech community is “[a]ny human aggregate characterized by regular and frequent interaction by means of a shared body of verbal signs and set off from similar aggregates by significant differences in language usage” (Gumperz 1968Gumperz, J. J. 1968 “The Speech Community.” In International Encyclopedia of the Social Sciences, edited by David L. Sills and Robert K. Merton, 381–386. New York, NY: Macmillan., 381). It is not difficult to imagine legal documents in which differing linguistic conventions of the different communities involved in drafting or interpreting the text might lead to different understandings of similar language.

Sometimes, the law seems to take into account differences in speech community. When interpreting a statute, courts will sometimes apply a rule of lenity, a canon of interpretation that states that “a court, in construing an ambiguous criminal statute that sets out multiple or inconsistent punishments, should resolve the ambiguity in favor of the more lenient punishment” (Garner 2014Garner, Bryan A. ed. 2014 Black’s Law Dictionary. 10th ed. St. Paul, MN: West., 1532). This means that courts will apply the interpretation of an ambiguous statute in favor of a defendant, though not necessarily one that reflects the linguistic conventions of the defendant’s community. Some contracts are executed between sophisticated commercial parties, with specialized knowledge of both contracting conventions of a particular industry and the characteristics of the underlying industry itself. Other contracts are more public-facing and are executed between firms and their customers. Some such differences are anticipated by interpretive rules in the context of the Uniform Commercial Code, where variations in trade usage are taken into account (White 1995White, James J. 1995 Uniform Commercial Code. St. Paul, Minn.: West., § 1–303 (defining usage of trade)). In such cases, where the question is whether competing industries, different geographical regions, lawyers or their clients, or sophisticated and unsophisticated parties use a given term differently, we might look to evidence of comparative usage in order to evaluate the plain meaning of contractual language.

3.Finding ordinary meaning

Even if it were the case that courts had a shared, generally agreed upon definition of the “Ordinary Meaning,” it is not clear that courts have any effective means for determining what that ordinary meaning actually is. This is because the tools that U.S. courts have at their disposal for determining ordinary meaning–including the judge’s own linguistic intuition and a variety of dictionaries–may not be reliable guides to ordinary meaning.

3.1Linguistic intuition

Courts sometimes make determinations about the meaning of a legal text without referencing any evidence of meaning extrinsic to the text itself. When courts reason from the text to determine its meaning, we can infer that judges are relying on their linguistic intuition. Yet human linguistic intuition has limitations that can frustrate the intention of judges to find a neutral, generally applicable meaning of the words in a text. To begin with, the most common words in a given language tend to be highly polysemous–that is, they tend to have a greater range of possible meanings (Crossley, Salsbury and McNamara 2010Crossley, Scott, Tom Salsbury, and Danielle McNamara 2010 “The Development of Polysemy and Frequency Use in English Second Language Speakers.” Language Learning. A Journal of Research in Language Studies 60:573–605. , 575). Put simply, the more commonly a word is used, the more likely it is to have many different senses. Yet language users are not particularly adept at objectively and predictably identifying and resolving lexical ambiguities when faced with high-frequency, highly polysemous words. The more common a word is, the more senses it has, and the more senses it has, the more likely two people are to disagree as to its meaning in a given context (Agirre and Edmonds 2007Agirre, Eneko and Phillip Edmonds 2007 Word Sense Disambiguation: Algorithms and Application. Chambersburg, PA: Springer., 91; Tsatsornis, Varlamis, and Nørvåg 2010Tsatsornis, George, Iraklis Varlamis, and Kjetil Nørvåg 2010 “An Experimental Study on Unsupervised Graph-based Word Sense Disambiguation.” Computational Linguistics and Intelligent Text Processing, edited by Alexander Gulbukh, 184–198. Chambersburg, PA: Springer. , 193; Brown and Allan 2009Brown, Keith and Keith Allan 2009 Concise Encyclopedia of Semantics. Oxford: Elsevier., 224). This leads to the counterintuitive result that judges and lawyers are more likely to disagree about the meaning of common words than the meaning of uncommon words.

