United States v. Tre Tate

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 21a0121p.06 UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT ┐ UNITED STATES OF AMERICA, │ Plaintiff-Appellee, │ > No. 20-5071 │ v. │ │ TRE RESHAWN TATE, │ Defendant-Appellant. │ ┘ Appeal from the United States District Court for the Eastern District of Tennessee at Knoxville. No. 3:19-cr-00026-1—Thomas A. Varlan, District Judge. Decided and Filed: May 28, 2021 Before: CLAY, READLER, and MURPHY, Circuit Judges. _________________ COUNSEL ON BRIEF: Erin P. Rust, FEDERAL DEFENDER SERVICES OF EASTERN TENNESSEE, INC., Chattanooga, Tennessee, for Appellant. Cynthia F. Davidson, UNITED STATES ATTORNEY’S OFFICE, Knoxville, Tennessee, for Appellee. READLER, J., delivered the opinion of the court in which CLAY, J., joined. MURPHY, J. (pp. 16–28), delivered a separate opinion concurring in the judgment. _________________ OPINION _________________ CHAD A. READLER, Circuit Judge. For the crime of bank robbery, the Sentencing Guidelines establish a tiered structure of sentencing enhancements based upon how an individual effectuates the robbery. See U.S.S.G. § 2B3.1(b)(2). When a defendant discharges a firearm while robbing a bank, the Guidelines prescribe a seven-level increase to his base offense level. No. 20-5071 United States v. Tate Page 2 Id. § 2B3.1(b)(2)(A). Less serious conduct results in a smaller boost to one’s offense level. For instance, a defendant who merely brandishes or possesses (but does not discharge) a firearm receives an increase of five offense levels. See id. § 2B3.1(b)(2)(C). Near the lower end of this sentencing hierarchy is the setting in which a defendant brandishes or possesses a “dangerous weapon” while committing a robbery. Id. § 2B3.1(b)(2)(E). Doing so subjects the defendant to a three-level increase to his base offense level. Id. Tre Reshawn Tate received this three-level enhancement for concealing his hand in a bag to suggest the existence of a dangerous weapon while robbing a bank. Given the text and context of § 2B3.1(b)(2)(E), the district court correctly included the enhancement in Tate’s sentence. We thus affirm. BACKGROUND A man wearing a black knit cap, sunglasses, and a dark jacket covering the lower portion of his face entered a bank with an opaque shoulder bag across his body. As he approached a bank teller, the man displayed a note that read: “You have 30 seconds to give me 20 thousand dollars in Big Bills an[d] I’ll let Everyone live Don’t Make A Sound!” The man then began audibly counting down from thirty. As he reached single digits, the man placed his right hand into his shoulder bag in a manner that led the teller to believe the robber was about to pull out a gun. The teller responded by reaching into a bank till to retrieve what ended up being $12,000 and handed it to the man. The man stuffed the cash into his bag and fled the scene. While initially unsuccessful in apprehending the robbery suspect, officers did obtain several items used in the robbery, including the suspect’s cap, jacket, and sunglasses. Forensic examiners were able to match a fingerprint found on the sunglasses to one that was recently added to an FBI database. The fingerprint was that of Tre Reshawn Tate, who lived a mile from the bank. Tate’s age and physical characteristics also matched those of the robber. And after obtaining a search warrant for Tate’s DNA, officers were able to confirm that Tate’s DNA was found on the cap and jacket. A federal grand jury indicted Tate on charges stemming from the robbery. Tate later pleaded guilty to violating 18 U.S.C. § 2113 (a), the federal bank robbery statute. During No. 20-5071 United States v. Tate Page 3 sentencing, the probation office suggested adding three levels to Tate’s total offense level in accordance with the “Specific Offense Characteristics” set out in U.S.S.G. § 2B3.1(b) due to Tate having “brandished or possessed” a “dangerous weapon” during a bank robbery. Tate objected, maintaining that there was insufficient evidence in the record to apply the § 2B3.1(b)(2)(E) enhancement because he did not possess a dangerous weapon during the robbery. Relying in part on a Guidelines application note indicating that a dangerous weapon can include using an object in a manner that creates the impression that the object was capable of inflicting serious injury, the district court overruled Tate’s objection. From the resulting Guidelines imprisonment range of 41 to 51 months, the district court imposed a sentence of 41 months’ imprisonment. Tate’s timely appeal followed. ANALYSIS Tate contends that his sentence is procedurally unreasonable because the district court incorrectly calculated his Guidelines range by applying § 2B3.1(b)(2)(E)’s dangerous-weapon enhancement. According to Tate, the district court improperly deemed Tate’s act of hiding his hand in a bag in a suggestive manner as amounting to “brandish[ing] or possess[ing]” a “dangerous weapon” worthy of applying the enhancement. To do so, says Tate, the district court had to expand the generally accepted meaning of the term “dangerous weapon” by reference to the Guidelines commentary section, thereby violating our command in United States v. Havis that the Guidelines’ “application notes are to be ‘interpretations of, not additions to, the Guidelines themselves.’” 927 F.3d 382, 386 (6th Cir. 2019) (en banc) (quoting United States v. Rollins, 836 F.3d 737, 742 (7th Cir. 2016) (en banc)) (holding that the Commission’s use of commentary to expand the reach of an otherwise clear Guideline “deserves no deference”); see also United States v. Riccardi, 989 F.3d 476, 479 (6th Cir. 2021) (“But guidelines commentary may only interpret, not add to, the guidelines themselves.”). Alternatively, Tate argues that even if the district court was correct in its reading of § 2B3.1(b)(2)(E), Tate’s action of placing his hand in the shoulder bag falls outside the range of conduct to which the enhancement should apply. 