concurring in the judgment:
¶ 46 Utah Code section 76-1-403 prescribes a rule of criminal claim preclusion. It states that the prosecution of “one or more offenses arising out of a single criminal episode” may bar the “subsequent prosecution for the same or a different offense arising out of the same criminal episode” if certain conditions are met. Utah Code § 76-1-403(1).
¶ 47 Here we are asked to define the scope of a “single criminal episode.” The majority does so by articulating a “totality of the circumstances” test turning on at least four factors—“the location where the crimes were committed, the nature of the offenses, whether the crimes involved different victims, and whether the defendant had the opportunity to deliberately engage in the next-in-time offense.” Supra ¶12. Applying this test, the court concludes that a prior prosecution of David Rushton on criminal tax charges does not foreclose the present prosecution on charges arising out of his failure to pay earned wages to employees. Because “[tjhree of the four factors” identified by the court “weigh against a finding of a single criminal objective,” the majority holds “that Mr. Rushton’s tax crimes and wage crimes did not have a single criminal objective.” Supra ¶40. And it therefore affirms the court of appeals’ decision allowing the prosecution on the wage offenses to move forward.
¶48 I agree with the judgment of the court—its affirmance of the court of appeals’ decision allowing the wage case to move forward. But I disagree with the majority’s analysis. I would not define the statutory terms by reference to a multi-factored balancing test. Instead, I would articulate a test based on the operative terms of the controlling statute. The operative language precludes subsequent prosecution for criminal “conduct” that is “closely related in time” and “incident to an attempt or an accomplishment of a single criminal objective.” Utah Code § 76-1-401. The phrase “incident to an attempt or an accomplishment of a single criminal objective” is controlling here. I would affirm on the ground that Rushton’s wage offenses were not incident to the attempt or accomplishment of the earlier-charged tax offenses. Accordingly, I would hold that the statute does not preclude prosecution for the wage offenses.
¶ 49 In the paragraphs below I first present the textual basis for the operative test as I see it. Next I identify a significant shortcoming of the majority’s “totality” test—the fact that it will not cover some conduct that is clearly “incident to an attempt or accomplishment” of a principal “criminal objective.” Last I close with some observations about the appropriate domain of balancing tests and the importance of predictability in a field like claim preclusion.
I
¶ 50 The majority rejects Rushton’s formulation of “single criminal episode.” It says that the wage offenses and tax offenses cannot be deemed to stem from a “single criminal objective” simply because they advance a nefarious “objective” framed at a high level *106of generality. Supra ¶16. Quite right. If the identification of an overarching bad purpose were enough to sweep in all charges deemed to advance it, the single criminal episode provision would be all-encompassing. Or at least it could potentially be so for a defendant whose lawyers are creative enough to formulate a nefarious purpose at a high level of generality. And that is not hard to do.
¶ 51 In this case, Rushton asserts that the relevant “single criminal objective” is to prop up his business through general “misappropriation.” That is certainly a nefarious objective; and it would encompass both the wage offenses and the tax offenses in this case. But why stop at this level of generality? An altei’-native formulation would be the purpose of advancing Rushton’s interests by engaging in criminal activity. And that criminal objective would sweep in all conceivable crimes—even crimes as diverse as a sexual assault committed on one day and a count of securities fraud committed three weeks later.
¶ 52 All of this tells us that “single criminal objective” cannot be defined in the abstract. If it were, the preclusive effect of section 76-1-402 would be all-encompassing. And, as the majority notes, this approach would eviscerate the permissive joinder statute, Utah Code section 77-8a-l(l), which permits join-der for offenses that are “based on the same conduct or are otherwise connected together in their commission” or “alleged to have been part of a common scheme or plan.” See supra ¶16. If all crimes said to advance a general purpose of criminality must be charged jointly because they are part of a single criminal episode, then joinder would never be permissive. That cannot be if we are to preserve meaning for the permissive joinder provision.
