Justice PARRISH, opinion of the Court

Sean D. Reyes, Att'y Gen., Ryan D. Tenney, Asst. Att'y Gen., Salt Lake City, for respondent.

Debra M. Nelson, Daniel Torrence, Salt Lake City, for petitioner.

Justice PARRISH authored the opinion of the Court, in which Justice DURHAM and Judge HARRIS joined in full, Chief Justice DURRANT joined except as to Part I.C, and Associate Chief Justice LEE joined except as to Part I.A, I.B, and I.C. Chief Justice DURRANT authored an opinion concurring in part and concurring in the judgment. Associate Chief Justice LEE authored an opinion concurring in part and concurring in the judgment. Due to his retirement, Justice NEHRING did not participate herein; District Court Judge RYAN HARRIS sat. Justice DENO G. HIMONAS became a member of the Court on February 13, 2015, after oral argument in this matter, and accordingly did not participate.

Justice PARRISH, opinion of the Court:

INTRODUCTION

¶ 1 Andy Rasabout fired twelve shots at a house in a gang-related drive-by shooting. A jury convicted him of twelve felony counts of unlawful discharge of a firearm, in violation of Utah Code section 76–10–508. The trial court merged the twelve counts into one, but the court of appeals reversed. We granted Mr. Rasabout's petition for certiorari and now affirm the ruling of the court of appeals, holding that the allowable unit of prosecution for unlawful discharge of a firearm is each discrete shot.

BACKGROUND

¶ 2 Mr. Rasabout is a member of the street gang known as the Tiny Oriental Posse. On November 1, 2007, Mr. Rasabout, riding shotgun in a Honda Civic, fired twelve shots from a Glock 9 mm semiautomatic pistol at a house and a car parked in front. Lee Tran, whom Mr. Rasabout knew to be a rival in the Original Laotian Gangsters, owned the car and was inside the house at the time. But Mr. Tran was not the only person in danger. Two young girls and their mother were asleep upstairs. Several others were playing cards in the basement. And one man was standing in the carport, enjoying the crisp morning air and a cigarette.

¶ 3 A jury convicted Mr. Rasabout of twelve felony counts of unlawful discharge of a firearm, in violation of Utah Code section 76–10–508. At Mr. Rasabout's request, the trial court merged the twelve counts and sentenced Mr. Rasabout on the basis of one conviction. The court of appeals concluded that the trial court erred and ordered that court to resentence Mr. Rasabout on all twelve convictions. We granted Mr. Rasabout's petition for certiorari and have jurisdiction pursuant to Utah Code section 78A–3–102(3)(a).

STANDARD OF REVIEW

¶ 4 On certiorari to the Utah Court of Appeals, we review the decision of that court for correctness; we do not review the decision of the trial court.

ANALYSIS

¶ 5 When Mr. Rasabout fired twelve shots into a house that he knew to be occupied by a rival gang member, Utah Code section 76–10–508 provided,

(1)(a) A person may not discharge any kind of dangerous weapon or firearm:

(i) from an automobile or other vehicle;

(ii) from, upon, or across any highway;

....

(vii) without written permission ... within 600 feet of ... a house, dwelling, or any other building;

....

(2) A violation of any provision of this section is a class B misdemeanor unless the actor discharges a firearm under any of the following circumstances ..., in which case it is a third degree felony and the convicted person shall be sentenced to an enhanced minimum term of three years in prison:

(a) the actor discharges a firearm in the direction of any person or persons, knowing or having reason to believe that any person may be endangered;

(b) the actor, with intent to intimidate or harass another or with intent to damage a habitable structure ..., discharges a firearm in the direction of any building; or

(c) the actor, with intent to intimidate or harass another, discharges a firearm in the direction of any vehicle.

Pursuant to this statute, a jury convicted Mr. Rasabout of twelve felony counts of unlawful discharge of a firearm. The trial court merged the twelve counts into one. The court of appeals reversed this ruling, reasoning that the allowable unit of prosecution for the offense is each discrete shot. Thus, the court concluded that the twelve discrete shots Mr. Rasabout fired support twelve convictions. We affirm the court of appeals.

¶ 6 Our analysis proceeds in three parts: First, we hold that the allowable unit of prosecution for unlawful discharge of a firearm is each discrete shot; accordingly, Mr. Rasabout was properly convicted of and may be punished for twelve separate counts because he fired twelve discrete shots. Second, we reject as inapplicable the single criminal episode statute and the single larceny rule, on which Mr. Rasabout relies. Third, we hold that the state and federal cruel and unusual punishment clauses are inapplicable because those clauses deal with punishment, which is distinct from the allowable unit of prosecution for an offense.

I. MR. RASABOUT WAS PROPERLY CONVICTED OF TWELVE COUNTS OF UNLAWFUL DISCHARGE OF A FIREARM BECAUSE HE FIRED TWELVE DISCRETE SHOTS

¶ 7 At the heart of this case is a single question: What is the allowable unit of prosecution for the crime of unlawful discharge of a firearm? The answer to that question determines whether Mr. Rasabout was permissibly convicted of twelve counts. The court of appeals concluded that “the unit of prosecution under the firearm discharge statute is each discrete shot.” Mr. Rasabout argues that this conclusion was in error. He points instead to the intent required by the enhancement provision and argues that the allowable unit of prosecution is the continuous intent that motivates one or more shots. He reasons that he violated the statute only once because a single continuous intent motivated him to fire all twelve shots. The State disagrees, arguing that the Legislature criminalized each discrete shot. We agree with the State and therefore affirm the court of appeals.

A. Identifying the Allowable Unit of Prosecution for an Offense

¶ 8 The allowable unit of prosecution for an offense determines whether a perpetrator's conduct constitutes one or more violations of that offense. Take the example of child pornography. It is a crime to “intentionally ... view [ ] child pornography.” If a perpetrator views multiple images of multiple victims over a period of time, how many times has he committed the offense? Perhaps there is one violation for each viewing session, regardless of the number of images or victims. Or maybe there is one violation for each victim or one for each image. The allowable unit of prosecution provided by the offense resolves this question. In the case of child pornography, the Legislature has provided that “[i]t is a separate offense under this section: (a) for each minor depicted in the child pornography; and (b) for each time the same minor is depicted in different child pornography.”

¶ 9 But not all statutes explicitly define the allowable unit of prosecution. State v. Morrison, which dealt with a prior version of the child-pornography statute, illustrates the appropriate analysis when the statute at issue is not so explicit. In that case, the defendant contended that he violated the child-pornography statute only once even though he possessed multiple visual representations of child pornography. Although the statute did not explicitly define the allowable unit of prosecution, we nevertheless focused on the statutory language that defined the offense as “ ‘knowingly ... possess[ing] ... material ... depicting’ ” child pornography and defined “material” as “ ‘any visual representation.’ ” We reasoned that “[t]he clearest reading of the statute is that each individual ‘visual representation’ of child pornography that is knowingly possessed by a defendant constitutes the basis for a separate offense.” Accordingly, on the basis of the statutory language, we held that the allowable unit of prosecution for child pornography is each visual representation.

¶ 10 In short, identifying the allowable unit of prosecution for an offense is a question of statutory construction. And when construing a statute, we seek to give effect to the intent of the Legislature. Indeed, this has been our practice from the time of statehood until now. To ascertain that intent, we look first to the text of the statute within its context. And while the ordinary meaning of a word is powerful evidence in understanding statutory text, it is not the only consideration because it is simply inconclusive as to the meaning intended in a particular context. Even if a word bears one meaning in the majority of cases where that word is used proximate to another, that does not foreclose the possibility that the Legislature intended a less-preferred meaning in a particular context. We therefore begin with the plain language of the provision at issue in our broader effort to ascertain the intent of the Legislature disclosed by the language of the act as a whole, the act's operation, and its purpose.

B. The Allowable Unit of Prosecution for Unlawful Discharge of a Firearm is Each Discrete Shot

¶ 11 The unlawful-discharge-of-a-firearm statute provides that “[a] person may not discharge any kind of dangerous weapon or firearm” in certain dangerous conditions. The allowable unit of prosecution turns on the meaning of the term “discharge” in the context of a “dangerous weapon or firearm.”

¶ 12 The word discharge is made up of two parts: a prefix and the root word. Adding the prefix dis provides the negative or opposite meaning of the root word. The root word charge includes the verb meaning “to place a charge (as of powder) in” and the noun meaning “the quantity of explosive used in a single discharge.” And the dictionary entry of the word discharge includes the meaning “to release from confinement” or simply to shoot. The definition of shoot confirms its synonymous meaning in the firearm context; it means “to eject or impel ... by a sudden release of tension” or “to drive forth ... by an explosion (as of a powder charge in a firearm ... ).” Under these dictionary entries, the clearest reading of the statute is that discharging a weapon or firearm means shooting a weapon or firearm.