In addition, the objectivity of a judge’s linguistic judgments may be limited by a number of cognitive biases, including false consensus bias–where parties systematically overestimate the degree to which they are likely to agree with other parties about the interpretation of a legal text (Solan, Rosenblatt, and Osherson 2008Solan, Lawrence, Terri Rosenblatt, and Daniel Osherson 2008 “False Consensus Bias in Contract Interpretation.” Columbia Law Review 108:1268–1300.). This phenomenon is referred to as false consensus bias. Courts and parties may fail to recognize false consensus bias and may, therefore, fail to recognize the legitimacy of different readings of the contract. As a consequence, “a judge may consider language to be plain when in fact different people do not understand it the same way, and this may happen even when the judge’s understanding is shared only by a minority of people in general” (Solan, Rosenblatt, and Osherson 2008Solan, Lawrence, Terri Rosenblatt, and Daniel Osherson 2008 “False Consensus Bias in Contract Interpretation.” Columbia Law Review 108:1268–1300., 1294). In addition, judges’ initial conclusions about the meaning of a text may be clouded by confirmation bias–the tendency to credit evidence that supports the judges’ prior commitments and ignore or undervalue contrary evidence (Galotti 2009Galotti, Kathleen M. 2009 Cognitive Psychology: In and Out of the Laboratory. Newbury Park, CA: SAGE., 357; Eysenck 2001Eysenck, Michael W. 2001 Principles of Cognitive Psychology. London: Psychology Press., 373).

Biases like false consensus bias and confirmation bias may prevent judges not only from reasoning objectively about the meaning of the text, but from recognizing the subjectivity of their own reasoning. In such circumstances, “[i]f one person says that both proposed readings of a statute seem plausible, and a colleague disagrees, finding one reading too strained, what is there to do about it but for each to stamp his foot?” (Farnsworth, Guzior, and Malani 2010Farnsworth, Ward, Dustin F. Guzior, and Anup Malani 2010 “Ambiguity About Ambiguity: An Empirical Inquiry into Legal Interpretation.” Journal of Legal Analysis 2:257–300. , 271).

3.2Dictionaries and the “Baffled Judge”

U.S. courts often turn to dictionaries in order to determine ordinary meaning. United States Supreme Court Justice Robert Jackson referred to the use of dictionaries as “the last resort of the baffled judge.”1111.Jordan v. De George, 341 U.S. 223, 234 (1951) (Jackson, J., dissenting). But dictionaries do not define words according to their ordinary meaning. Ordinary meaning is a legal concept, not a linguistic or lexicographical one. The editors of the Webster’s Third New International Dictionary make clear that their dictionary “does not evaluate senses or establish an enduring hierarchy of importance among them” and that the best sense is the one that most aptly fits the context of an actual genuine utterance” (Gove 1971Gove, Philip Babcock 1971 Webster’s Third New International Dictionary of the English Language Unabridged, Springfield, MA: G. & C. Merriam Company., 17a).

But even if dictionary editors set out to define the ordinary meaning of the words in legal texts, dictionary editors cannot possibly anticipate all of the contexts in which a word will be used in a legal text (Sonpal 2003Sonpal, Rickie 2003 “Old Dictionaries and New Textualists.” Fordham Law Review 71:2177–2226., 2206). As Professors Henry Hart and Abert Sacks observed in their influential Legal Process lectures: “A dictionary, it is vital to observe, never says what meaning a word must bear in a particular context” (Hart and Sacks 1994Hart, Henry M. and Albert M. Sacks 1994 The Legal Process: Basic Problems in the Making and the Application of Law, edited by William N. Eskridge Jr. and Philip P. Frickey. St. Paul, MN: Foundation Press.). Dictionaries are merely repositories of recorded senses–of the ways in which a word has been used. The production of dictionaries is a decidedly human endeavor. Many dictionaries relied upon by judges were assembled without the use of language evidence from linguistic corpora. Instead, they were assembled with usage evidence from citation files that are liable to be “unrepresentative of the language as a whole,” and reviewed by editors who “all too often ignore common usages and give disproportionate attention to uncommon ones ….” (Landau 2001Landau, Sidney I. 2001 Dictionaries: The Art and Craft of Lexicography. 2nd ed. Cambridge: Cambridge University Press., 104; Biber, Conrad, and Reppen 1998Biber, Douglas, Susan Conrad, and Randi Reppen 1998 Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press. , 26).