¶ 53 And that means that we must embrace a narrower notion of “single criminal objective.” The majority does so by considering the “totality of the circumstances,” meaning the series of considerations that the court deems relevant to this inquiry—the location of the two crimes, the nature of the offenses, the identity of the victims, and the opportunity to “deliberately engage in the next-in-time offense.” Supra ¶12. Yet the court proffers no linkage between the factors it includes in its balancing test and the operative terms of the statute.1 Instead, the court imports a test set forth in a federal court of appeals decision under a federal statute that bears little resemblance to the operative Utah provisions. See supra ¶¶36-39.2
*107¶ 54 The cited federal case is United States v. Letterlough, 63 F.3d 332 (4th Cir. 1995). Letterlough is a case arising under the federal Armed Career Criminal Act, 18 U.S.C. § 924(e)(1). The cited statute provides for a sentencing enhancement—a mandatory minimum of fifteen years to life—for a person convicted of unlawful possession of a firearm under 18 U.S.C. § 922(g) who “has three previous convictions ... for a violent felony or serious drug offense, or both, committed on occasions different from one another.” 18 U.S.C. § 924(e)(1). The Letterlough court articulated factors of relevance to the determination whether past crimes were “committed on occasions different from one another.” Id, .Citing cases from other circuits, the Letter-lough court held that “a multiplicity of factors” help to indicate “when more than one conviction constitutes a separate and distinct criminal episode.” 63 F.3d at 335. And the cited factors include those embraced by the majority in this case—“whether the offenses arose in different geographic locations; whether the nature of the offenses was substantively different; and whether the offenses involved multiple victims or multiple criminal objectives.” Id. at 335-36 (footnote omitted)'. In addition, the Letterlough court also considered whether the defendant had “the opportunity to make a conscious and knowing decision to engage in [the next-in-time offense.]” Id. at 337.
¶ 55 These factors may well be appropriate in assessing whether a series of crimes were “committed on occasions different from one another.” But the quoted language comes from 18 U.S.C. § 924(e)(1). And those terms appear nowhere in the operative Utah provision, Utah Code section 76-1-401.
¶ 56 The question under section 401 is not whether a later-charged crime was committed on an occasion different from that .of an earlier-charged offense. It is whether the two sets of offenses are “closely related in time” and are “incident to an attempt or an accomplishment of a single criminal objective.” Utah Code § 76-1-401. Thus, so long as the time-relatedness element is met,3 subsequent prosecution is barred only if it arises out of conduct that is “incident to an attempt or an accomplishment of a single criminal objective.” Id.
¶ 57 The statutory phrase “incident to” modifies or extends to both “an attempt ... of a single criminal objective” and “an accomplishment of a single criminal objective.” It accordingly defines the scope of “conduct” that comprises a “single criminal episode.” Conduct that is not “incident to” an “attempt!’ or “accomplishment” of a “single criminal objective” may be charged separately because it does not comprise a “single criminal episode” under the statute.
¶ 58 The words “attempt” and “accomplishment” have well-defined meanings in the criminal law. An “attempt” is itself a crime involving conduct that is a “substantial step toward commission of [a specified] crime” wherein the defendant “intend[ed] to commit the [specified] crime.”, Utah Code § 76-4-101(1).4 And the term “accomplishment” or “accomplish” refers to the completion of all the elements of a crime identified in the *108criminal code.5 Because this statute appears in the criminal code, speaks in terms of a criminal objective, and identifies timing requirements for criminal prosecutions, I would give these terms their criminal law meanings. See In re Adoption of Baby E.Z., 2011 UT 38, ¶ 109, 266 P.3d 702 (Lee, J., concurring) (“ “Words of art bring their art with them,’ and courts have commonly assumed that “where [a legislature] borrows terms of art ..., it presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken and the meaning its use will convey to the judicial mind unless otherwise instructed.’ ” (second alteration in original) (footnote and citations omitted)).
¶ 59 In their criminal law senses, both attempt and accomplishment are used in relationship to an identifiable offense in the criminal code. They refer, in other words, to a specific crime that was attempted or accomplished. Accordingly, in the context of the statutory definition of “single criminal episode,” these terms have reference to the “objective” of either attempting or accomplishing a specific offense. So the “objective” referenced in the statute is not, as Rushton asserts, some hazy nefarious purpose. Nor is it an unlawful purpose defined vaguely by a totality of the circumstances. It has reference, rather, to a particular crime—a “single” “objective” crime that the defendant either “attempt[ed]” or “accomplishfed].” See Utah Code § 76-1-401. Thus, for the prosecution of a crime to be barred by the statute, the conduct underlying that crime must have been “incident to” the attempt or accomplishment of the identified objective crime.