¶ 13 This understanding is confirmed by the statutory definition applicable to this offense. It defines a firearm as “any device ... from which is expelled a projectile by action of an explosive.” In other words, it contemplates the expulsion of a single projectile with a single explosion. Similarly, that section defines a handgun as a “firearm of any description ... from which any shot, bullet, or other missile can be discharged, the length of which ... does not exceed 12 inches.” This definition also contemplates a single “shot, bullet, or other missile” that “can be discharged.” These definitions thus confirm that the term discharge contemplates a discrete shot or explosion.

¶ 14 Additionally, it was reasonable for the Legislature to criminalize each shot fired because each shot carries an independent harm. Otherwise, an individual shooting once would have no punitive deterrent from continuing to shoot as many times as possible—essentially a buy one, harm as much as you like discount. Thus, convicting Mr. Rasabout of a separate offense for each shot fired creates an independent punitive deterrent for each pull of the trigger.

¶ 15 In urging his position that the allowable unit of prosecution is not each shot fired, Mr. Rasabout points to the “intent to intimidate” language in subsection (2) of the statute. He argues that this language suggests that the Legislature defined the allowable unit of prosecution for this offense based on the singular intent of the perpetrator, no matter haw many shots are actually fired. We are unpersuaded by Mr. Rasabout's argument for two reasons. First, the “intent to intimidate” language is not part of the provision defining the crime. Rather, it is located in the enhancement provision. Because we presume that the Legislature used the term discharge consistently throughout the statute, it must mean the same thing whether or not the offense is enhanced. Because the word discharge means shoot in the provision defining the crime, it must mean the same thing when the perpetrator discharges a firearm with an “intent to intimidate.” Second, there is nothing about an “intent to intimidate” as an enhancement that suggests that the intimidation supercedes the flying bullet as the principal harm that the Legislature sought to criminalize. Just as each flying bullet is dangerous, so too is each flying bullet that was fired with the intent to intimidate. Accordingly, the allowable unit of prosecution for the unlawful-discharge-of-a-firearm statute under which Mr. Rasabout was convicted is each discrete shot, whether or not the shot is fired with a mens rea that enhances the classification of the offense.

C. Sua Sponte Corpus Linguistics Research is Not Appropriate

¶ 16 Justice Lee charges the court with coming to this conclusion by plucking a definition from a dictionary on the basis of cloaked intuition. Instead, he advances what he deems a “more transparent” interpretation methodology that utilizes corpus linguistics research. We decline to adopt his approach because, among other reasons, it is unfair to the parties and it attempts scientific research that is not subject to scientific review.

¶ 17 To begin, Justice Lee argues that we should decide this case against Mr. Rasabout on the basis of the corpus linguistics research he has conducted sua sponte. But because his rationale is so different in kind from any argument made by the parties, Mr. Rasabout has never had a reasonable opportunity to present a different perspective. This violates the very notion of our adversary system, which “assures fairness by exempting a party from the inequity of [losing] on appeal on a ground that [he] had no opportunity to address.” “[W]e should not dilute [the protections of our adversary system] by stretching their standards to justify our consideration of [an argument] we find interesting or important.” Moreover, deciding this case on the basis of an argument not subjected to adversarial briefing is a recipe for making bad law.

¶ 18 Additionally, it would be entirely inappropriate for this court to conduct the independent scientific research that serves as the basis for Justice Lee's approach. Justice Lee admits that he is not an expert in this field and does not completely understand its methodologies, but asserts that we must “try.” Linguistics is a scientific field of study that uses empirical research to draw findings. And just as with other fields of scientific study, simply trying harder will not lead us to a better answer. The knowledge and expertise required to conduct scientific research are “usually not within the common knowledge” of judges, so “testimony from relevant experts is generally required in order to ensure that [judges] have adequate knowledge upon which to base their decisions.” We regularly refuse to conduct legal research, a field in which we are experts. We should similarly refuse our inclination to contrive of interesting research projects that require expertise in fields in which we have no training.

¶ 19 Moreover, as Justice Lee points out, “[m]ost judges are generalists.” Indeed, we are aware of almost no one sitting on the bench or practicing law in this state who has the kind of scientific expertise required to reliably conduct the research Justice Lee requires. And issues of statutory construction where “both sides are able to marshal dictionary definitions in support of their view” permeate the majority of cases in this state. But in every such case, Justice Lee would require ad hoc linguistics research that could only be reliably conducted by dueling linguistics experts. Imposing such a significant financial burden on so many of the litigants coming through the doors of our courts would be tantamount to locking those doors for all but the most affluent. Moreover, it would place an unbearable burden upon our already thinly stretched district judges. That is simply unacceptable.

¶ 20 Justice Lee's appeal to linguistics research would not be so disquieting if he were not conducting the research himself, but merely citing findings that have been advanced by the parties, published in a scholarly journal, or authored by a respected source. Such sources include the reports of expert witnesses, published academic articles, and widely available dictionaries. In those cases, findings are subject to review by the relevant field of study or the opposing party's expert before we rely upon them. But if we conduct our own research, the parties are bound by our decision even if our methods or findings are subsequently found to be flawed. Accordingly, it is unfair, and indeed unwise, for us to decide a case on the basis of scientific research that is subject to neither prior review by the relevant field of study or adversarial briefing.

¶ 21 We are not experts in this field and accordingly have no intention of meeting Justice Lee on the merits of his research. But there are at least a few questions that are immediately apparent with regard to his methodology. To begin, of the eighty-six “hits” Justice Lee found with his search, he simply ignores thirty-six as having “insufficient detail to indicate whether the discharge at issue had reference to a single shot or the emptying of a magazine.” And thirty-one of the remaining hits required an interpretive assumption or were “a bit more ambiguous.” So many apparently neutral or questionable data points raise the question of whether the trend that Justice Lee draws from the data is statistically insignificant—“the figures extracted from the data are simply the result of random chance, and do not indicate what they seem to indicate.” It certainly calls into question his claim that “discharge of a weapon is used overwhelmingly in the single shot sense.” Additionally, the data set on which Justice Lee relies “ ‘contains more than 410 million words of text and is equally divided among spoken, fiction, popular magazines, newspapers, and academic texts.’ ” The point of a large data set is to avoid “ ‘basing conclusions on a few speakers' idiosyncrasies.’ ” But the “idiosyncrasies” of the Utah Legislature constitute the rule of law in this state. And the only way to identify those idiosyncrasies is through the text of the Utah Code, which is wholly absent from Justice Lee's data set. Apparently, there is no statutory or legal text of any kind in the data set. Accordingly, in an attempt to eschew the influence of any one speaker, Justice Lee's data set ignores the only speaker that matters. These are just two of the many questions raised by Justice Lee's methodology. And we have no doubt that more would surface if his methodology were subject to adversarial briefing or review by an expert in this field.

D. The Rule of Lenity Does Not Apply Because the Statute Is Not Ambiguous

¶ 22 Mr. Rasabout argues that we should use the rule of lenity to find in his favor. The rule of lenity requires that we interpret an ambiguous statute in favor of lenity toward the person charged with criminal wrongdoing. Thus, in the context of determining the allowable unit of prosecution under an ambiguous criminal statute, “doubt [must] be resolved against turning a single transaction into multiple offenses.” But as both parties agree, the rule of lenity is not implicated unless a statute is ambiguous. A statute is ambiguous when “its terms remain susceptible to two or more reasonable interpretations after we have conducted a plain language analysis.” In this case, we have conducted a plain language analysis and concluded that there is only one reasonable interpretation of the statute at issue. Accordingly, the rule of lenity is not implicated here. ¶ 23 Nevertheless, we take this opportunity to vacate the erroneous conclusion of the court of appeals with respect to the rule of lenity. The court of appeals held that the rule of lenity has been abrogated by the strict construction statute, Utah Code section 76–1–106. It reasoned that the statute may permissibly abrogate the rule of lenity because “the rule is one of statutory construction, not constitutional law.” The State defends this reasoning on appeal. Mr. Rasabout argues that the rule of lenity cannot be abrogated by statute because it is based in due process. We agree.

¶ 24 The United States Supreme Court has held that the rule of lenity “is rooted in fundamental principles of due process which mandate that no individual be forced to speculate, at peril of indictment, whether his conduct is prohibited.” The reasoning behind the Court's holding is simple. An ambiguous statute violates the notice requirement of due process because it “fails to give a person of ordinary intelligence fair notice that his contemplated conduct is forbidden by the statute.” The Due Process Clause of the Utah Constitution provides an alternative basis for the same protection. That provision “do[es] not permit enforcement of a statute that forbids an act in terms so vague that [persons] of common intelligence must necessarily guess at [the statute's] meaning and differ as to its application.” Thus, the rule of lenity is dictated by the notice protection afforded by both federal and state due process. Accordingly, insofar as the strict construction statute purports to abolish the rule of lenity, we hold that it does not.

E. Mr. Rasabout's Twelve Convictions Do Not Violate the Rule Against Multiplicity Under the Double Jeopardy Clause of the United States Constitution

¶ 25 Mr. Rasabout contends he committed only one offense and accordingly the State cannot punish him for twelve counts without violating the rule against multiplicity, which arises from the Double Jeopardy Clause of the United States Constitution. We disagree.