No dictionary is a complete repository of definitions of every word, or of every sense in which a word has been used. Lexicographers have estimated the total number of English words at two million or more:

[e]ven the largest unabridged American dictionaries contain well under half this total (…). Abridged and collegiate dictionaries include a smaller number of words and–more relevant–they contain fewer and shorter definitions for these words. (…) all dictionaries must deal with space limitations that affect the usage listings and amplifications contained in their definitions.(Brudney and Baum 2013Brudney, James J. and Lawrence Baum 2013 “Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras.” William & Mary Law Review 55:483–580. , 513)

Sometimes, courts will support a claim about ordinary meaning by citing to a number of dictionaries. However, if dictionaries do not contain ordinary meaning, then citing multiple dictionaries simply merely compounds the problem. It is not always clear that dictionaries have arrived at their definitions independently–“[t]he history of English lexicography usually consists of a recital of successive and often successful acts of piracy” (Landau 2001Landau, Sidney I. 2001 Dictionaries: The Art and Craft of Lexicography. 2nd ed. Cambridge: Cambridge University Press., 43). This is because “[d]ictionary editors look at each other’s books, and though editors form their own opinions about what ground should be covered, they dare not depart too far from the area laid out by their competitors.” (Ibid. 402).

Sometimes courts will appeal to the ranking of senses in a dictionary, claiming that a given sense should be given priority because it is listed “first” in the dictionary (Mouritsen 2010Mouritsen, Stephen 2010 “The Dictionary Is Not a Fortress: Definitional Fallacies and a Corpus-Based Approach to Plain Meaning.” BYU Law Review 5:1915–1980., 1924–1929). Yet the dictionaries most commonly relied upon by judges do not rank their senses according to ordinariness. Courts will also sometimes arbitrarily choose a single sense from a dictionary and insist they have found the ordinary meaning, all the while ignoring other competing senses.1212. State v. Rasabout, 356 P.3d 1258, 1274 (Utah 2015) (Lee, Associate C.J., concurring in part and concurring in the judgment).

U.S. courts do not have a principled or well-defined basis for choosing between legal dictionaries and general-use dictionaries (Brudney and Baum 2013Brudney, James J. and Lawrence Baum 2013 “Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras.” William & Mary Law Review 55:483–580. , 510). Courts have failed to provide any principled basis for selecting among different types of dictionaries (Brudney and Baum 2013Brudney, James J. and Lawrence Baum 2013 “Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras.” William & Mary Law Review 55:483–580. , 507). In addition, when called upon to interpret an older text, courts will sometimes appeal to a dictionary that was published around the same time that the legal text in question was drafted. While intuitively appealing, this practice has a number of flaws.

The listed publication date of a dictionary does not always tell you when that dictionary was written. Some historical dictionaries were merely reprinted with a new addition and were not updated to reflect contemporary usage (Sonpal 2003Sonpal, Rickie 2003 “Old Dictionaries and New Textualists.” Fordham Law Review 71:2177–2226., 2209–2210). Additionally, some dictionaries may rely on examples “composed centuries before the dictionary was compiled” (Sonpal 2003Sonpal, Rickie 2003 “Old Dictionaries and New Textualists.” Fordham Law Review 71:2177–2226., 2207). The editors may simply copy usage examples from prior dictionaries (Reed 1962Reed, Jr., Joseph W. 1962 “Noah Webster’s Debt to Samuel Johnson.” American Speech 37:95–105. , 95). Courts do not have explicit, well-defined rules about when they will appeal to a historical dictionary (Brudney and Baum 2013Brudney, James J. and Lawrence Baum 2013 “Oasis or Mirage: The Supreme Court’s Thirst for Dictionaries in the Rehnquist and Roberts Eras.” William & Mary Law Review 55:483–580. , 511).