¶ 60 The “incident to” element is the language of connection or causation.6 One event or act is “incident to” another if it arises out of it or is otherwise connected to it—as in a risk “incident to” employment or a search “incident to” an arrest. I would interpret the statutory reference to an offense “incident to an attempt or an accomplishment” of the objective crime in this sense. I would require proof of some connection or relationship between the two offenses.7
¶ 61 The statutory test accordingly consists of two steps. First, the court must identify the “single criminal objective”—the objective crime—at issue.8 Second, the court *109must determine whether the defendant’s other criminal conduct was incident to the attempt or accomplishment of the objective crime.
¶ 62 Because the statute prohibits successive “prosecution” for offenses that are part of a single criminal episode, see Utah Code § 76-1-403, I would conduct this analysis by reference to the documents setting forth the basis for the prosecution (the criminal information or indictment). And I would hold that Rushton’s wage offenses were not incident to an attempt or accomplishment of his tax offenses. Rushton made no effort to show any relationship between the conduct underlying his tax offenses and the conduct underlying his wage offenses—let alone one that could be described as “incident to,” under the statute. He argued only that both the tax offenses and the wage offenses fall under the same general objective of misappropriating funds. That is plainly insufficient. To fall within the preclusive sweep of the statute, Rushton’s tax offenses would have to be shown to be incident to—or directly and immediately relating to—an attempt or accomplishment of his wage offenses. No such showing was made here, and Rushton’s argument should fail on that basis.
II
¶ 63 The majority takes a completely different tack. It adopts instead a multi-fac-tored balancing test. But that test is completely disconnected from the operative statutory language. And it will also lead to problematic results.
¶ 64 An inchoate offense, in my view, is clearly “incident to an attempt or an aecom-plishment of a single criminal objective.” Utah Code § 76-1-401. Surely a conspiracy to commit bank robbery, for example, is “incident to an attempt or an accomplishment” of the bank robbery. So if such offense is “closely related in time” to the principal offense it is aimed at facilitating, it must necessarily be barred if not included in the prosecution of the principal offense.
¶ 66 Yet the majority’s multi-factored balancing test could easily lead to the opposite conclusion. Consider a case involving a bank robbery preceded by solicitation of an accomplice. The act of solicitation is undoubtedly “incident to” the “accomplishment” of the “criminal objective” of bank robbery. So the crime of solicitation should be foreclosed if closely related in time and not charged in connection with an initial prosecution for the bank robbery. But that crime may not be barred under the majority’s test. If, for example, the accomplice resides in a different state, the majority’s factors could easily weigh against a requirement of joinder in the initial bank robbery case, since the offenses could be said to arise in a “different geographic location[],” supra ¶36; the solicitation offense is “substantively different” in “nature” from bank robbery, supra ¶37; and the defendant “had ‘the opportunity to make a conscious and knowing decision to engage in’ the next-in-time offense,” supra ¶39.9 That makes three of four factors that cut against the conclusion that these offenses are “incident to” the “accomplishment” of the “criminal objective” of bank robbery.
¶ 66 This is troubling. If the factors we identify fail to sweep in'classic crimes “incident to an attempt or accomplishment of a single criminal objective,” then it’s time to go *110back to the drawing board. I would do so by giving the terms attempt, accomplishment, and criminal objective the meaning those terms have in the criminal law. And I would hold that Rushton’s wage offenses are not precluded here because he has made no effort to identify any relationship between the conduct underlying the two distinct sets of crimes.
¶ 67 The majority objects to my test as underinclusive, citing an example of a series of thefts “from1 ten separate accounts belonging to ten separate clients of the same bank,” brought about by a “computer program” written by a defendant who admits to a goal of stealing “small enough increments that (in his view) were less likely to immediately trigger the bank’s anti-fraud measures.” Supra ¶¶24-25, In the court’s view, these offenses “ha[ve] a single criminal purpose underlying” them, as witnessed by the defendant’s contemporaneous admission of his intent. Id. ¶26. And because my test would allow “serial prosecutions” of these multiple choate offenses, the majority insists that my approach must be dismissed as un-derinclusive. Id.
¶ 68 I disagree for several reasons. For one thing, the court’s conclusions are premised on its own take-our-word-for-it sense of the scope of “single criminal episode.” It confidently announces its belief that “[t]he state should not be allowed to bring serial prosecutions against” a defendant in these circumstances. Id. But it offers no basis in the words of the statute for that conclusion. So the court’s criticism (of the underinclu-siveness of my test) is entirely circular. In my view, it is the majority’s test that is under- and overinclusive—or at least potentially so, in that the multi-factored balancing test makes it impossible to predict with any certainty which crimes will be deemed to be encompassed within a “single criminal episode.”