¶ 26 The Double Jeopardy Clause protects a defendant from “be[ing] twice put in jeopardy” “for the same offence.” From this protection springs the rule against multiplicity, which prohibits “ ‘multiple punishments for the same offense.’ ” The rule against multiplicity is violated if the State punishes a defendant for more counts of an offense than are allowed by the intended unit of prosecution for that offense. But it is not implicated in determining the allowable unit of prosecution for an offense or when the state charges a defendant with multiple violations consistent with the statutorily defined unit of prosecution.

¶ 27 In this case, Mr. Rasabout was convicted of twelve separate counts of unlawful discharge of a firearm on the basis of twelve discrete shots. Because the allowable unit of prosecution for that offense is each discrete shot, the State may charge and punish Mr. Rasabout for twelve counts without implicating Mr. Rasabout's Fifth Amendment rights.

II. THE SINGLE CRIMINAL EPISODE STATUTE AND THE SINGLE LARCENY RULE ARE INAPPLICABLE

¶ 28 In merging Mr. Rasabout's twelve convictions, the trial court relied on the single criminal episode statute, Utah Code section 76–1–401, and a case applying the single larceny rule. Mr. Rasabout urges us to adopt its reasoning. We conclude that neither the single criminal episode statute nor the single larceny rule apply here.

¶ 29 First, the single criminal episode statute, Utah Code section 76–1–401, is inapplicable because it does not dictate the merger of offenses. While this section provides that a “ ‘single criminal episode’ means all conduct which is closely related in time and is incident to an attempt or an accomplishment of a single criminal objective,” there is nothing in this or the operative sections that follow mandating that a court merge offenses committed within a single criminal episode. To the contrary, subsection (1) of section 76–1–402 specifically provides that “[a] defendant may be prosecuted in a single criminal action for all separate offenses arising out of a single criminal episode.” And subsection (2) explicitly accounts for a situation when “conduct may establish separate offenses under a single criminal episode.” Additionally, while the operative sections provide specific double-jeopardy-like protections, such as protecting a defendant from being “subject to separate trials for multiple offenses,” being “convicted of both the offense charged and [an] included offense,” and being “punished [for the same act] in different ways under different provisions of [the] code,” none of these concerns are implicated in this case. Mr. Rasabout was convicted in a single trial of twelve separate counts of violating a single offense. Thus, the trial court's reliance on section 76–1–401 for merging Mr. Rasabout's convictions was misplaced.

¶ 30 The single larceny rule is also inapplicable here because this case does not involve larceny. The single larceny rule governs the aggregation of multiple thefts over a period of time when the statutory text does not otherwise provide an allowable unit of prosecution. In the context of theft, a statutory descendent of common law larceny, we have held that “ ‘[i]f there is but one intention, one general impulse, and one plan, even though there is a series of transactions, there is but one offense.’ ” This doctrine was developed in the context of common law larceny and has been extended to larceny's statutory progeny in this state, including theft and robbery. But Mr. Rasabout has failed to provide us with a rationale for extending this common law doctrine outside the realm of larceny and into a statutory scheme that independently defines the allowable unit of prosecution.

¶ 31 In sum, we hold that the single criminal episode statute is inapplicable because Mr. Rasabout was convicted of twelve counts of a single statutory offense in a single trial. And we hold that the single larceny rule is inapplicable because this case is governed by the allowable unit of prosecution defined by the Legislature in the statute criminalizing unlawful discharge of a firearm.

III. CRUEL AND UNUSUAL PUNISHMENT IS NOT IMPLICATED IN DETERMINING THE ALLOWABLE UNIT OF PROSECUTION FOR AN OFFENSE

¶ 32 Mr. Rasabout finally argues that defining the allowable unit of prosecution as each discrete shot may result in a sentence that is cruel and unusual in violation of the United States Constitution and the Utah Constitution. The State argues that this issue is not implicated here. We agree with the State because this appeal raises only the allowable unit of prosecution, an issue that is distinct from punishment.

¶ 33 Both the United States Constitution and the Utah Constitution prohibit cruel and unusual punishment. But these protections are a constitutional backstop to the overall punishment imposed for criminal conduct; they do not dictate the allowable unit of prosecution. In this case, the question is how many convictions are permissible. On remand, the trial court will determine the appropriate punishment for Mr. Rasabout's criminal conduct. While the number of convictions will be one factor in that determination, the trial court will consider numerous other factors. It is at that point that Mr. Rasabout may challenge the total sentence as cruel and unusual.

¶ 34 In sum, we hold that the question of cruel and unusual punishment is not implicated in a determination of the allowable unit of prosecution for an offense.

CONCLUSION

¶ 35 Mr. Rasabout fired twelve shots at a house he knew to be occupied by a rival gang member. A jury convicted him of twelve counts of unlawful discharge of a firearm, in violation of Utah Code section 76–10–508. That was permissible because each shot fired was a separate discharge under the statute. We accordingly affirm the court of appeals and remand this case for further proceedings consistent with this opinion.

DURRANT, C.J., concurring in part and concurring in the judgment.

Chief Justice DURRANT, concurring in part and concurring in the judgment:

¶ 36 I concur in the majority's affirmance of the court of appeals' judgment and in all of the majority's analysis—except the portion addressing Justice Lee's corpus linguistics interpretive method. While I too would decline to use that method in this case, I applaud Justice Lee for his thoughtful exploration of corpus linguistics as a potential additional tool for our statutory interpretation tool box. I am open to the possibility that in certain cases it may well prove useful in our assessment of the ordinary meaning of statutory terms. But I do not consider it necessary in this case, because I find the majority's analysis of the meaning of the term “discharge,” an analysis conducted using our long-established methods of statutory construction, to be compelling. And I would further note that although the dictionary alternatively defines “discharge” as “to shoot” or “to unload,” the former definition applies specifically to firearms while the latter seems to encompass the more general concept of unloading cargo or emptying a container. Granted, the cartridge of a firearm is in some sense a “container” that can be “unloaded,” but in the context of a statute that involves firearms, I am confident the legislature intended the term “discharge” to mean “to shoot,” not “to unload.” So unlike Justice Lee, I do not believe this is a case where the dictionary “fails to dictate the meaning that the statutory terms must bear.”

¶ 37 Further, I am concerned about our use of corpus linguistics in a case where it has not been argued by the parties. Now, it is certainly true, that we may, and often have, employed dictionaries, canons of construction, or other tools for statutory interpretation that have not been argued by the parties. But primary linguistics research seems to me a step removed from these traditional tools. I fear that a sua sponte venture into such territory may be fraught with the potential for error. I think this is illustrated by Justice Lee's critique of the flaws in Judge Posner's sua sponte attempt at such research. The fact that a jurist of Judge Posner's intellect and stature is capable of such missteps suggests to me that, even in those cases where corpus linguistics may be a useful addition to our traditional methods of statutory construction, it would be best employed by us, or by other judges, only after the parties have raised it and argued it. This would allow the respective sides in the dispute to challenge each other's database, methodologies, and conclusions.¶ 38 Finally, in our exploration of whether or when corpus linguistics should play a role in our interpretation of statutes, it is important that we weigh the potential usefulness of the approach against its potential cost. In some cases, the linguistic analysis may be sufficiently complicated that an expert would be required. This would be an unfortunate cost to add to the already-too-expensive litigation process. But I think that in other cases the lawyers could themselves conduct the linguistic analysis, outline their methods and conclusions in their briefs, and present argument as to why we should adopt their approach over their opponents'. Because so few judges and, perhaps, even fewer members of the bar are familiar with corpus linguistics, I believe it is simply too soon to know whether the benefits of using this new tool warrant the increased expense.

¶ 39 Finally, I wish to make clear that while I would not employ Justice Lee's corpus linguistics analysis in this case for the reasons I have specified, and because of other concerns I have not discussed here nor resolved for myself, I look forward to our continued debate in this area. Should we elect to embark down this path, we should tread slowly and cautiously. And caution dictates that this potential method of statutory interpretation be fully tested in the crucible of the adversarial process, rather than simply applied sua sponte by our court.

LEE, A.C.J., concurring in part and concurring in the judgment.

Associate Chief Justice LEE, concurring in part and concurring in the judgment:

¶ 40 I concur in the judgment of the court and in elements of the majority's analysis. Thus, I share the majority's goal of “giv[ing] effect to the intent of the Legislature,” and of looking for such intent in “the text of the statute.” Supra ¶ 10. And I agree that our interpretation of the operative statutory text “turns on the meaning of the term ‘discharge’ in the context of a ‘dangerous weapon or firearm.’ ” Supra ¶ 11. Ultimately, moreover, I concur in the court's construction of this statutory phrase—as defining the element of prosecution in terms of “each discrete shot” expelled from a gun.Supra ¶ 1.