1. When reviewing a sentence’s procedural reasonableness, we ordinarily apply an abuse-of-discretion standard. See Gall v. United States, 552 U.S. 38, 51 (2007). No. 20-5071 United States v. Tate Page 4 The government, however, asserts that our review of Tate’s challenge to the calculation of his Guidelines range is confined to “plain error” due to Tate’s failure to preserve the argument. On that score, we note that while Tate did challenge the application of the § 2B3.1(b)(2)(E) enhancement before the district court, the precise nature of his challenge was somewhat different than how he frames his argument today. Whether that distinction means Tate’s argument today was unpreserved, and thus subject to plain error review, is a fair point of debate. Cf. United States v. Fleming, 894 F.3d 764, 771 (6th Cir. 2018) (evaluating the defendant’s argument under both plain error and abuse of discretion where the defendant’s general objection to an upward variance was “arguably sufficient to preserve” the specific procedural unreasonableness at issue). But we need not resolve that debate here, as Tate’s challenge fails even under the more forgiving abuse-of-discretion standard, by which we review legal errors de novo. United States v. Stubblefield, 682 F.3d 502, 510 (6th Cir. 2012). a. Section 2B3.1(b)(2)(E) imposes a three-level enhancement if a “dangerous weapon was brandished or possessed” during a covered robbery. The interpretative question Tate raises is whether one’s hand, confined in a bag, amounts to a “dangerous weapon [] brandished or possessed” during his bank robbery. Tate believes it does not. A “dangerous weapon,” says Tate, is not one’s hand when covered in a bag. Instead, it must be something that, when used in its ordinary course, is “able or likely to cause injury [when] used against another.” See Appellant’s Br. at 18 (citing Merriam-Webster Dictionary’s definition of “dangerous” and “weapon”). In matters of textual interpretation, however, “literal or dictionary definitions of words will often fail to account for settled nuances or background conventions that qualify the literal meaning of language and, in particular, of legal language.” John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387 , 2393 (2003). As a result, we customarily measure a term’s meaning with an eye on the broader context in which the term appears as well as the term’s original public meaning at the time of enactment. See United States v. Grant, 979 F.3d 1141, 1144 (6th Cir. 2020) (observing that a court should not “mechanistically pars[e] down each word of the statute to its dictionary definition, no matter the resulting reading that would give the law”); Antonin Scalia, A Matter of Interpretation 23 (1997) (“A text should not be construed No. 20-5071 United States v. Tate Page 5 strictly, and it should not be construed leniently; it should be construed reasonably, to contain all that it fairly means.”). In other words, we do not woodenly interpret a legal text “in a vacuum,” see Abramski v. United States, 573 U.S. 169 , 179 (2014), but instead discern “the meaning of a statement” in a law from the “context in which it is made,” see United States v. Briggs, 141 S. Ct. 467, 470 (2020). We customarily employ these settled principles in interpreting acts of Congress. See, e.g., New Prime Inc. v. Oliveira, 139 S. Ct. 532, 539 (2019). So too our interpretation of the Sentencing Commission’s words, which are a product of delegated authority for rulemaking akin to an administrative agency when enacting a legislative rule. See Stinson v. United States, 508 U.S. 36, 44 (1993); see also United States v. Reaves, 253 F.3d 1201, 1203 (10th Cir. 2001) (“We interpret the Sentencing Guidelines according to accepted rules of statutory construction.”). Turning to the use of the phrase “dangerous weapon” in the federal robbery Guidelines, the phrase’s meaning encompasses not only objects that are per se dangerous, but also those that, by their objective appearance, create the possibility of danger. Telling here is the fact that the Supreme Court adopted this reading of the phrase “dangerous weapon” in an analogous context just a year before the Guidelines first employed the term. In McLaughlin v. United States, a unanimous Supreme Court held that the term “dangerous weapon” as used in 18 U.S.C. § 2113 (d)—a statutory penalty enhancement applicable to one who uses a dangerous weapon during a bank robbery—extended to the use of an unloaded gun during the robbery. 476 U.S. 16 , 17–18 (1986). Employing a functional approach to that phrase, the Supreme Court provided three “independently sufficient” reasons why an unloaded gun was nonetheless a “dangerous weapon”: (1) a gun is “typically and characteristically dangerous”; (2) “the display of a gun instills fear in the average citizen,” risking a violent response; and (3) a gun can be used as a bludgeon. Id. McLaughlin did not arrive at this conclusion from an unsettled foundation. Instead, it built on decisions from state and federal courts holding that a robber can effectuate his crime with a “dangerous weapon” even if the instrument in question could not have produced injury if used in its customary manner. See, e.g., Commonwealth v. Tarrant, 326 N.E.2d 710 , 713–14 (Mass. 1975) (discussing Massachusetts’s long-held interpretation of its armed robbery statute); No. 20-5071 United States v. Tate Page 6 see generally Baker v. United States, 412 F.2d 1069, 1072 (5th Cir. 1969) (recounting “[n]umerous cases hold[ing] that one may be convicted of robbery by means of a dangerous weapon notwithstanding the fact that the gun allegedly used was unloaded”); Brief for the United States, McLaughlin v. United States, 476 U.S. 16 (1986) (No. 85-5189), 1985 WL 670258 at *16 (observing that “states have nearly unanimously concluded that an unloaded gun used in a robbery is a dangerous weapon”). As in McLaughlin, these courts understood that a robbery committed with a non-inherently dangerous instrument can produce fear “equivalent” to when an inherently dangerous weapon is present, “naturally lead[ing]” to the same risks of “resistance and conflict.” Tarrant, 326 N.E.2d at 713 (quoting Commonwealth v. Mowry, 93 Mass. (11 Allen) 20 , 22–23 (1865)); see also United States v. Bennett, 675 F.2d 596, 599 (4th Cir. 1982); United States v. Crouthers, 669 F.2d 635, 639 (10th Cir. 1982); United States v. Beasley, 438 F.2d 1279, 1283 (6th Cir. 1971). In other words, the federal robbery Guidelines’ commentary merely echoes how some courts have long viewed a dangerous weapon to include both objects that are (1) per se dangerous, as well (2) those that are used in a manner that is likely to endanger life. See generally 67 Am. Jur. 2d Robbery § 97 (“To be considered a ‘dangerous weapon,’ the weapon need not actually be capable of inflicting severe bodily harm or injury upon another; rather, it may be considered dangerous if it instills fear in the average citizen, creating an immediate danger that a violent response will follow.”). Well before the adoption of U.S.S.G. § 2B3.1(b)(2)(E), courts had held that a robber who used an unloaded gun, an inoperable firearm, or a body part to create the credible impression, through concealment, that he possessed a firearm was viewed as employing a deadly or dangerous