Faced with these challenges in legal interpretation, Professor Arthur L. Corbin observed:

It is true that when a judge reads the words of a contract he may jump to the instant and confident opinion that they have but one reasonable meaning and that he knows what it is. A greater familiarity with dictionaries and the usages of words, a better understanding of the uncertainties of language, and a comparative study of more cases in the field of interpretation, will make one beware of holding such an opinion so recklessly arrived at.(Corbin 1952Corbin, Arthur L. 1952 Corbin on Contracts. One Volume Edition. St. Paul, Minn.: West., § 535)

Dictionaries can still be useful to legal interpreters, just not in the way that they are typically used. Dictionaries can be used by judges to define unknown terms, the use of which is almost certainly uncontroversial (Hoffman 2003Hoffman, Craig 2003 “Parse the Sentence First: Curbing the Urge to Resort to the Dictionary when Interpreting Legal Texts.” New York University Journal of Legislation and Public Policy 6:401–438., 416; Solan 1993Solan, Lawrence 1993 “When Judges Use the Dictionary.” American Speech 68(1):50–57. , 55). Dictionaries can also be used to confirm that a given sense of a word is attested–that is, that the sense of a word advocated by the parties was not invented by the parties but has a documented record of prior usage. In this respect, unabridged dictionaries are “historical records… of the meanings with which words have in fact been used by writers of good repute. They are often useful in answering hard questions of whether, in an appropriate context, a particular meaning is linguistically permissible” (Hart and Sacks 1994Hart, Henry M. and Albert M. Sacks 1994 The Legal Process: Basic Problems in the Making and the Application of Law, edited by William N. Eskridge Jr. and Philip P. Frickey. St. Paul, MN: Foundation Press., 1375–1376). Dictionaries can give interpreters a sense of the range of possible uses a given word may have had. Dictionaries can also model for judges, who are often called upon to make fine distinctions among different senses, how to describe differences in meaning. Yet dictionaries cannot tell their readers the meaning a word must bear in a particular context and in a particular legal text because they do not contain ordinary meaning.

4.Evaluating ordinary meaning claims with linguistic corpora

Recent scholarship has both recognized these difficulties with both conceptualizing ordinary meaning and evaluating claims about ordinary meaning and has suggested the use of language evidence from linguistic corpora as a tool for addressing these difficulties with ordinary meaning (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879.).

Corpus linguistics is the study of language using evidence from electronic collections of texts (McEnery and Hardie 2012McEnery, Tony and Andrew Hardie 2012 Corpus Linguistics: Method, Theory and Practice. Cambridge: Cambridge University Press.). Linguistic corpora are designed to be representative samples of the speech or writing of a speech community (McEnery and Wilson 2011McEnery, Tony and Andrew Wilson 2011 Corpus Linguistics: An Introduction. Edinburgh: Edinburgh University Press. , 75; Morgan 2014Morgan, Marcyliena H. 2014 Speech Communities: Key Topics in Linguistic Anthropology. Cambridge University Press: Cambridge. , 1). They are designed “to actually ‘represent’ a domain of language use with a corpus of texts” (Biber and Reppen 2015Biber, Douglas and Randi Reppen 2015 The Cambridge Handbook of English Corpus Linguistics. Cambridge: Cambridge University Press. , 1). These corpora can allow their users to make observations about language that cannot be made through introspection (McEnery and Wilson 2011McEnery, Tony and Andrew Wilson 2011 Corpus Linguistics: An Introduction. Edinburgh: Edinburgh University Press. , 12), and to gather evidence of language use that is replicable and falsifiable (McEnery and Hardie 2012McEnery, Tony and Andrew Hardie 2012 Corpus Linguistics: Method, Theory and Practice. Cambridge: Cambridge University Press., 66). Corpora can be used to show evidence of how frequently a keyword occurs, and how frequently a given sense of a keyword is used in a particular context, by a particular speech community, and at a particular point in history.