¶ 41 I write separately, however, because I cannot resolve the ambiguity in the term discharge as the court does—by mere resort to the dictionary. I see the need to look elsewhere. I would interpret the terms of the statute by looking for real-world examples of its key words in actual written language in its native context. This sort of analysis has a fancy name—corpus linguistics. But it is hardly unusual. We often resolve problems of ambiguity by thinking of examples of the use of a given word or phrase in a particular linguistic context. I propose to do that (as I have in a couple of prior opinions) on a systematic scale—by computer-aided searches of online databases in an effort to assemble a greater number of examples than I can summon by memory on my own.

¶ 42 I begin with a description of the question presented in this case as I understand it. Then I outline my approach to answering it. And I close with a response to criticisms to my methodology.

I

¶ 43 My starting point is in line with the majority's: I would seek to discern the intent of the legislature, and to find it in the language it enacted into law. When the text is plain that enterprise is straightforward. But in cases litigated to an appeal that is rarely the case. The statutory text in such cases is often ambiguous. One form of ambiguity concerns a conflict between the “ordinary” and “specialized” meaning of a term. Thus, a threshold question is whether the legislative text conveys some specialized meaning—as in the case of a statutorily defined term, a scientific phrase, or a legal term of art.¹ If so that is the end of the matter; the specialized meaning controls if it was intended by the legislature.²

¶ 44 Another form of ambiguity concerns a contest between two different definitions of a word in common usage. But “[t]he fact that the statutory language may be susceptible of multiple meanings” does not leave us without an answer, as all but one of the meanings is often “eliminated” by the structure or context of the statute. Olsen v. Eagle Mountain City, 2011 UT 10, ¶ 13, 248 P.3d 465. When neither of two competing meanings can be so eliminated, however, we opt for the meaning we deem as more “ordinary.” See Hi–Country Prop. Rights Group v. Emmer, 2013 UT 33, ¶¶ 18–19, 21–22, 304 P.3d 851.

¶ 45 This case presents this latter question. No one has suggested that discharge is a specialized term,³ or that the structure of the statute eliminates one of the parties' proposed definitions. So we are left to construe the relevant statutory language to convey the “ordinary meaning” of its terms—the meaning its words would have in the mind of a “reasonable person familiar with the usage and context of the language in question.” Olsen, 2011 UT 10, ¶ 9, 248 P.3d 465.⁴

¶ 46 Dictionaries are a good “starting point” for analyzing ordinary meaning. Hi–County Prop. Rights Grp. v. Emmer, 2013 UT 33, ¶ 19, 304 P.3d 851. They are “useful in cataloging a range of possible meanings that a statutory term may bear.” Id. (citing Henry M. Hart, Jr. & Albert M. Sacks, The Legal Process: Basic Problems in The Making and Application of Law 1375–76 (William N. Eskridge, Jr. & Phillip P. Frickey eds., 1994) [hereinafter Hart & Sacks ] ). Dictionaries provide “ ‘an historical record, not necessarily all-inclusive, of the meanings which words in fact have borne.’ ” Id. (quoting Hart & Sacks, at 1190). “Such a record, however, will often fail to dictate ‘what meaning a word must bear in a particular context.’ ” Id. (quoting Hart & Sacks, at 1190). “That question will often require further refinement—of selecting the best meaning among a range of options, based on other indicators of meaning....” Id.

¶ 47 This case requires such “further refinement.” It is “one of those cases where the dictionary fails to dictate the meaning that the statutory terms ‘must bear’ in this context.” State v. Canton, 2013 UT 44, ¶ 14, 308 P.3d 517. That is because, as in Canton, “dictionary definitions” of the operative terms of the statute “leave the statute semantically open to” both parties' positions. Id. On one hand, as the majority indicates, discharge is susceptible to the definition advanced here by the State—of discharge as “to release from confinement” or “to shoot.” Supra ¶ 12 (quoting Merriam–Webster's Collegiate Dictionary YY 356 (11th ed.2012)).⁵ Yet that is not the only sense of discharge identified in the dictionary. Discharge is also understood as “to empty of a cargo: UNLOAD.” SeeWebster's third Int'l Dict. 644 (3d ed.2002). And that is the sense of discharge advanced here by the defense. Rasabout urges a unit of prosecution defined by unloading or emptying the contents of a weapon, asserting that the latter sense of discharge supports his position.

¶ 48 The majority paints part of the relevant picture. It cites the first of the above definitions and adopts it as the operative sense of discharge of a firearm under Utah Code section 76–10–508(1)(a). But the majority fails to acknowledge the second dictionary definition set forth above. And I am unsure of its basis for preferring the first definition over the second.

¶ 49 In context, I see three possible grounds for the court's conclusion. But each falls short.

¶ 50 First, the definition invoked by the court may be listed first in the dictionary, seeMerriam–Webster's Collegiate Dictionary at 356 (listing, as the first sense of the verb discharge, “to relieve of a charge, load, or burden”), but there is no such thing as a “main” or “primary” dictionary definition.⁶ “The dictionaries most relied upon by courts in statutory interpretation make no claims about the ordinariness of the words they define or the senses they assign to those words.” J.M.W. v. T.I.Z. (In re Baby E.Z.), 2011 UT 38, ¶ 98, 266 P.3d 702 (Lee, J., concurring) (citing Stephen C. Mouritsen, The Dictionary Is Not a Fortress: Definitional Fallacies and the Corpus–Based Approach to Plain Meaning, 2010 BYU L. Rev. 1915, 1925–45 (discussing problems with dictionary usage by courts)). “Nor do they present their lexical information in a way that reveals ‘ordinary’ usage.” Id. “A number of dictionaries simply rank their definitions according to evidence of historical usage.” Id. (Emphasis added). The dictionary cited by the majority, Webster's, “expressly disavows any attempt to establish a hierarchy of ordinariness in the ranking of its senses, admitting that sometimes an ‘arbitrary’ listing of senses is used.” Id.; see alsoMerriam–Webster's Collegiate Dictionary at 20a. So the dictionary generally cannot provide the basis for preferring one of its definitions over another.¶ 51 Second, the majority seems to attribute significance to the apparent meaning of the component parts of the term discharge. In embracing the shoot sense of discharge, the court notes that the term “is made up of ... a prefix [‘dis'] and the root word [‘charge’].” Supra ¶ 12. And because “dis” conveys a “negative or opposite,” and “charge” indicates a “quantity of explosive,” the majority concludes that “the clearest reading of the statute is that discharging a weapon or firearm means shooting a weapon or firearm.” Supra ¶ 12. The ordinary meaning of a word is not always discernible by simple addition, however. Many words take on meanings that bear a strained relation to the sum of their constituent parts—including words with the prefix “dis,” like disease or disheveled. The former is more than a mere state of unease.⁷ And the latter is not the opposite of sheveled.

¶ 52 In this case, discharge does appear to convey the undoing of a charge. But that conclusion still begs the question of the nature of the charge being undone. Even if the relevant charge is, as the majority says, a “quantity of explosives,” we must decide here what quantity of explosives must be undone to count as a discharge. The majority says that it's a “single projectile with a single explosion.” Supra ¶ 13. That sounds fine. But Rasabout's alternative is no less defensible; all of the bullets in a gun's magazine also qualify as a “quantity of explosives.” So the majority's deconstruction of the component parts of discharge does not yield a conclusive answer to the interpretive question presented.

¶ 53 That leaves a third explanation for the majority's conclusion: The court's sense of discharge as shoot may simply be an expression of the majority's linguistic intuition. That is the way most problems of lexical ambiguity are resolved by our courts.⁸ Instead of acknowledging and rejecting contrary senses of a statutory term, judges tend to ignore them—identifying only the sense of a word they deem ordinary without acknowledging any others.

¶ 54 Reliance on judicial intuition is entirely appropriate. When confronted with problems of ambiguity in legal text, judges must employ their intuition to discern the proper sense of its terms. Too often, however, we proffer our intuition as if it were authoritative; and when we do so in a manner cloaking alternative senses of the operative terms, we deprive the parties and the public of a clear understanding of the complex nature of the linguistic problems that we confront. We also risk confirmation bias, or just old-fashioned mistake. The former risk is well- known.⁹ And it is magnified in cases where our analytical methods are opaque. Our human intuition of ordinary meaning, moreover, is fallible. As linguists have noted, our analysis of word meaning and usage “remains hidden from introspection,” and may thus too often be mistaken.¹⁰

¶ 55 For these reasons we should be more transparent about our means of resolving the difficult problems of interpretation that come before us. When confronted with a contest between two competing constructions that each find tenable support in our lexicon, we should openly acknowledge the ambiguity. And when we resolve that ambiguity based purely on instinct or intuition, we should be open about that.

¶ 56 I would do so here. Instead of embracing one accepted sense of discharge of a firearm and ignoring the contrary understanding, I would acknowledge both. To the extent my intuition tells me (as it does) that the single shot sense of discharge is the more ordinary sense of discharging a firearm, moreover, I would openly present that as the basis for our decision. I would also take the matter a step further, however. I would also check my intuition against publicly available means for assessing the ordinary meaning of a statutory phrase. My proposed means for doing so are outlined below.