weapon in a robbery. See, e.g., People v. Raleigh, 16 P.2d 752, 753 (Cal. Ct. App. 1932) (unloaded gun); Crum v. State, 227 A.2d 766, 767 (Md. 1967) (inoperable gun); Stewart v. State, 443 So. 2d 1362, 1364 (Ala. Crim. App. 1983) (hidden body part). (While some states created an evidentiary presumption that a robber simulating possession of a firearm had a “dangerous weapon,” see, e.g., Commonwealth ex rel. Johnson v. Myers, 189 A.2d 331, 332 (Pa. Super. Ct. 1963) (per curiam), we see little difference, other than semantics, between those cases and more modern cases expressly holding that a simulated weapon can constitute a dangerous weapon.) True, as our concurring colleague notes, some states, as is their right, have not adopted the functional view of a “dangerous weapon” when interpreting their own state robbery laws, while others take a more precise route to cover No. 20-5071 United States v. Tate Page 7 simulated weapons. Yet the United States Supreme Court, in McLaughlin, was well aware of that competing state authority when it embraced a functional view of what amounts to a “dangerous weapon” in the context of federal bank robbery law. 476 U.S. at 17–18. That holding deserves emphasis when assessing the Sentencing Commission’s adoption of the phrase “dangerous weapon.” Just a year after McLaughlin, the Commission included the same term in its Guideline addressing federal robbery offenses, including those under § 2113. See U.S.S.G. § 2B3.1 (Oct. 1987). When the Supreme Court has authoritatively interpreted a term as it is used in a particular field of law, the term acquires a “technical legal sense . . . that should be given effect in the construction of later-enacted” laws and rules governing that field. See Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 324 (2012). Not surprisingly, then, in the years that followed § 2B3.1’s adoption, courts routinely understood the Guidelines’ use of the term “dangerous weapon” in the robbery context to have incorporated McLaughlin’s view of the term. See United States v. Dixon, 982 F.2d 116, 123 (3d Cir. 1992) (“In Guideline cases, courts have thus relied on McLaughlin to hold that a defendant brandished, possessed or displayed a ‘dangerous weapon’ . . . .”); United States v. Gray, 895 F.2d 1225, 1226 (8th Cir. 1990) (per curiam) (holding that the McLaughlin standard applies to the dangerous weapon enhancement); United States v. Laughy, 886 F.2d 28, 30 (2d Cir. 1989) (per curiam) (same). For our part, we have observed that the dangerous-weapon sentencing enhancement works part and parcel with the law McLaughlin interpreted in that the Guidelines’ enhancement applies when the defendant’s conduct falls short of satisfying the statutory sentencing enhancement in § 2113(d) for using a dangerous weapon during a bank robbery. See United States v. Perry, 991 F.2d 304 , 310 n.2 (6th Cir. 1993). In view of this reading of the phrase “dangerous weapon,” the Sentencing Commission, we note, later amended its application notes to their current form to make clear that the courts’ uniform application of McLaughlin in interpreting § 2B3.1(b) was correct. See U.S.S.G. App. C amend. 601 (2000); see also United States v. Rodriguez, 301 F.3d 666, 668 (6th Cir. 2002) (observing that the amendment’s purpose was to “make clear” that an object used to create the impression that it was capable of inflicting death or seriously bodily injury was a “dangerous weapon”); United States v. Davis, 635 F.3d 1222, 1225 (D.C. Cir. 2011) (explaining that the No. 20-5071 United States v. Tate Page 8 “current version of the Commentary all but codifies” holdings embracing the functional view of dangerous weapon). Today, those application notes explain that a dangerous weapon is an object that is “capable of inflicting death or seriously bodily injury” or one that is used “in a manner that creates the impression that the object was such an instrument.” See U.S.S.G. § 2B3.1 cmt. n.2 (incorporating § 1B1.1 cmt. n.1(E)). Albeit not expressly, in applying the Guidelines’ federal robbery provisions, we have read those notes as simply echoing the Guidelines themselves, including their adoption of McLaughlin’s functional approach. See, e.g., United States v. Woodard, 24 F.3d 872 , 873–74 (6th Cir. 1994) (holding that a toy gun could be a dangerous weapon for purposes of the enhancement, as the toy’s use could put “people’s lives in jeopardy”); see also Rodriguez, 301 F.3d at 668–69 (applying the functional approach to conclude that a Styrofoam sandwich box that a robber brought with him into a bank could have been “reasonably regarded as a dangerous weapon; namely, a bomb” under the enhancement). As such, the unambiguous text of the Guidelines enhancement for dangerous weapons applies whether a robber is, or merely pretends to be, armed. Our holding, we emphasize, is specific to the federal dangerous weapons sentencing enhancement as applied to robbery offenses. Cases arising with respect to different facts or different laws could dictate a different result. In the context of federal robbery law, McLaughlin’s functional approach requires examination of how a particular object was used under the specific circumstances at issue. See Woodard, 24 F.3d at 873–74. More broadly, the term “dangerous weapon” need not necessarily be interpreted and applied in the same manner across every law. See Yates v. United States, 574 U.S. 528 , 537 (2015) (plurality opinion) (“In law as in life, . . . the same words, placed in different contexts, sometimes mean different things.”); see, e.g., People v. Peralta, 770 N.Y.S.2d 339, 341 (N.Y. App. Div. 2004) (holding that an unloaded gun is not a “dangerous instrument” under New York penal law defining that term as an object “readily capable of causing death or other serious physical injury”), State v. Godfrey, 20 P. 625, 628 (Or. 1889) (concluding that a “dangerous weapon” within the meaning of Oregon’s criminal assault statute must be inherently “capable of producing death or great bodily harm”). For instance, whether an unconcealed part of the human body amounts to a dangerous weapon for purposes of various state and federal crimes is a well-joined debate. See 67 A.L.R. 6th 103 (2011). That debate, however, customarily arises outside the robbery context No. 20-5071 United States v. Tate Page 9 and/or where the threat in question does not implicate the same dangers of a robber pretending to have a concealed firearm. Compare State v. Hinton, 639 S.E.2d 437, 441 (N.C. 2007) (concluding that a “defendant’s hands, in and of themselves, cannot be dangerous weapons” under N.C.G.S. § 14–87), with Bennett, 675 F.2d at 599 (explaining the unique risks raised by a robber pretending he has a threatening weapon in a robbery). Instead, those cases query whether a particular exposed body part presents a serious enough risk of injury to be deemed a “dangerous weapon.” Compare United States v. Rocha, 598 F.3d 1144, 1157 (9th Cir. 2010) (holding that a federal prisoner, by pulling a fellow inmate’s ankles and forcing him to the floor, did not commit an assault with a “dangerous weapon” under the federal assault statute, 18 U.S.C. § 113 ), with United States v. Sturgis, 48 F.3d 784, 785 (4th Cir. 1995) (concluding that a jury “could reasonably have concluded” that a HIV-positive prisoner who bit two correctional officers used his mouth and teeth as a “dangerous weapon” in violation of 18 U.S.C. § 113 ). Here, of course, no one would dispute that a gun is a “dangerous weapon,” leaving the question whether creating the impression of possessing a gun can trigger the sentencing enhancement applicable to robberies. We hold that it does. b. Tate sees things differently. He contends that the enhancement could only be applied to him through reliance on the application notes to the Guidelines. Here, Tate explains that while no definition of the term “dangerous weapon” is included in the Guidelines themselves, the relevant application notes interpret the term to include both an instrument that is “capable of inflicting death or serious bodily injury” and an object that is not such an instrument but either closely resembles such an instrument or is used in a manner to create the impression that the object is such an instrument. See U.S.S.G. § 1B1.1 cmt. n.1(E); see also id. § 2B3.1 cmt. n.2. And, as Tate fairly and persuasively contends, because the “commentary to the Guidelines,” unlike the Guidelines themselves, “never passes through the gauntlets of congressional review or notice and comment,” Havis, 927 F.3d at 386 , a court may not rely on a commentary note that inconsistently expands the scope of the corresponding Guideline. Id. But contrary to Tate’s contention, the application notes are not the tail that wags the Guidelines’ dog. Rather, the Guidelines themselves incorporate McLaughlin’s view of a dangerous weapon. As a result, the commentary Tate invokes “does not purport to add to No. 20-5071 United States v. Tate Page 10 (or contradict) the text of the Guidelines,” meaning “it poses no problem under this circuit’s precedent in Havis.” United States v. Murphy, 815 F. App’x 918 , 925 (6th Cir. 2020) (Thapar, J., concurring). Today’s case, in other words, does not ask us to defer to an agency’s interpretation of an undefined or ambiguous phrase in a regulation, thereby allowing the agency to be the judge and jury of the limits of its own power. See Kisor v. Wilkie, 139 S. Ct. 2400 , 2433–34, 2437 (2019) (Gorsuch, J., concurring). c. For similar reasons, we appreciate yet ultimately are unpersuaded by our concurring colleague’s conclusion that application of the “dangerous weapon” enhancement is limited to (1) inherently dangerous objects and (2) non-inherently dangerous objects that can actually be used to inflict bodily harm under the circumstances (like a chair or a cane). The concurrence begins its thoughtful analysis by asking how an “ordinary English speaker” would understand what it means to possess a dangerous weapon, concluding that “[n]o ordinary English speaker would say that a robber possess a dangerous weapon when the robber merely pretends to possess one.” Id. at 1. A fair inquiry, to be sure, were we writing on a clean slate. But in American law, at least, few slates these days are clean. That is the case here. While an ordinary English speaker may not consider a robber who uses an unloaded firearm to be using a dangerous weapon, the Supreme Court has said otherwise, a pronouncement that informed the Sentencing Commission’s use of the same term in the bank robbery Guidelines one year later. Those developments shape the pages upon which we write today. So too, for that matter, do prior decisions from our Court. See Little v. BP Expl. & Oil Co., 265 F.3d 357, 362 (6th Cir. 2001) (“[W]e are bound by Sixth Circuit precedent unless it is overruled by either our court sitting en banc or the Supreme Court.”). With only cursory references to the commentary, we have previously held that the text of § 2B3.1(b)(2)(E), through its use of the term “dangerous weapon,” necessarily covered the use of simulated weapons in robbery. See, e.g., Rodriguez, 301 F.3d at 668 (referencing the clarifying Guideline commentary only with a “see also” signal); Woodard, 24 F.3d at 874 (relying on United States v. Medved, 905 F.2d 935 , 939–40 (6th Cir. 1990), a case interpreting 18 U.S.C. § 2113 (d) in light of McLaughlin to hold that a toy gun could be viewed as a dangerous weapon). In other words, binding precedent from two sources, both the Supreme Court as well as this circuit, channels our application of § 2B3.1(b)(2)(E). No. 20-5071 United States v. Tate Page 11 To our mind, that largely sums up the differences between today’s two opinions. Rather than beginning with McLaughlin’s functional view or our own decisions adhering to McLaughlin, the concurring opinion measures the meaning of the term “dangerous weapon” seemingly as a matter of first impression. Relying on dictionary definitions and an extensive survey of state robbery law (some of which postdates McLaughlin and the Guidelines), the concurring opinion rejects a broad reading of dangerous weapon, instead viewing the term as applying (at most) only to items that are “used in such a way as is likely to produce death or grievous bodily harm.” But see United States v. Reed, 94 F.3d 341, 344 (7th Cir. 1996) (observing that “the Guidelines must be interpreted in accordance with federal law,” even when state law may be relevant as a background principle). Only near its closing passage does the concurrence consider McLaughlin. And when it does, the opinion concedes that McLaughlin controls the Guidelines’ definition of dangerous weapon. As McLaughlin’s construction does not comport with the concurrence’s broader view that non-inherently dangerous objects only become dangerous when the object itself can inflict harm, however, the concurring opinion would limit McLaughlin to its precise facts and would ignore its “independently sufficient” reasons for its holding. That approach, to our minds, is at odds with the understanding that when a judicial interpretation has settled the meaning of an existing legal provision, “repetition of the same language” in a subsequent law incorporates the judicial interpretation—and not just the narrow