An example of how to use linguistic corpora to evaluate claims about ordinary meaning can be found in the literature discussing the United States Supreme Court’s decision Muscarello v. United States, 524 U.S. 125 (1998). The Muscarello decision has received a lot of academic attention in the law and corpus linguistics literature (Goldfarb 2017Goldfarb, Neal 2017 “A Lawyer’s Introduction to Meaning in the Framework of Corpus Linguistics.” Brigham Young University Law Review 6:1359–1417. ; Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879.; Mouritsen 2010Mouritsen, Stephen 2010 “The Dictionary Is Not a Fortress: Definitional Fallacies and a Corpus-Based Approach to Plain Meaning.” BYU Law Review 5:1915–1980.; Solan and Gales 2017Solan, Lawrence M. and Tammy Gales 2017 “Corpus Linguistics as a Tool in Legal Interpretation.” Brigham Young University Law Review 2017:1311–1358.), at least in part because Muscarello was the first Supreme Court decision that involved “crudely … searching computerized newspaper data bases,” in order to “make certain that there is no special ordinary English restriction” on the meaning of the text in question (Muscarello v. United States, 524 U.S. 125, 129 (1998)).

The Muscarello case required the interpretation of the phrase carries a firearm in the Omnibus Crime Control and Safe Streets Act of 1966 (18 U.S.C. § 924(c)(1)). Specifically, the case required a determination of whether a person “carries a firearm” when the firearm is locked in a glovebox, but not being held (Ibid. 126–127). Frank Muscarello was arrested during a minor drug transaction, but at the time of his arrest he had a handgun locked in his glovebox (Ibid. 127). A five-to-four majority of the Court concluded that Section 924(c) suggested the conveyance in a vehicle meaning (Ibid. 139).

The Muscarello majority framed the dispute by stating that “[a]lthough the word ‘carry’ has many different meanings, only two are relevant here” (Ibid. 128). The Court then noted that carry’s “first, or primary, meaning” reflects the notion carrying as conveyance (hereafter vehicle-carry); while only a “different, rather special” meaning of carry reflects the notion of carrying upon one’s person (hereafter personal-carry) (Ibid. 128).

The Court referenced a number of dictionaries to conclude vehicle-carry is carry’s ordinary meaning. They made reference not only to how the word was defined, but whether the definition in question was listed first in the dictionary, and which of the contested definitions were consistent with the word’s etymology (Ibid. 127–128, 131).

But the dictionaries most cited by the Supreme Court, including Webster’s Third New International Dictionary and Oxford English Dictionary, rank their senses historically–oldest to newest–and not according to their “ordinariness” (Gove 1971Gove, Philip Babcock 1971 Webster’s Third New International Dictionary of the English Language Unabridged, Springfield, MA: G. & C. Merriam Company., 17a; Oxford English Dictionary 1989Oxford English Dictionary 1989 2nd ed. Oxford: Oxford University Press., xxix). Moreover, Webster’s Third, makes clear that “[s]ometimes an arbitrary arrangement or rearrangement is the only reasonable and expedient solution to the problems of ordering senses” (Gove 1971Gove, Philip Babcock 1971 Webster’s Third New International Dictionary of the English Language Unabridged, Springfield, MA: G. & C. Merriam Company., 17a). Consequently, the editors of Webster’s Third also counsels that “[t]he best sense is the one that most aptly fits the context of an actual genuine utterance” (Gove 1971Gove, Philip Babcock 1971 Webster’s Third New International Dictionary of the English Language Unabridged, Springfield, MA: G. & C. Merriam Company., 17a).

The Muscarello Court also appealed to the etymology of carry, citing an etymological dictionary for the proposition that the word carry comes from the Latin carum meaning car or cart (Muscarello v. United States, 524 U.S. 125, 128 (1998)). Yet appealing to the etymology of a word to determine its contemporary meaning is obviously fallacious. This is an error in reasoning so common that it has its own name–the Etymological Fallacy (Sihler 2000Sihler, Andrew L. 2000 Language History: An Introduction. Philadelphia: John Benjamins. , 131). If words were interpreted according to their etymology, then December would be the tenth month, not the twelfth.1313.See December, Online Etymology Dictionary, https://​www​.etymonline​.com​/search​?q​=December (noting that December was the “tenth month of the old Roman Calendar” and comes from the Latin decem).