II

¶ 57 In this age of information, we have ready access to means for testing our resolution of linguistic ambiguity. Instead of just relying on the limited capacities of the dictionary or our memory, we can access large bodies of real-world language to see how particular words or phrases are actually used in written or spoken English. Linguists have a name for this kind of analysis; it is known as corpus linguistics.

¶ 58 The fancy Latin name makes this enterprise seem esoteric and daunting. It is not. We all engage in it even if we don't attach the technical label to it. A corpus is a body, and corpus linguistics analysis is no more than a study of language employing a body of language.¹¹ When we communicate using words we naturally access a large corpus—the body of language we have been exposed to during our lifetimes—to decode the groups of letters or sounds we encounter. The most basic corpus linguistics analysis involves our split-second effort to access the body of language in our heads in our ongoing attempt to decode words or phrases we may be uncertain of. We all do that repeatedly every day.

¶ 59 The first time we heard a skateboarder described as “so sick,” we may have misperceived that description as negative—if not an indication of illness, then at least a general notion of repellence. Oxford English Dictionary online, “sick” (definitions 1a and 4a(d)). In time we learned to decode the above use of this term in this context. We came to understand that “so sick” in this context was not criticism but high praise. Oxford English Dictionary online, “sick” (“slang (now esp. Skateboarding and Surfing ). Excellent, impressive; risky”). Our means of sensing that meaning, moreover, was not a matter of looking the word up in the dictionary (as the word sick itself can mean all of the above, and this new meaning does not yet appear in most dictionaries). It was a matter of accessing the body of language we have encountered to view the meaning of the word in its broader context—the object of the adjective (skateboarder), the broader phrase (“so sick”), and even the demographics of the speaker (millennial).

¶ 60 Judges have long employed similar methods of decoding the language of the law. To resolve ambiguities in statutes, judges access their memory of the use of uncertain terms in the context in which they have heard them used. In so doing they are engaged in corpus linguistics analysis. And no one bats an eye, because this is a natural, accepted method for humans to resolve ambiguities in language.

¶ 61 It is a small step to utilize a tool to aid our linguistic memory. Judges do this with some frequency as well. Naturally. If judges are entitled to consult the corpus of language in our heads (and how could we not?), we must also be permitted to supplement and check our memory against publicly available sources of language.

¶ 62 A dictionary, in fact, is a corpus. Its material—definitions of an extensive body of words—is compiled from broader linguistic corpora. And some dictionaries include snippets of concordance lines from those corpora, as examples of instances in which a particular sense of a word might be intended. Judges also use dictionaries as a means of this kind of corpus analysis.

¶ 63 Judges have also looked to databases available in Westlaw or Lexis, or more broadly through an internet search engine, to help us recall how particular words or phrases are commonly used in written or spoken English. We have done so in opinions for this court. The United States Supreme Court has also taken this approach. And judges have long analyzed the meaning of legal words and phrases by seeing how other courts have used them. That too is corpus analysis.

¶ 64 I would employ this sort of tool in resolving this case. In part II.A. below, I present the results of a Google News search. The results of that search demonstrate that the verb discharge as used in conjunction with a firearm is almost always used in the sense of a single shot (and not the emptying of all of the bullets available in the magazine).

¶ 65 I would also take this analysis a step further, however. For reasons described below, there are reasons to look beyond Google News to consult an additional tool for understanding the notion of discharge of a weapon as that phrase is commonly used in written or spoken English. In part II.B., I present a parallel analysis employing an online tool for analyzing a large online collection of written and spoken English—the Corpus of Contemporary American English (COCA), available at corpus.byu.edu/coca. The search results from this analysis confirm this same understanding of discharge of a weapon, but with greater transparency and reliability.

A

¶ 66 A natural place to go for access to a large body of linguistic data is the internet. Most all of us (even judges) are familiar with this source of data, and with the search tools for accessing it. And thoughtful analysis using the results of a common search engine can generate data we may use as an empirical check on our (imperfect) linguistic intuition.

¶ 67 Judges on this court and elsewhere have begun to access this resource as we tackle problems of lexical ambiguity. Over the past few years, members of this court have authored several separate opinions presenting internet search results in support of our construction of ambiguous statutory text. More recently, in State v. Canton, 2013 UT 44, 308 P.3d 517, we employed such an approach in an opinion of the court. Our unanimous Canton opinion resolved an ambiguity in a tolling provision for our criminal statutes of limitations, Utah Code section 76–1–304(1). The critical statutory phrase tolls the statute of limitations during any period in which the defendant is “out of the state.” And we resolved a critical ambiguity in that phrase—as to whether it referred to physical absence beyond a state's boundaries, or mere unavailability for legal process—by use of a Google News search.

¶ 68 In Canton we concluded that “[d]ictionaries [we]re ... insufficient by themselves to resolve the interpretive task before us.” Id. ¶ 20. And we therefore indicated a need to “look elsewhere to determine the ordinary meaning of the language of the tolling statute.” Id. The “elsewhere” was a Google News search, which we employed to assess the “way the full phrase [‘out of the state’] is typically used in common parlance.” Id. ¶ 26. The results of that search confirmed that “[w]hen the phrase ‘out of the state’ is used in its full context, it refers to the physical territory of a state, not its political power or influence.” Id. ¶ 27.

¶ 69 We described the search and its results as follows:

This conclusion is based on results of a Google News search, http://news.google.com, considering 150 instances in which the phrase “out of the state” was used in news stories published in May 2013—27 of which involved references to the relationship between a person and the state. Not one of those 27 relevant references use “out of the state” in a manner involving absence of a person from the legal authority or influence of a state. Every single one of them makes unequivocal reference to the physical confines of a state.

Id. ¶ 27 n. 6. On the basis of these search results, we concluded that “although Canton's construction” of the Utah tolling provision was “semantically plausible based on dictionary definitions of ‘out of’ and ‘the state,’ it cannot be reconciled with the uniform understanding of the extended statutory phrase ‘out of the state.’ ” Id. ¶ 27. Because “[t]hat phrase is not used in the way that Canton construes it,” we rejected it as incompatible with the ordinary meaning of the statutory text. Id.

¶ 70 I would follow the Canton model in this case. Because both sides are able to marshal dictionary definitions in support of their view of discharge, we must reach beyond the dictionary to resolve this case. And, as in Canton, a Google News search confirms the conclusion that the majority adopts but cannot justify on the basis of the dictionary, or etymology, or mere intuition.

¶ 71 Google is a widely used, well-known search engine. As most people know, this search engine can do more than search the world-wide web. It may also be employed to search an extensive body of published newspaper articles.

¶ 72 Such a corpus is a reliable resource for assessing the ordinary meaning of a statutory term. Published newspaper articles are a useful body of natural language due to their size. The Google News corpus, moreover, is both well-understood by the public (including judges and lawyers) and easily adapted for analysis (in that a search may be time-delimited and the search results may be reviewed in a systematic, transparent way). Unlike a dictionary, moreover, a Google search allows for analysis of phrases—or key words in a particular context—and not just individual words.

¶ 73 A Google News search confirms the single shot sense of discharge when that term is used in connection with a firearm. On a search conducted on June 4, 2015, the phrase “discharge a firearm” generated forty-three hits for newspaper articles published between March 1 and June 4, 2015. Fifteen of those hits were inconclusive in terms of the intended sense of “discharge.” I omitted those hits from my analysis. Of the remaining twenty-eight articles, none of them clearly support Rasabout's notion of discharge involving emptying all bullets in a gun's magazine. And twenty-seven clearly employ the single shot sense of the term.¶ 74 The twenty-seven relevant hits linked to articles cited below. Each of them conclusively indicates that when we speak of discharging a firearm, we are talking about a single shot, and not a complete emptying of a gun's magazine. In nine of the articles it is clear from the context that only one shot was fired. In two, each discharge is described as a separate event. In ten others, the term discharge is used in close connection with the terms shoot, fire or similar words describing a single firing of a gun. And in the remaining six, the use of discharge simply doesn't make sense grammatically if read as a reference to multiple shots.

¶ 75 I would employ the above corpus analysis to sustain the conclusion that each single shot is a separate discharge of a weapon under Utah Code section 76–10–508(1)(a). That is the conclusion suggested by my linguistic intuition. But, more transparently, it is also confirmed by an analysis of the use of natural language in current newspaper articles.

B

¶ 76 Most any analysis of public sources of real-world language, in my view, is better than a judge's take-my-word-for-it assertion of ordinariness. But not all searches are created equal. A Google News search bears some of the key hallmarks of reliable analysis. It is certainly more transparent and easier to replicate than a judge's intuition.

¶ 77 Yet a Google News search is hardly unimpeachable. The Google algorithm is proprietary and thus not fully transparent. So we cannot tell exactly what factors affect the results of any given search on Google News. Another concern goes to the replicability of a given search. My search terms and results are memorialized in the above footnotes. But because the Google algorithm is hidden, and the results of any given search may be affected by factors unknown to (or particularized for) an individual user, there is no guarantee that the same search performed at another time on another computer will generate identical results.