facts from which that interpretation was born. See Merrill Lynch, Pierce, Fenner & Smith Inc. v. Dabit, 547 U.S. 71 , 85–86 (2006) (quoting Bragdon v. Abbott, 524 U.S. 624, 645 (1998)) (rejecting a narrow construction of a statutory term when the Supreme Court had adopted a broad construction of the identical term in another statute). That view likewise overlooks the fact that, for decades, we and other circuits have concluded that the fear caused by possessing an unloaded firearm is materially indistinguishable from that caused by attempting to simulate possession of a firearm during a robbery. While § 2B3.1(b)(2)(E) thus may result in a sentence enhancement for, as the concurring opinion describes it, “fear-inducing behavior” unassociated with possession of an actual weapon, that is simply the product of the Guidelines’ reliance on McLaughlin. A result, it bears noting, that may likewise be the same under the concurring opinion’s “ordinary English speaker” standard, which relies upon dictionary and criminal treatise definitions of “dangerous weapon.” After all, a bank robber’s use of a hand and shoulder No. 20-5071 United States v. Tate Page 12 bag risks “produc[ing] death or grievous bodily harm,” see 3 Ronald Anderson, Wharton’s Criminal Law and Procedure § 961, at 113 (1957), given the dangers inherent to a perceived armed robbery. See McLaughlin, 476 U.S. at 17–18. We offer no view whether, as a policy matter, Tate’s crime should be punished more harshly than the Sentencing Commission (following McLaughlin’s lead) has resolved in § 2B3.1(b)(2)(E). Nor, it bears repeating, does our analysis of § 2B3.1(b)(2)(E) turn on subsequent commentary added by the Commission. Instead, we rest on the text of the Guidelines alone, as informed by McLaughlin and the uniform precedent holding that a dangerous weapon extends to the employment of a simulated weapon. 2. Having resolved that § 2B3.1(b)(2)(E) covers the robber who creates the impression of having an inherently dangerous weapon, Tate argues that his action of placing his hand in his shoulder bag was insufficient to create such an impression. To Tate’s mind, using one’s hand to suggest the “possession” of a separate weapon is outside the ambit of the enhancement. See U.S.S.G. § 2B3.1(b)(2)(E). As there is no dispute that Tate fairly preserved this issue below, we review de novo whether the district court correctly applied the Guidelines in light of the undisputed facts. See United States v. Kimble, 305 F.3d 480, 485 (6th Cir. 2002). Whether a robber should be subject to the dangerous weapons enhancement is an “objective” inquiry under the McLaughlin standard. Rodriguez, 301 F.3d at 668 . It turns on whether a “reasonable individual” would believe that the robber “brandished or possessed” a dangerous weapon, U.S.S.G. § 2B3.1(b)(2)(E), that is, a weapon “capable of inflicting serious bodily injury under the circumstances.” See United States v. Tolbert, 668 F.3d 798, 801 (6th Cir. 2012) (citations omitted). Employing this reasonable-belief standard, we have applied the enhancement to robberies involving non-inherently dangerous objects disguised to be more sinister items, including a toy gun, see Woodard, 24 F.3d at 873 , and a small Styrofoam box, see Rodriguez, 301 F.3d at 669 . And while we have not yet been faced with the facts presented by Tate’s robbery, other circuits have uniformly applied the dangerous weapons enhancement to a robber using his concealed hand to suggest the existence of a lethal instrument. See, e.g., United States v. Smith, 617 F. App’x 941, 942 (11th Cir. 2015) (per curiam) (defendant implied he had a firearm by nudging his right hand inside a jacket pocket during the robbery); United States v. No. 20-5071 United States v. Tate Page 13 Hoffa, 587 F.3d 610 , 615–16 (3d Cir. 2009) (defendant threatened bank teller while holding his hand in his pocket); United States v. Stitman, 472 F.3d 983 , 986–87 (7th Cir. 2007) (similar); United States v. Farrow, 277 F.3d 1260, 1268 (10th Cir. 2002) (similar); United States v. Souther, 221 F.3d 626, 629 (4th Cir. 2000) (similar). We now join this uniform line of cases treating a robber that uses his concealed hand to reasonably suggest the existence of a weapon as having committed an act sufficient to satisfy § 2B3.1(b)(2)(E). Clad in a dark disguise, Tate entered a bank and delivered a note to a teller giving the teller a thirty second deadline to provide Tate $20,000 at the risk of not letting “[e]veryone” in the bank live. Tate then immediately followed up on his threat by audibly counting down. And as the count neared the end, Tate thrust his hand into the opaque shoulder bag covering his body, seemingly acting on his threat of not letting “[e]veryone” survive. Through his latter action, Tate “create[d] an appearance” that he possessed a dangerous weapon to effectuate the crime. Dixon, 982 F.2d at 119 . Through Tate’s menacing words coupled with using his hand and his shoulder bag, a “reasonable individual” would believe that Tate had a weapon capable of inflicting serious bodily injury. Tolbert, 668 F.3d at 801 . And, it bears adding, while the subjective state of mind of the victim alone is insufficient to determine whether it was reasonable to believe that Tate had a dangerous weapon, see Woodard, 24 F.3d at 874 , the teller here, we note, responded to Tate’s actions by turning over the money, suggesting that he believed Tate’s threat to be credible. See United States v. Hall, 763 F. App’x 722 , 725–26 (10th Cir. 2019) (observing that the teller’s impressions can inform whether it was objectively reasonable to believe that the bank robber possessed a dangerous weapon); Smith, 617 F. App’x at 942 (similar). Resisting this mountain of authority, Tate turns to a single case from the Second Circuit, United States v. Taylor, 961 F.3d 68 (2d Cir. 2020). There, the Second Circuit considered whether to apply the dangerous weapons enhancement to a defendant who, during one robbery, gestured with his hand toward his waistband, and, during another, held his belt. Id. at 71–72. Quoting the Guidelines’ commentary, the Second Circuit held that these acts did not amount to using an “object in a manner that created the impression that the object was an instrument No. 20-5071 United States v. Tate Page 14 capable of inflicting death or serious bodily injury.” Id. at 75 (emphasis in original) (quoting U.S.S.G. § 2B3.1 cmt. n.2(B)). Taylor is a poor guide here. Start with the facts. Taylor involved the use of an unconcealed hand during a robbery, meaning there was “no indication” that the robber used his hand to “create the impression” that he had