Instead of relying on dictionaries for information they do not contain in such cases, advocates for the use of corpus linguistics in legal texts have proposed gathering evidence of the use of the verb carry from a corpus, particularly in the circumstances relevant here where there is an implied human agent and a firearm as the thing that is being carried.1414.This section is drawn largely from the discussion of Muscarello in Lee and Mouritsen (2018)Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879.. One recent study looked at evidence of the way carry was used in this context, using the News on the Web (NOW) Corpus. The NOW Corpus was employed to provide a “randomized sample of concordance lines featuring carry” in circumstances with similar linguistic features to § 924(c)(1).1515.In the context of Section 924(c), carry is a transitive verb (or “has a transitive argument structure”), with an implied “human subject and a non-human, inanimate, weapon object.” A search in a corpus can reveal uses of the verb carry in similar contexts. Id. at 833, 846.

The authors searched the NOW Corpus for instances of carry within a few words of firearm (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879., 846 n.237). They then broadened the search to include the most common synonyms of firearm that can be revealed by looking at the collocates of carry, like gun(s), pistol(s), handgun(s), and rifle(s) (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879., 847). Of the 271 instances of carry reviewed, some 104 instances of carry had the personal-carry sense while only five instances of carry had the vehicle-carry meaning. Similarly, the authors used the Corpus of Historical American English (COHA) to provide contemporaneous usage evidence from the decade of the statute’s enactment, the 1960s. The COHA includes a collection of annotated texts from each decade going back to 1810, with somewhere between 10 and 20 million words per decade. That means that fewer examples of the use of carry are available in the COHA. There are some twenty-eight instances of carry co-occurring with firearm(s), gun(s), pistol(s), handgun(s), or rifle(s) in the COHA. Of these, eighteen were instances of personal-carry and two were instances of vehicle-carry.

Evidence from these two corpora suggests that in contexts similar to Section 924(c)(1), personal-carry occurs much more commonly than vehicle-carry. Whatever is meant by the “ordinary meaning” of carry, it cannot mean that vehicle-carry is the most common way that the phrase carries a firearm is used.

Another study examining the same case also found that evidence from corpora reveal that “carry is used more frequently to talk about acts of personally carrying objects (for instance, in one’s hands or arms, or strapped to one’s back), than about events in which objects are transported or carried in a vehicle” (Goldfarb 2017Goldfarb, Neal 2017 “A Lawyer’s Introduction to Meaning in the Framework of Corpus Linguistics.” Brigham Young University Law Review 6:1359–1417. , 1408). When speakers use carry in the vehicle-carry sense, they tend to make explicit reference to the vehicle in question (Goldfarb 2017Goldfarb, Neal 2017 “A Lawyer’s Introduction to Meaning in the Framework of Corpus Linguistics.” Brigham Young University Law Review 6:1359–1417. , 1408; Solan and Gales 2017Solan, Lawrence M. and Tammy Gales 2017 “Corpus Linguistics as a Tool in Legal Interpretation.” Brigham Young University Law Review 2017:1311–1358., 1346).

The corpus evidence can provide a better picture of the way the phrase carries a firearm is used. It can also provide some insight into the circumstances in which the vehicle-carry sense is implicated. This information cannot be found in a dictionary, and may not be available through introspection. The searches performed in a corpus can be replicated by any investigator. Corpora can be designed to represent the speech or writing of any speech community for which there are surviving texts or transcriptions of speech. Corpora can also provide evidence of historical usage for which contemporary language cannot be expected to have accurate instructions.

5.Limitations and future work

While corpora may prove useful in gathering language evidence that can be employed to evaluate claims of ordinary meaning, the use of linguistic corpora to perform these tasks raises a host of questions and presents a number of avenues for future research.