¶ 78 These problems will be heightened—and others added to them—in a broader Google search of the world-wide web (as opposed to the Google News database). A search for relative hit counts is especially problematic. Consider the Seventh Circuit's analysis in United States v. Costello, 666 F.3d 1040, 1044 (7th Cir.2012). In Costello the court was faced with the question of the scope of the crime of “habor[ing]” an illegal alien under 8 U.S.C. § 1324(a)(1)(A)(iii) (2012). Judge Posner, writing for the court, conducted a Google search to try to resolve an ambiguity—of whether the statute used harbor in the sense of “to give shelter or refuge to” (seeWebster's Third New International Dictionary, sense 1a(1)) or “to receive clandestinely and conceal” (id., sense 1a(2)). To Judge Posner's credit, the Costello opinion acknowledged the limits of dictionaries. It noted that “[d]ictionary definitions are ... acontextual, whereas the meaning of sentences depends critically on context, including all sorts of background understandings.” 666 F.3d at 1042. And the court's instincts were generally in line with the tools and methods that I propose. Instead of purporting to resolve the lexical ambiguity by resort to a dictionary, Judge Posner sought to do so by means of a corpus analysis of the world-wide-web—via a Google search.

¶ 79 Yet Judge Posner's search highlights some deficiencies in standard Google internet searches. His search, as described in the Costello opinion, was as follows:

A Google Search ... of several terms in which the word “harboring” appears—a search based on the supposition that the number of hits per term is a rough index of the frequency of its use—reveals the following:

“harboring fugitives”: 50,800 hits

“harboring enemies”: 4,730 hits

“harboring refugees”: 4,820 hits

“harboring victims”: 114 hits

“harboring flood victims”: 0 hits

“harboring victims of disasters”: 0 hits

“harboring victims of persecution”: 0 hits

“harboring guests”: 184 hits

“harboring friends”: 256 hits (but some involve harboring Quakers—“Friends,” viewed in colonial New England as dangerous heretics)

“harboring Quakers”: 3,870 hits

“harboring Jews”: 19,100 hits

¶ 80 From these results, the Costello opinion concludes that “[i]t is apparent ... that ‘harboring,’ as the word is actually used, has a connotation—which ‘sheltering,’ and a fortiori ‘giving a person a place to stay’—does not, of deliberately safeguarding members of a specified group from the authorities, whether through concealment, movement to a safe location, or physical protection.” And on that basis the court reversed the conviction under review, concluding that there was no evidence to support the conclusion that Costello had “harbored” her boyfriend, an illegal alien, in the sense of seeking to conceal him from the authorities.

¶ 81 The Costello opinion is a step in the right direction. But it also highlights some deficiencies of Google web searches as a basis for assessing ordinary meaning. The Costello analysis is vulnerable for reasons noted above—that hit counts are unreliable because the Google algorithm is unknown, as underscored by the fact that different searches at different times on different computers may reveal very different results.

¶ 82 But the Costello approach is also vulnerable to other attacks. One is that Judge Posner's choice of search terms seems arbitrary. To assess the ordinary meaning of harbor, he searches only the present participle form of the verb (harboring ). And he chooses objects of the verb (fugitives, enemies, refugees, flood victims, victims of disasters, victims of persecution, guests, friends, Quakers, and Jews ) on grounds that are not stated. (Curiously, the statutory object—alien —is not included.)

¶ 83 These search terms are understandable. But they are somewhat arbitrary (presumably based on Posner's linguistic intuition), and could easily skew the results. The use of intuition is, again, understandable. But if the goal of this type of analysis is to check the judge's intuition, this move undermines a key benefit of the approach.

¶ 84 These problems can be addressed by means of a more transparent, reliable search tool. The tool that I would employ is one developed by a renowned corpus linguist, Professor Mark Davies. This corpus is known as the Corpus of Contemporary American Usage (COCA). See Mark Davies, The Corpus of Contemporary American English: 410+ million words, 1990–present, http://corpus.byu.edu/coca (2008). COCA is “the largest freely-available corpus of English, and the only large and balanced corpus of American English.... The corpus contains more than 410 million words of text and is equally divided among spoken, fiction, popular magazines, newspapers, and academic texts.” Id.

¶ 85 Like Google or Westlaw, the COCA search engine is easy to use. But unlike Google, and to a lesser extent Westlaw, COCA is also completely transparent, and it generates search results that are easily replicable. COCA, moreover, avoids the shortcomings of a Google web search as noted above, and illustrated in the Costello opinion. With the COCA search engine, there is no need for a user to think up her own objects of the verb harbor. COCA allows the user to generate a list of the most common words used near harbor. Significantly, moreover, the user can search only for the verb forms of harbor. So the COCA user can generate the most common nouns near harbor as a verb, instead of guessing at what they might be. That list of collocates (essentially, “word neighbors”) in itself will give some insight into the ordinary meaning of harbor.

¶ 86 COCA also allows the user to do more than get a simple “hit count.” A COCA search yields a display of each use of harbor, exactly like Google or Westlaw, in the context of each of its most common word neighbors. By examining each instance of harbor with its common word neighbors, we can assess the frequency of each of the attested meanings of the verb.

¶ 87 COCA also facilitates transparency and replicability. A search performed on the COCA site can be saved and linked for future reference, allowing a party, counsel, or interested commentator to review the search results and assess the court's analysis.

¶ 88 I would accordingly utilize a COCA search to analyze the meaning of discharge of a firearm. My search identified eighty- six instances of the verb discharge within five words of the nouns firearm, firearms, gun, and weapon. By examining the instances of discharge in connection with these nearby nouns, I confirmed that the single shot sense of this verb is overwhelmingly the ordinary sense of the term in this context.

¶ 89 Twelve of the eighty-one hits in the COCA search clearly linked discharge to a single bullet. In sixteen other hits, the discharge was accidental. I deemed those hits as also consistent with the single shot sense of discharge, as it seems highly unlikely if not impossible that an accidental trigger-pull could result in a release of all of the bullets in a gun's magazine. Fifteen other hits were a bit more ambiguous; but on closer examination, the discharge in question seemed to imply a single shot (based on the nature of the weapon, the circumstance of the discharge, or description of the resulting damage).

¶ 90 In thirty-six other instances I concluded that there was insufficient detail to indicate whether the discharge at issue had reference to a single shot or to the emptying of a magazine. One additional hit was rejected as irrelevant (involving a patient being discharged from a hospital, which happened to occur within five words of the word gun ).

¶ 91 In all, I found only one instance of discharge of a weapon that seemed consistent with the firing of multiple shots. Based on the context of this usage, it seemed likely that the discharge in question referred to a stream or burst of bullets instead of a single shot.

¶ 92 This COCA search accordingly confirms our linguistic intuition and is consistent with the Google News analysis above. It indicates that discharge of a weapon is used overwhelmingly in the single shot sense. Of eighty-one hits, (forty-four that were conclusive and relevant) only one seems consistent with Rasabout's notion of a burst of bullets (and even then, there is no basis for concluding that the burst involved emptying the entire contents of a gun's magazine). Thus, almost every conclusive instance of discharge of a weapon involves a single shot.

¶ 93 This provides strong confirmation of the basis of our holding in this case. And it does so on the basis of a transparent database that is publicly available, created by linguists, and subject to replication by anyone seeking to confirm (or reject) my analysis.

III

¶ 94 Novel tools for tackling old problems naturally prompt skepticism. That reaction is all the more predictable when the new tool implicates unfamiliar technology.

¶ 95 For decades lawyers used paper digests and other hard-copy compilations to find judicial opinions to support their arguments. When computer-searchable databases of opinions were first introduced, some were dubious. For years some even predicted that computer-based research would undermine the lawyer's craft. Not many of us think that way now. The addition of the tool of computer-based research is widely viewed as an enhancement, particularly when used in combination with other, traditional methods.

¶ 96 The point is not that corpus linguistic analysis is the next Westlaw or Lexis. Its utility is obviously more narrow. But the analogy is apt in that the tool I'm proposing would not replace but add to existing methods, and it would do so in a manner that takes advantage of technology available to us in the computer age.

¶ 97 That said, I accept many of the points raised by Chief Justice Durrant in his concurrence. I too see a need to proceed with caution, supra ¶ 39, and to “weigh the potential usefulness” of corpus analysis “against its potential cost.” Supra ¶ 38. And I certainly agree that this analysis “would be best employed by us, or by other judges, ... after the parties have raised it and argued it.” Supra ¶ 37. That is true for anything we do. We depend on the adversary process. Our opinions are better when adversary briefing is complete and in-depth.