a dangerous weapon. Id. at 75–76 (observing that while there was a “possibility” the robber appeared to have a weapon, the record was “far too general to support application of the enhancement”). Tate, in contrast, used a verbal threat and employed his concealed hand and his shoulder bag to create the objectively credible fear that he possessed an inherently dangerous weapon. And then consider the law. Taylor did not cite McLaughlin or purport to employ its functional test. Instead, the Second Circuit questioned whether the facts were consistent with the plain language of the Guideline’s commentary, seemingly requiring unity between (1) the object being used to create the impression that there is a dangerous weapon (in Taylor, the defendant’s hand) and (2) the imagined weapon itself. Id. at 76–77. Yet as Tate himself acknowledges, the Guidelines’ commentary has “no independent legal force.” Havis, 927 F.3d at 386 . What is binding is the Guidelines’ text, which imposes the enhancement for brandishing or possessing a “dangerous weapon,” U.S.S.G. § 2B3.1(b)(2)(E), a term that extends to items that reasonably appear to be dangerous. See McLaughlin, 476 U.S. at 17–18. By employing the phrase “dangerous weapon,” the text of § 2B3.1(b)(2)(E) suggests that a robber who uses one instrument to imply credibly the existence of a separate dangerous instrument could be viewed as having brandished or possessed such a weapon. Perhaps merely placing an unconcealed hand on a belt during a robbery, without more, does not justify applying the enhancement. See Taylor, 961 F.3d at 75–76. But that conclusion follows not from an additional rule created by the Guidelines’ commentary, but because a reasonable individual ordinarily would not understand a person touching his belt to suggest the possession of a dangerous weapon. Finally, Tate argues that imposing the enhancement based on his gesture with his concealed hand would eliminate the need for the final enhancement in § 2B3.1(b)(2)’s hierarchy. To that end, the Guidelines’ tiered structure imposes a two-level enhancement for making a “threat of death” during robbery. See U.S.S.G. § 2B3.1(b)(2)(F). Because § 2B3.1(b)(2)’s No. 20-5071 United States v. Tate Page 15 commentary instructs that a threat can occur through a “gesture,” see id. § 2B3.1 cmt. n.6, Tate contends that applying the enhancement to his specific act would make the threat-of-death enhancement superfluous. But here again, this reading elevates the Guidelines’ commentary beyond the plain text of the threat-of-death enhancement. It likewise erroneously assumes that, for purposes of the enhancement, all physical gestures are equally suggestive. Whether the dangerous weapons enhancement applies depends on whether a reasonable individual would believe that the defendant brandished or possessed such a weapon. And not all gestures do. Moving one’s hand in a towel as if you have a gun is a gesture that creates a threat of death and, critically, reasonably implies the existence of a firearm, see id. § 2B3.1 cmt. n.2, allowing for the application of the more significant “dangerous weapons” enhancement. By comparison, drawing a bare, empty hand across your throat in a slashing motion is a gesture that threatens death, see id. § 2B3.1 cmt. n.6, but not one that implies the existence of a dangerous weapon, thereby leaving the defendant eligible for only the lesser “threat of death” enhancement. There is ample room, in other words, for these enhancements harmoniously to coexist. CONCLUSION All told, under the circumstances presented here, the district court did not err in increasing Tate’s sentence by three levels. Accordingly, we affirm the judgment of the district court. No. 20-5071 United States v. Tate Page 16 ______________________________________ CONCURRING IN THE JUDGMENT ______________________________________ MURPHY, Circuit Judge, concurring in the judgment. Did Tre Tate “possess” a “dangerous weapon” when he stuck his hand into a bag to mislead a bank teller into believing that he had a gun? U.S.S.G. § 2B3.1(b)(2)(E). I do not think so. No ordinary English speaker— even one who has spent years reading legal decisions—would say that a robber possesses a dangerous weapon when the robber merely pretends to possess one. Yet the definition of “dangerous weapon” in the Sentencing Commission’s commentary makes this atextual leap by “equat[ing] the image of a ‘dangerous weapon’ with its reality[.]” United States v. Dixon, 982 F.2d 116, 121 (3d Cir. 1992). The commentary thus unlawfully expands the scope of the dangerous-weapon enhancement beyond the text of the guideline itself. See United States v. Riccardi, 989 F.3d 476, 486 (6th Cir. 2021). That said, Tate’s challenge to the commentary in this court is not the same challenge that he raised in the district court. So I would review his claim only for plain error. And the majority’s thoughtful opinion shows that no error was “plain.” I thus concur in the judgment. I The robbery guideline imposes a three-level enhancement to an offense level if a “dangerous weapon was brandished or possessed” during a robbery. U.S.S.G. § 2B3.1(b)(2)(E). The commentary defines the phrase “dangerous weapon” to cover not just an instrument capable of inflicting injury, but also an object used to create a false impression that it is such an instrument: “Dangerous weapon” means (i) an instrument capable of inflicting death or serious bodily injury; or (ii) an object that is not an instrument capable of inflicting death or serious bodily injury but (I) closely resembles such an instrument; or (II) the defendant used the object in a manner that created the impression that the object was such an instrument (e.g. a defendant wrapped a hand in a towel during a bank robbery to create the appearance of a gun). Id. § 1B1.1 cmt. n.1(E); id. § 2B3.1 cmt. n.2. Tate offers two reasons why the district court wrongly held that he possessed a dangerous weapon. He argues (1) that his conduct did not No. 20-5071 United States v. Tate Page 17 satisfy the commentary’s broad definition of the phrase “dangerous weapon” and (2) that, even if it did, this definition impermissibly enlarges the guideline’s scope. I agree with my colleagues’ rejection of the first claim. Tate contends that the second part of the commentary’s “dangerous weapon” definition requires that a defendant use an “object” to create the impression that the object itself is the weapon. Id. § 1B1.1 cmt. n.1(E)(ii)(II). It is not enough, Tate says, for a defendant to use an object in a way that suggests that the defendant has a weapon elsewhere. See United States v. Taylor, 961 F.3d 68 , 