¶ 69 Second, the practical effect of my test is not nearly as sweeping as the court imagines. “[Sjerial prosecutions” will ensue only if the prosecution chooses to exercise its discretion to charge these crimes separately. And a rational prosecutor seems highly likely to charge these hypothetical crimes together for a range of practical reasons. Even if related cases are charged separately, moreover—as with the court’s secondary example of “a hacker who simultaneously acquires unauthorized access to one million computers,” id.— there is no reason to expect “one million trials.” Id. Cases can be charged separately but ultimately consolidated for trial. Utah R. Crim. P. 34. So the majority overstates the practical concern implicated by my test,
¶ 70 Finally, the majority’s concerns about underinclusiveness are not necessarily alleviated by its test. The virtue of a “totality of the circumstances” test is its flexibility. But that is also its principal vice. And I cannot see how a prosecutor could predict how the multiple bank theft or computer hacking counts would fare under the majority’s test. In both cases there are some factors pointing in favor of mandatory joinder (location, similar conduct) and others pointing the other way (different victims, neither offense is incident to the attempt or accomplishment of the other). The majority, moreover, goes out of its way to say that the listed considerations “are certainly not the only factors relevant to the mandatory joinder analysis.” Supra ¶35. So it seems impossible to anticipate with any certainty how the court’s own hypotheticals would come out under its test. It is entirely possible that the examples it lists would come out the same under both tests.
Ill
¶ 71 A totality-of-the-circumstances test is a tempting response to a complex legal problem. And such a test may have a place in the law—in a field, for example, where precision is untenable (or unimportant) and flexibility is at a premium.10 But this is not such a field.11 The flipside of flexibility is unpredict*111ability. And predictability is at a premium in the field of claim preclusion.
¶ 72 A prosecutor faced with the question of how widely to frame an information or indictment must anticipate the preclusive effect of such a decision.12 A misjudgment in this field will produce drastic consequences— the preclusion of a criminal claim without any consideration of the merits. Such a consequence should be the predictable result of a reasoned decision. And we thwart that possibility when we do no more than articulate a series of “factors” to be balanced in some unspecified way.
¶73 I would avoid that problem here. I would do so by rejecting the majority’s “totality” test in favor of a more objective rule rooted in the operative terms of the governing statute. And I would affirm under that test.
. I see a parallel between the majority opinion here and that in State v. Shickles, 760 P.2d 291, 295 (Utah 1988). Shickles identified factors to *107guide decisions weighing evidence under rule 403 of the Utah Rules of Evidence. Id. at 295-96. Yet the Shickles factors were not rooted in the text of the operative rule; they were “drawn from” the McCormick on Evidence treatise. State v. Lucero. 2014 UT 15, ¶ 32, 328 P.3d 841, And in time we were forced to repudiate. the Shickles factors—identifying unforeseen consequences arising from an attempt to formulate factors not rooted in the text of the governing law, and backing away from the Shickles factors and pointing the courts back to rule 403. Id,
I foresee a similar path ahead under the single criminal objective statute. The Rushton factors seem destined to be applied in unforeseen ways in future cases, leading to results that cannot be squared with the operative text of the statute. When that happens we will be forced to formulate a test more tied to the terms of the statute.
. I would consider the time-relatedness factor separately. That is an independent element of a • “single criminal episode.” And it confuses the matter—and double-counts the time element—to consider time as a factor of relevance to the requirement of a "single criminal objective.” See infra ¶65 n.9.
. See also Black's Law Dictionary 146 (10th ed. 2014) (defining "attempt” as "1. The act or an instance of making an effort to accomplish something, esp. without success. 2. Criminal law. An overt act that is done with the intent to commit a crime but that falls short of completing the crime.”).
. See, e.g., State v. Castonguay, 663 P.2d 1323, 1326 (Utah 1983) (stating in an attempted murder case that "it must be borne in mind that an attempt transcends intent, yet fails to culminate in its planned accomplishment”); State v. Musser, 110 Utah 534, 175 P.2d 724, 731 (1946) (stating that "a criminal conspiracy ... consists of an unlawful agreement plus some overt act or acts done to further or to accomplish the object of such an agreement"; also noting that "[a]n act done in furtherance of an agreement need not succeed in accomplishing its objective in order to fulfill the requirements of the statute”) (emphases added), vacated and remanded on other grounds in Musser v. Utah, 333 U.S. 95, 68 S.Ct. 397, 92 L.Ed. 562 (1948); State v. Mortensen, 95 Utah 541, 83 P.2d 261, 262-63 (1938) (reversing attempted rape conviction on the ground that there was “no act which can be said to be designed to accomplish the act of intercourse”).