¶ 98 Yet I do not see these as barriers for employing corpus analysis in this case. The parties have squarely presented the issue of the meaning of discharge of a weapon for our resolution, and they have given their best shot at offering linguistic analysis (using dictionaries and the lawyers' own linguistic intuition) on that issue. To some extent all members of the court have gone beyond the parties' briefing in deciding that issue. It is no affront to the adversary system for us to do that, as it is not just inevitable but entirely appropriate. (No one thinks that a careful judge should suspend independent analysis.) And it adds mostly transparency to extend our discussion to encompass the corpus analysis set forth in my opinion.

¶ 99 I concede that the Google News and COCA tools that I have employed will be unfamiliar to many. But it is not the case that my “rationale” is “different in kind from any argument made by the parties,” or even from analysis presented by my colleagues. Supra ¶ 17. Every member of this court is addressing the same issue, and we all are engaged in the same basic analysis—of seeking the ordinary sense of discharge when that term is used in connection with a firearm. We can do that using our intuition and the corpus-based information compiled in a dictionary (as my colleagues do), or we can extend the analysis (as I have) to look at examples of real-world language compiled by a Google News or COCA search. Neither approach is “subject to scientific review.” Supra ¶ 16. And neither yields to Mr. Rasabout “a reasonable opportunity to present a different perspective,” supra ¶ 17, if by that we mean the chance to engage with us at the opinion-writing stage to respond to our arguments.

¶ 100 Parties never have that opportunity, however. And they have no more opportunity to do that with regard to the majority's approach than with mine. The majority's insistence that discharge means shoot in this context is based on the majority judges' linguistic intuition—informed, no doubt, by the body of language that they have encountered during their lifetime. Certainly that analysis is not “subject to scientific review,” much less to pushback from Mr. Rasabout.

¶ 101 The majority's remaining concerns are a bit more substantial, or at least deserving of a more fulsome response. In the paragraphs below, I respond to the court's charges (a) that corpus analysis is a “scientific field of study” that is the domain of an expert witness and not a generalist judge, supra ¶¶ 18–19; (b) that judges lack the expertise necessary to conduct reliable corpus analysis, supra ¶¶ 19–21; (c) that the introduction of corpus analysis will “place an unbearable burden” on the judicial system by requiring “dueling linguistics experts” “in every case,” supra ¶ 19; (d) that my approach ignores corpus analysis of “the text of the Utah Code,” supra ¶ 21, and suggests that “ordinary meaning is the exclusive tool available to us in our effort to effectuate legislative intent,” supra ¶ 10 n. 16; and (e) that my corpus analysis is problematic in that it discards a certain number of hits “as having ‘insufficient detail’ ” to be conclusive, supra ¶ 21.

A

¶ 102 The legal or ethical propriety of “sua sponte” corpus analysis by a judge is a valid question. See supra ¶ 17. On a range of scientific matters, we think of research or analysis as the domain of experts. And we rightly eschew a judge's independent attempts at his own scientific “findings.” Supra ¶ 20. Such analysis, in fact, would be a breach of judicial ethics.

¶ 103 In Utah, our Code of Judicial Conduct provides that “[a] judge shall not investigate facts in a matter independently, and shall consider only the evidence presented and any facts that may properly be judicially noticed.” Utah R.J. Admin.Code Canon 2, Rule 2.9(C). So the majority would be right to reject my corpus analysis if it were a matter of independent factual investigation. That is usually what is involved in the “scientific field[s] of study” cited by the court. Supra ¶ 18. And on such questions, it would undoubtedly be improper for us to “conduct scientific research” or to “contrive of interesting research projects that require expertise in fields in which we have no training.” Supra ¶ 18.

¶ 104 But that is not at all what I am engaged in. Independent investigation is foreclosed only as to “facts,” not law. That is significant. Judicial analysis of the meaning of language, using corpus analysis or anything else, is aimed at interpreting the law. That is the judge's role. In performing the core function of deciding what the law is or should be, we cannot properly be restricted from consulting sources that inform our understanding.

¶ 105 The operative rule makes that abundantly clear. It opens the door to the consideration of “facts that may properly be judicially noticed.” Utah R.J. Admin.Code Canon 2, Rule 2.9(C). To the extent the analysis presented above encompasses “facts” (e.g., as to the way the verb discharge is typically used in our language in connection with the noun firearm or its synonyms), moreover, our law makes clear that we are free to consider them as a matter of judicial notice. Otherwise my colleagues' opinions also cross the line, as they also consider evidence not presented in the briefs regarding the ordinary meaning of discharge. A judge's judicial notice power is addressed by rule 201 of our rules of evidence. That rule indicates that the limitations in the rule are addressed to “an adjudicative fact only, not a legislative fact.” Utah R. Evid. 201(a). And the advisory committee note explains the distinction. It says that “[s]ince legislative facts are matters that go to the policy of a rule of law as distinct from the true facts that are used in the adjudication of a controversy they are not appropriate for a rule of evidence and best left to the law-making considerations by appellate and trial courts.” Utah R. Evid. 201 advisory committee's note. Thus, this type of analysis is an appropriate method of judicial analysis because it goes to “the policy of a rule of law as distinct from the true facts that are used in the adjudication of a controversy.” Id. ¶ 106 On reflection, moreover, this must be correct. A contrary conclusion would call into question a wide range of opinions of this court and many others. If we were foreclosed from considering outside material that informs our resolution of open questions of law, we would be barred from engaging in historical analysis relevant to a question of original meaning of a provision of the constitution, or from considering social science literature in resolving a difficult question under the common law. Linguistic analysis is no different; to the extent we charge our judges with resolving ambiguities in language, we cannot (and do not ) reasonably restrict their ability to do so on a well-informed basis—even on grounds not presented by the parties, and not within the domain of judges' professional training. Such a restriction, after all, would not just foreclose corpus analysis; it would also prevent us from consulting a dictionary or our own experience with language.

B

¶ 107 It is likewise fair to question our ability to perform corpus analysis properly. Most judges are generalists. And few if any have specialized training in corpus linguistics. So I concede the point that judges will not bring to bear the kind of training possessed by “linguistics experts” retained by the parties. Supra ¶ 19.

¶ 108 But that, respectfully, is not the point. We judges are experts on one thing—interpreting the law. And the fact that that enterprise may implicate disciplines or fields of study on which we lack expertise is no reason to raise the white flag. It is reason to summon all our faculties as best we can, and to overcome any weaknesses we may possess. This is not a matter of dreaming up “interesting research projects.” Supra ¶ 18. It is a matter of doing our job—of doing all we can to understand and implement the will of the legislature as expressed in the terms of its statutes, and to convey our grounds for doing so in a written opinion.

¶ 109 That job isn't always easy. It involves not just linguistic analysis but also historical inquiry—e.g., in finding original meaning. Few of us have training in historical research. It may even be said that lawyers and judges “are for the most part extremely bad historians,” and may “make up an imaginary history and use curiously unhistorical methods.” Yet judges of all stripes engage in historical analysis, particularly in their interpretation of the constitution. So the response to our lack of historical training is not to back away from the enterprise; it is to arm ourselves with the tools necessary to do the best history we can.

¶ 110 We face a parallel problem when it comes to our analysis of the meaning of language. When it comes to training or experience in methods of linguistic analysis, most of us lack specialized training. So there is certainly a risk, to paraphrase Max Radin, of judges using curiously unscientific linguistic methods.

¶ 111 But the proper response to this risk is not the abandonment of the enterprise of linguistic analysis. That enterprise is an integral element of judging. Judges cannot do their job without assessing the ordinary meaning of words. So the question is not whether to engage in linguistic analysis; it is whether to do so with the aid of—instead of in open ignorance of or rebellion to—modern tools developed to facilitate that analysis.

¶ 112 We could continue to judge the ordinary meaning of words based on intuition, aided by the dictionary. But those tools are problematic, for reasons explained above. And the impact of a judge's mere gut intuition is entirely opaque. So it is our current methodology and tools that involve bad linguistics produced by unscientific methods. If the concern is reliability, the proper response is to embrace—and not abandon—corpus-based analysis.¶ 113 To do so well, we judges must seek to understand this field better. We are not experts. At least I am not; I do not possess a complete understanding of the methodologies at our disposal. But I am convinced that the approach employed above is essential to a more reliable, transparent fulfillment of this judicial task.

¶ 114 Corpus analysis, in all events, is not rocket science. At some level we all do it intuitively in our minds. It's a small leap to check our intuition against examples of real-world language revealed by a Google- or COCA-based search of a body of written English. We don't need much expertise to do that well.

¶ 115 Corpus analysis is like math—at one level it's something basic that everyone does; at another level it's something complicated that only “experts” can do. The type of corpus analysis I am doing—and advocating—is the former. I just think we should be using a calculator instead of doing it in our heads.

¶ 116 As to historical analysis, Justice Antonin Scalia and his co-author Bryan Garner have aptly repudiated the charge that “ ‘no one can reconstruct original understanding precisely’ ” with a powerful reminder of the judicial task: “Our charge is to try.” The same can be said of linguistic analysis. It may not be possible to resolve questions of ordinary meaning with absolute certainty. But we must try. And in so doing we must bring to bear the methods and tools developed in the 21st Century to better understand the meaning of language for this crucial element of judging.