To begin with, as noted repeatedly above, judges and lawyers do not have a shared, well-defined understanding of what ordinary meaning actually means. If corpora are used to gather evidence to evaluate claims about ordinary meaning, then investigators must first determine what the evidence is meant to prove or disprove. One candidate for ordinary meaning (especially in cases of lexical ambiguity) is the frequency with which a given sense of a word occurs. Where a given sense is overwhelmingly more common, we might conclude that sense is the ordinary meaning. This notion of frequency of occurrence is consistent with one of the ways in which the word “ordinary” is defined.1616.“Of language, usage, discourse, etc.: that most commonly found or attested.” ( Oxford English Dictionary 1989Oxford English Dictionary 1989 2nd ed. Oxford: Oxford University Press.). Interpreting the family law word with its most frequent sense, when divorced from context would render the legal text in question entirely incoherent. Professor Richard Fallon recently observed, “As corpus linguistic research recurrently teaches, the most common uses of words or phrases are typically not the only, or the only linguistically eligible, ones” (2019Fallon Jr., Richard 2019 “The Statutory Interpretation Muddle.” Northwestern University Law Review 114:269–334., 269).

Another advantage of corpus studies has been highlighted by Taylor as follows: “Corpus-based studies have also revolutionized our understanding of polysemy, lexical polysemy in the first instance, but also the various semantic values which attach to larger constructions” (2016Taylor, James R. 2016 “Cognitive Linguistics.” In Routledge Handbook of Linguistics, edited by Keith Allan, 455–469. Abingdon: Routledge., 466). Yet as courts and attorneys rely on corpus linguistics in making claims evaluating claims about ordinary meaning, they must also be sensitive to how “[e]ach sense of a word is associated with a distinct lexico-syntactic context,” and how “the context serves to prime the relevant sense, thereby suppressing the other possible readings” (Taylor 2016Taylor, James R. 2016 “Cognitive Linguistics.” In Routledge Handbook of Linguistics, edited by Keith Allan, 455–469. Abingdon: Routledge., 466).

In addition, some emerging scholarship has proposed evaluating claims of ordinary meaning not with corpus evidence, but instead with survey responses (Ben-Shahar and Strahilevitz 2017Ben-Shahar, Omri and Lior J. Strahilevitz 2017 “Interpreting Contracts Via Surveys and Experiments.” New York University Law Review. 92:1753–1827. ; Tobia 2020Tobia, Kevin P. 2020 “Testing Ordinary Meaning.” Harvard Law Review 134:726–806.). These survey methods may involve the systematic collection of linguistic judgments from survey respondents, particularly through modern survey administration platforms like Amazon’s Mechanical Turk (Bedi and Reibstein 2020Bedi, Suneal and David Reibstein 2020 “Measuring Trademark Dilution by Tarnishment.” Indiana Law Journal 95:683–734., 707 n.126). Using survey evidence to evaluate claims of ordinary meaning offers some of the same advantages as using corpus linguistics to perform this task. Surveys and corpora both provide evidence of meaning that may not be available via introspection and that is not recorded in a dictionary. Surveys may be administered to members of any speech community. Surveys also have some advantages over the use of linguistic corpora when examining questions of usage and meaning. The survey architect has greater control of the framing of the question and the linguistic context in which the language under examination will appear. With that said, surveys may prove more costly than examining a usage question in a corpus. Moreover, unlike corpora, surveys may be limited in time because “there are no members of the relevant societies left for us to gather opinions from and there were no public surveys of social attitudes undertaken then” (McEnery, Baker, and Dayrell 2020McEnery, Tony, Helen Samantha Baker, and Carmen Dayrell 2020 “Working at the Interface of Hydrology and Corpus Linguistics.” In Using Corpus Methods to Triangulate Linguistic Analysis, edited by Jesse Egbert and Paul Baker, 52–84. London: Routledge., 54). Corpora may allow us to examine historical usage, but surveys do not.