. See Black's Law Dictionary 146 (10th ed. 2014) (defining "incident” as “[d]ependent on, subordinate to, arising out of, or otherwise connected with (something else, usu. of greater importance) <the utility easement is incident to the ownership of the tract>”); Webster’s Third New International Dictionary 1142 (3d ed. 2002) (setting forth the "law” definition of the word "incident” when used as an adjective: "dependent on or appertaining to another thing: directly and immediately relating to or involved in something else though not an essential part of it.”).
. Perhaps this inquiry will introduce some uncertainty. That is one of the majority's complaints about my approach. See supra ¶24. But the "incident to” element is an explicit component of the statutory test. If we are concerned about the fuzziness of the inquiry then we should do our best to bring clarity to the analysis; we are in no position to write it out of the law (as the majority does in its test). See supra ¶45 ("Because Mr. Rushton's conduct did not have a single criminal objective, it was not a single criminal episode under the mandatory joinder statute.” (emphasis added)). I would bring clarhy to "incident to” by drawing upon analogies in other areas of the law, which treat "incident to” as the language of connection or causation. Our cases can iron out any wrinkles in this formulation in due time.
The majority, moreover, is in no position to make the plea for certainty. Its "totality of the circumstances” test requires uncertainty from case to case. See infra ¶¶71-72 (noting further concerns about the unpredictability of the majority's test).
. Either of two crimes could be the "objective.” The analysis would turn on whether the conduct underlying one of the crimes at issue was incident to the attempt or accomplishment of another crime.
. The majority's test also seems problematic for an additional reason: Even if we assume the propriety of a multi-factored test, the majority's standard seems to double-count the timing element. The operative statute tells us that time-relatedness is an inquiry distinct from the "single criminal objective1' element. See Utah Code § 76-1-401 (defining "single criminal episode" as one involving two sets of offenses that are "closely related in time" and "incident to an attempt or an accomplishment of a single criminal objective”). So the assessment of "opportunity to make a conscious and knowing decision" seems to amount to double-counting.
I have no quarrel with the notion that "overlap is often a feature of legal tests.” Supra ¶39 n.17. But my point is not to raise a general objection to "overlap” in all circumstances; it is to suggest that the separate existence of a "closely related in time” test calls into question the statutory basis for a time component of the "single criminal objective” test. Thus, the existence of overlap in legal tests formulated in other areas of the law tells us nothing about the appropriate standard for analyzing a "single criminal objective.” And the separate nature of the “closely related in time” standard undermines the existence of a time component for assessing "single criminal objective.”
. See United States v. Mead Corp., 533 U.S. 218, 241, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) (Scalia, J., dissenting) (deriding "th’ol’ 'totality of the circumstances’ " as "that test most beloved by a court unwilling to be held to rules (and most feared by litigants who want to know what to expect)”).
. There is a rich literature on the virtues and vices of objective rules and subjective standards. Nearly everyone agrees that there is a place in *111the law for both. But even the pro-standards crowd acknowledges that balancing tests are problematic in fields in which predictability is at a premium. Compare Antonin Scalia, The Rule of Law As a Law of Rules, 56 U. Chi. L. Rev. 1175, 1182 (1989) (stating that "predictability is destroyed” when courts adopt tests based on the "totality of the circumstances” or "a balancing of all the factors involved") with Cass R. Sunstein, Must Formalism Be Defended Empirically?, 66 U. Chi. L. Rev. 636, 654 (1999) (acknowledging that although the "issue[] [is] hard to resolve in the abstract,” ”[w]here predictability is especially important—such as in areas involving commercial and criminal law—formalism [which seeks to limit judicial discretion] might be favored.”).
. Surely this is an additional “purpose of the mandatory joinder statute.” See supra ¶35 (listing other purposes). We should not assume that a statute like this one is aimed at accomplishing only a limited set of objectives on one side of the scale.. We should recognize the reality that this statute, like most all others, is aimed at balancing competing objectives, including that of the prosecution in the exercise of its discretion. See Olsen v. Eagle Mountain City, 2011 UT 10, ¶ 23 n.6, 248 P.3d 465 ("[M]ost statutes represent a compromise of purposes advanced by competing interest groups, not an unmitigated attempt to stamp out a particular evil.”).