C

¶ 117 I see no reason to expect that the introduction of corpus analysis in judging ordinary meaning will require “dueling linguistics experts” “in every ... case” where both sides present plausible dictionary definitions to support their position. Supra ¶ 19. First, as noted, this isn't rocket science. Just as lawyers in constitutional litigation have learned to present historical analysis of original meaning, counsel in statutory cases can compile Google-or COCA-based examples of the use of statutory phrases in real-world written English. To do this well, lawyers will have to inform themselves a bit in basic linguistic methodology. But lawyers are crafty and ingenious. They learn new methods and tools all the time. I see no barrier to their presentation of helpful analysis in this field.

¶ 118 Perhaps there will be an occasional case where linguistics expertise could be useful, and where a party may wish to retain an expert. But even if an expert were retained in every case in which corpus analysis were determinative, we would hardly see experts in “every ... case” in which the parties proffer dueling dictionary definitions. Again the majority misunderstands my approach. Corpus analysis is something of a last resort. It comes into play only if we find that the legislature is not using words in some specialized sense, and only if we cannot reject one of the parties' definitions based on the structure or context of the statute. See supra ¶¶ 43–44. Corpus analysis comes in, in other words, as something of a tie-breaker where we find no better way of resolving the matter. In my five years on this court, I have employed such analysis only a very few times. In the many other statutory cases that come before us, I have disposed of the matter using more traditional tools of interpretation.

¶ 119 The “unbearable burden” imagined by the majority, supra ¶ 19, is a strawman. Lawyers can easily learn to present corpus analysis in the vast run of cases that implicate this approach, and the need for experts will at most be very occasional.

D

¶ 120 The majority's next criticisms of my corpus analysis—that it ignores consideration of “the text of the Utah Code,” supra ¶ 21, and suggests that “ordinary meaning is the exclusive tool available to us in our effort to effectuate legislative intent,”supra ¶ 10 n. 16—are based on misconceptions. First, a corpus analysis of the “text of the Utah Code” would make sense if we were thinking of discharge of a firearm as a legal term of art. In that case I would agree wholeheartedly with the notion of searching for uses of the statutory language to “identify ... idiosyncrasies” in the way this term is used in the “rule of law in this state.” Supra ¶ 21. But no one (certainly not the majority) has proffered the view that discharge is a legal term of art subject to specialized meaning in the law. Everyone agrees that this term is being used in its ordinary sense. So the critique for the lack of this kind of analysis is puzzling.

¶ 121 Second, I am not at all seeking to limit the enterprise of statutory interpretation to a search for “ordinary meaning” in all cases. None of my (unanimous) opinions string-cited dismissively by the majority “suggest that ordinary meaning is the exclusive tool available to us in our effort to effectuate legislative intent.” Supra ¶ 10 n. 16 (citing, in reference to this criticism, Barneck v. Utah Dep't of Transp., 2015 UT 50, ¶ 28, 353 P.3d 140 ; State v. Bagnes, 2014 UT 4, ¶ 13, 322 P.3d 719 ; Hi–Country Prop. Rights Grp. v. Emmer, 2013 UT 33, ¶ 18, 304 P.3d 851 ; State v. Canton, 2013 UT 44, ¶¶ 12–13, 308 P.3d 517 ; Olsen v. Eagle Mountain City, 2011 UT 10, ¶ 9, 248 P.3d 465 ). All of these opinions are in line with the approach I have outlined here—of asking first about specialized meaning, looking next for structural or contextual grounds for eliminating one of the parties' definitions, and resorting to ordinary meaning only at the end of the interpretive road. None proposed a search for ordinary meaning as “the only consideration” in statutory interpretation. See also supra ¶ 43 n. 1.

¶ 122 I have no quarrel, of course, with the goal of “giv[ing] effect to the intent of the Legislature.” Supra ¶ 10. That is the aim of any thoughtful jurist on a matter of statutory interpretation. And it is the enterprise that we are all engaged in here.

¶ 123 Sometimes we speak of a difference in the theory or methodology for getting there. There is a purported division—sometimes acknowledged in judicial opinions, and more often in law journals—between judges who approach matters of interpretation from the standpoint of “purposivism” and those who see their enterprise as a matter of “textualism.” But these “label [s] tend[ ] toward caricature”; “textualism and purposivism ... share more conceptual common ground than” is “sometimes emphasized.” “[N]o ‘textualist’ favors isolating statutory language from its surrounding context, and no critic of textualism believes that statutory text is unimportant.” So the overlap is significant. Both approaches look for legislative intent, both seek to find it in the statutory text, and both acknowledge the relevance of context in interpreting the text.

¶ 124 In my opinions on statutory interpretation, I tend to speak of our role in interpreting the text of the statute, as that is what was voted on and signed into law by the legislature and the governor. This is a basic premise of textualism. But my approach to interpretation is founded on the notion of legislative supremacy and is aimed at finding legislative intent. And it acknowledges the need to consider context—and even a carefully informed sense of legislative purpose—in resolving ambiguities in statutory language. To a large extent this is just a flipside of modern purposivism.

¶ 125 I have long understood my colleagues on the court to share this essential view of our role. We may occasionally use different labels to describe our aims, but it has always seemed to me that we are engaged in the same essential function. Thus, I have not viewed our court's opinions as embracing “old-school” purposivism—of a prerogative of “adjust [ing] the enacted text to capture what [the legislature] would have intended” if it had thought about the issue before the court more carefully. My colleagues' purposivism, as I understand it, is one that looks to statutory language to find legislative intent, and is bound by the meaning of those words as we interpret them.

¶ 126 I am comfortable that this is a close cousin to the textualist approach that I have taken. So I have joined opinions authored by my colleagues even though they do not speak in explicitly textualist terms. And my colleagues have joined my opinions even though they are framed in textualist terms (and may even include corpus analysis ).

¶ 127 My textualism has never been about finding “ordinary meaning” as the “exclusive tool available to us in our effort to effectuate legislative intent.” Supra ¶ 10 n. 16. The court's critique along those lines is simply mistaken. In this case I'm asking the exact same question the court is asking; I'm just resorting to an additional tool for answering it.

¶ 128 The tool I employ (COCA) and the methodology I use (corpus analysis), moreover, are not just for textualists. They are for anyone who is interested in analyzing text. All judges—textualists and purposivists and others—focus on the meaning of statutory language. The question is how to do so—whether to stick with using our intuition and dictionaries only, or whether to check our intuition by searching a database of language. I see more upside than downside to taking that step.

E

¶ 129 I welcome focused critiques of the corpus analyses I have presented as grounds for my decision. A main point of my approach, after all, is to facilitate transparency and discourse. But the majority's criticism of my exclusion of thirty-six “inconclusive” hits from my COCA analysis is puzzling, as is its challenge to the “statisitical[ ] significan[ce]” of my conclusions. Supra ¶ 21. The decision to exclude thirty-six of the hits in the COCA analysis is straightforward. I didn't “ignore” them. Supra ¶ 21. I found “insufficient detail to indicate whether the discharge at issue had reference to a single shot or to the emptying of a magazine.” Supra ¶ 90.

¶ 130 I am unsure of the majority's discomfort with that decision. If it thinks the thirty-six hits that I excluded cut against my conclusion, it should say so. The sentences I analyzed are available to the majority at the link I provided, so the majority could easily have examined those thirty-six hits I left out to determine whether I did so appropriately. Absent a specific critique, I will simply stand by the above analysis.

¶ 131 The same goes for the “interpretive assumption[s]” that led to my treatment of other hits. See supra ¶ 21. The COCA data are preserved and available for review. Chief Justice Durrant apparently reviewed them and was satisfied with my conclusions. See supra ¶ 37 n. 10 (indicating that he “found no ... flaws” in his “own examination” of my COCA analysis). The majority is free to do the same. If it disagrees with my “interpretive assumption[s]” or my exclusion of inconclusive hits it should offer reasons for so concluding.

¶ 132 I am unsure what the majority is looking for in terms of statistical significance. I found only one sole instance of discharge of a firearm that had arguable reference to multiple bullets. That was out of eighty-one hits, or forty-four hits after excluding those that were inconclusive or irrelevant. That is statistically significant under any measure. ¶ 133 In all events, the majority's criticisms again are a boomerang for its position. If the concern is for statistical significance, you can't get more insignificant than a data point of one—particularly if the data point is not unbiased. The problem with the majority's analysis, at its core, is that it is rooted in the subjective linguistic intuition of an individual judge (or a panel of them). The aspiration for statistical significance is a good one, but that goal is hardly advanced by the insistence on a judicial prerogative of resolving ambiguity based on pure intuition without resort to objective means for testing it.

IV

¶ 134 The resolution of ambiguities in legal language is one of the most important of all judicial tasks. It is also one of the most opaque; for that reason it is fraught with the potential for bias and error. We should do what we can to minimize those risks. The approach I have outlined is a step in that direction.