There is also the challenge of what to do when survey responses and corpus evidence do not match. For example, the United States Supreme Court’s decision in Taniguchi v. Kan Pacific Saipan, Ltd., 566 U.S. 560 (2012), addressed the question of whether a person hired to translate written documents is an interpreter for the purposes of a statute authorizing costs to prevailing parties for interpreters (Ibid. 562, 566). In Taniguchi, the court held that “an interpreter is normally understood as one who translates orally from one language to another” (Ibid. 569). The Court conceded that the text-to-text translation sense of “interpreter” is possible, but concluded that it is “hardly a common or ordinary meaning” (Ibid. 569). The Court even went so far as to characterize this potential sense of interpreter as “obsolete” (Ibid. 569). In examining the usage evidence for interpreter in the NOW Corpus, the authors of one corpus-based study did not find a “single instance of anyone referred to as an interpreter performing a text-to-text translation” of one language to another (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879., 850). This, the authors note, raises the question not merely of whether interpreter is commonly used to refer to text-to-text translation, but whether such a reading is even possible–that is whether the word interpreter is ever used in reference to text-to-text translation (Lee and Mouritsen 2018Lee, Thomas R. and Stephen C. Mouritsen 2018 “Judging Ordinary Meaning.” Yale Law Journal 12(4):788–879., 850).

There is at least some evidence that when presented with interpretive problems similar to those faced by judges, survey respondents will sometimes reach conclusions that are at odds with usage evidence from a corpus (Tobia 2020Tobia, Kevin P. 2020 “Testing Ordinary Meaning.” Harvard Law Review 134:726–806., 764). If usage evidence disagrees with survey responses, can we meaningfully say that one represents “ordinary meaning” and the other does not, or should we simply conclude that the results of either approach are in error? One reason for the difference may be that “[e]ven the best designed elicitation tasks are removed from how people use (and think about) language in everyday life, and people’s reports of their linguistic usage may or may not match up with what they actually do” (Shilling 2014Shilling, Natalie 2014 “Surveys and Interviews.” In Research Methods in Linguistics, edited by Robert J. Podesva and Devyani Sharma, 96–115. Cambridge: Cambridge University Press. , 103). The better question for researchers is not whether corpus evidence and survey responses diverge, but why they diverge. A better understanding of the basis for that divergence will not only help researchers construct better surveys but may also lead to a better theorized notion of ordinary meaning.

Finally, the language recorded in a linguistic corpus will reflect the linguistic biases of the speech community it is intended to study (Jennejohn, Nelson, and Núñez 2021Jennejohn, Matthew, Samuel Nelson, and D. Carolina Núñez 2021 “Hidden Bias in Empirical Textualism.” Georgetown Law Journal 109:767–811.). While corpora may be constructed to represent the linguistic conventions of any speech community, from any region, ethnic or racial background, or socioeconomic background for which there are existing texts or transcribed speech, it is nevertheless the case that many of the most widely used corpora (particularly in the law and corpus linguistics literature) are based on written sources made up mostly of books, newspapers, and magazines. These corpora will necessarily privilege those who write books, newspapers, and magazines, which will create a bias in favor of language usage from mostly white and mostly male writers and speakers. This effect will be further amplified the more one examines historical corpus evidence. In this respect, the corpus does not necessarily create the bias, but reports biases that already exist. However, using comparative corpora, the corpus user can also use comparative corpora to identify, study, and record such linguistic bias, and examine differences in language usage in different communities. Bespoke corpora may be constructed with the intention of representing the linguistic conventions of a variety of speech communities and combating bias in legal interpretation.

6.Conclusions

For centuries judges and lawyers have appealed to plain or ordinary meaning when attempting to puzzle out the meaning of a legal text. Yet judges have appealed to ordinary meaning without a shared understanding of what ordinary meaning actually means, and without effective methods for evaluating claims of ordinary meaning. Evidence of language use from linguistic corpora may allow judges and lawyers to have a more meaningful discussion about what is meant by a legal text’s ordinary meaning and may provide jurists with more evidence-based method for evaluating claims about ordinary meaning. Yet jurists must proceed cautiously, taking account of both linguistic context and speech community, when relying on corpus evidence to evaluate claims of meaning.