75–77 (2d Cir. 2020). Even under this reading, however, Tate’s hand could qualify as the relevant “object”—as the commentary’s hand-in-a-towel example shows. Indeed, many decisions have read the commentary to cover defendants who conceal their hand in a pocket or a bag to create the impression that they have a gun. See, e.g., United States v. Davis, 635 F.3d 1222 , 1224–25 (D.C. Cir. 2011); United States v. Stitman, 472 F.3d 983, 987 (7th Cir. 2007). In the case on which Tate relies, by contrast, the defendant’s unconcealed hand could not have been mistaken for a gun. Taylor, 961 F.3d at 75 . That leaves Tate’s attack on the commentary’s definition of “dangerous weapon.” In my view, we should review this claim for plain error because Tate did not raise it in the district court. See Puckett v. United States, 556 U.S. 129 , 134–35 (2009). To preserve the claim, Tate needed to raise it with the level of “specificity” that would alert the district court he was challenging the commentary. See United States v. Bostic, 371 F.3d 865, 871 (6th Cir. 2004) (citation omitted). Yet his brief below objected to the enhancement only on the commentary’s own terms: He claimed that his conduct fell outside its definition of “dangerous weapon” because he did not create the impression that he had a gun. Objections, R.22, PageID#60–61. Not once did Tate’s brief challenge this definition or cite cases limiting the commentary’s scope. See Stinson v. United States, 508 U.S. 36, 46 (1993); United States v. Havis, 927 F.3d 382, 386 (6th Cir. 2019) (en banc) (per curiam); see also Kisor v. Wilkie, 139 S. Ct. 2400 , 2414–18 (2019). At sentencing, Tate again argued only that he did not create the impression that he had a gun and again failed to challenge the commentary. Tr., R.35, PageID#135–137. The district court rejected Tate’s argument because “almost anything can count as a dangerous weapon” under the commentary. Id., PageID#139–45. The court did not address the commentary’s No. 20-5071 United States v. Tate Page 18 validity, confirming that it did not view Tate as raising such a challenge. Lastly, when asked about any other objections at the end of sentencing, Tate stayed silent about the commentary’s validity. Id., PageID#156; see Bostic, 371 F.3d at 872–73. This record triggers plain-error review. We cannot expect the district court to have addressed—on its own initiative—Tate’s claim that the commentary’s “dangerous weapon” definition is invalid. An analogy proves the point. Suppose a defendant argues in the district court that certain conduct does not fall within a statute. Would that statutory argument preserve the defendant’s contention that the statute is unconstitutional as applied to the conduct? No, those statutory and constitutional theories are separate “claims” under our preservation rules (not separate arguments in support of a single claim). United States v. Dedman, 527 F.3d 577, 591 (6th Cir. 2008); see United States v. Reed, 993 F.3d 441, 453 (6th Cir. 2021). That logic covers this case. A world of difference exists between the claim that the commentary’s text does not apply to certain conduct and the claim that this text is unlawfully overbroad. A contrary rule would permit sentencing “sandbagging” wasteful of judicial resources. See Puckett, 556 U.S. at 134 . Busy district courts should be able to rely on the commentary unless and until a party challenges it. If Tate had raised his claim in the district court, the court also could have given reasons why this issue does not matter. See id. It, for example, might have said that it would have imposed the same sentence under the advisory guidelines even without the dangerous-weapon enhancement. See Molina-Martinez v. United States, 136 S. Ct. 1338, 1346 (2016). Tate’s appellate briefing all but concedes that plain-error review applies. The government invoked this standard in response to Tate’s challenge to the commentary. Tate’s reply offered no counterargument. It merely claimed that he should win “even under plain error review.” Reply Br. 1. Tate’s appellate silence on the standard of review reads as an implicit concession that he did not preserve the issue. Cf. Croce v. N.Y. Times Co., 930 F.3d 787, 793 (6th Cir. 2019). The plain-error standard makes this case easy. The standard requires, among other things, that an error be “clear or obvious, rather than subject to reasonable dispute.” Puckett, No. 20-5071 United States v. Tate Page 19 556 U.S. at 135 . And the district court did not commit an error that was “clear under current law.” United States v. Olano, 507 U.S. 725, 734 (1993). The majority’s well-reasoned opinion shows as much. Before this case, moreover, we had applied § 2B3.1’s commentary without examining its validity. See United States v. Rodriguez, 301 F.3d 666, 668 (6th Cir. 2002); United States v. Woodard, 24 F.3d 872 , 873–74 (6th Cir. 1994). Other courts had likewise applied the commentary in similar circumstances. See, e.g., Stitman, 472 F.3d at 987 ; Dixon, 982 F.2d at 121–23. This state of affairs shows plenty of room for debate over Tate’s challenge. Cf. United States v. Al-Maliki, 787 F.3d 784, 794 (6th Cir. 2015). I thus would affirm Tate’s sentence enhancement on the ground that no error was “plain.” But I would leave it at that. II Because my colleagues reject Tate’s challenge to the commentary on the merits, I offer my thoughts on that subject too. The rules for Tate’s challenge are now clear in this circuit. If a phrase in a guideline is ambiguous, we defer to the commentary’s clarification of the ambiguity. See Stinson, 508 U.S. at 45 . To receive deference, however, the commentary’s reading of a guideline must “fall ‘within the zone of ambiguity’ that exists.” Riccardi, 989 F.3d at 480 (quoting Kisor, 139 S. Ct. at 2416 ). The commentary cannot expand the guideline under the guise of “interpreting” it. See Havis, 927 F.3d at 386–87. This case thus turns on a straightforward question: Can the phrase “possessed” a “dangerous weapon” in § 2B3.1(b)(2)(E) be reasonably read to cover a robber who only pretends to have such a weapon by concealing his hand in a bag? See id. §§ 1B1.1 cmt. n.1(E), 2B3.1 cmt. n.2. Both the plain meaning of “dangerous weapon” and the legal backdrop against which the Commission first used this phrase show that we should take this commentary for what it is: an improper enlargement of the guideline’s scope. A How would an ordinary English speaker understand what it means to “possess” a “dangerous weapon”? See U.S.S.G. § 2B3.1(b)(2)(E). Dictionaries from when the Commission adopted § 2B3.1 provide an initial data point. The noun “weapon” typically meant “any instrument or device for use in attack or defense in combat, fighting, or war, as a sword, rifle, or No. 20-5071 United States v. Tate Page 20 cannon.” Random House Unabridged Dictionary 2153 (2d ed. 1993); see also American Heritage Dictionary of the English Language 2022 (3d ed. 1992). The adjective “dangerous” added one limit on the covered instruments: They must be “able or likely to do harm.” American Heritage Dictionary, supra, at 472. The verb “possessed” added another: The defendant must have “actual and physical control” of the weapon. Black’s Law Dictionary 1162 (6th ed. 1990). (The government does not argue that Tate “brandished” a dangerous weapon, the other type of conduct covered by § 2B3.1(b)(2)(E).) But, as my colleagues point out, perhaps we should not ask how the average person would understand the words “weapon” and “dangerous” in the abstract. Ordinary meaning often gives way when the context suggests that a phrase has a technical or legal meaning. See Neder v. United States, 527 U.S. 1 , 21–22 (1999). A tomato might be a “vegetable” in a statute directed to a lay audience but a “fruit” in one directed to a scientific audience. See Nixon v. Hedden, 149 U.S. 304, 307 (1893). In this criminal setting, too, many words (like “fraud” or “conspiracy”) have developed unique legal meanings over time. Neder, 527 U.S. at 22 ; United States v. Wheat, 988 F.3d 299, 307 (6th Cir. 2021). “Dangerous weapon” is a similar phrase. It has long been used in the criminal law to increase the punishment for “armed” robberies or assaults. See, e.g., 2 Wayne R. LaFave & Austin W. Scott, Substantive Criminal Law § 8.11(f), at 455–57 (1986). Yet I, for one, do not see much daylight between the phrase’s legal meaning and its ordinary meaning. Legal sources defined the phrase “dangerous weapon” as “any article which, in circumstances in which it is used, attempted to be used, or threatened to be used, is readily capable of causing death or other serious injury.” Black’s Law Dictionary, supra, at 394. Whether assessed based on its ordinary or legal meaning, the phrase can cover two types of articles. It obviously covers instruments like guns or daggers that are “typically” or “inherently” dangerous weapons—that is, those that are “designed or constructed for offensive or defensive purposes or for the destruction of life or the infliction of bodily injury.” 3 Ronald Anderson, Wharton’s Criminal Law and Procedure § 961, at 113 (1957); see McLaughlin v. United States, 476 U.S. 16, 17 (1986); United States v. Wallace, 800 F.2d 1509 , 1512–13 (9th Cir. 1986). Yet the phrase need not be limited to things that we typically call “weapons.” “Weapon” can reach “any” object used or threatened to be used for the purpose of inflicting No. 20-5071 United States v. Tate Page 21 bodily harm. Random House Unabridged Dictionary, supra, at 2153; Black’s Law Dictionary, supra, at 394. It is perfectly natural to say that an aggressor used a chair or a cane “as a weapon” if the aggressor attacked or threatened someone with it. See United States v. Loman, 551 F.2d 164, 169 (7th Cir. 1977); United States v. Johnson, 324 F.3d 264, 266 (4th Cir. 1963). When the Commission first used the phrase, therefore, ordinary items regularly qualified as “dangerous weapons” if the items were “used in such a way as [was] likely to produce death or grievous bodily harm.” Anderson, supra, at 113; Johnson, 324 F.3d at 266 (listing examples). Even if we read “dangerous weapon” in this broad way, though, I still do not see how Tate “possessed” a “dangerous weapon” when he stuck his hand into a bag to mislead a bank teller into thinking he had a gun. As most courts have long held, a gun (loaded or unloaded) can certainly be described as a “dangerous weapon.” See McLaughlin, 476 U.S. at 17–18; LaFave, supra, § 8.11(f), at 456. But the guideline required Tate to “possess” the gun. He did not. What, then, could qualify as Tate’s “weapon” (even setting aside the word “dangerous”)? His hand? I do not think so. For starters, the word “weapon” typically connotes an item distinct from the person; it would be a strange usage to refer to one’s hand as an “instrument or device.” Random House Unabridged Dictionary, supra, at 2153. (While those skilled in martial arts (and not lacking in self-confidence) might describe their fists as “deadly weapons,” that is because the English language includes “a figure of speech” known as a metaphor, not because body parts are commonly thought of as weapons. MCI Telecomms. Corp. v. Am. Tel. & Tel. Co., 512 U.S. 218, 228 (1994).) I thus tend to agree with the majority of courts that refuse to treat body parts as “dangerous weapons.” See United States v. Rocha, 598 F.3d 1144 , 1154–57 (9th Cir. 2010) (citing cases). That is particularly true for this guideline, which prohibits a robber from “possessing” a dangerous weapon. See U.S.S.G. § 2B3.1(b)(2)(E). It is even more unnatural to say that Tate “possessed” his hand. Regardless, Tate’s hand (unlike a gun) is at least not an “inherently” dangerous weapon. Wallace, 800 F.2d at 1512–13. An object that is not traditionally viewed as a “weapon” might qualify as one only if the defendant used or threatened to use it in a manner that was capable of inflicting serious bodily injury. See Loman, 551 F.2d at 169 . For example, even though a pen might perhaps qualify as a “weapon” if a robber threatens to stab a bank teller with it, see, e.g., No. 20-5071 United States v. Tate Page 22 State v. Barragan, 9 P.3d 942 , 946–47 (Wash. Ct. App. 2000), nobody would say that the robber used the pen “as a weapon” if the robber wrote a note demanding money on threat of violence. In this case, too, Tate did not use his hand in the required assaultive manner. He did not hit a bank teller or threaten to do so. And his finger was not actually capable of discharging lethal bullets. Tate’s shoulder bag likewise could not qualify as a weapon for the same reason. The bag itself was not used or threatened to be used in a way capable of causing serious physical harm (perhaps, say, by suffocating the teller with it). I thus would hold that Tate’s conduct fell outside any plausible interpretation of “possess[ing]” a “dangerous weapon” in § 2B3.1(b)(2)(E). B All this said, I understand the impetus behind imposing harsher punishment on Tate’s conduct. Robbers who imply that they have a weapon will enhance their victims’ fear and increase the risk of a violent encounter (from an armed guard, for example). See Stitman, 472 F.3d at 988 . Perhaps there are “good reasons” why the conduct should be subject to increased punishment. United States v. Shores, 966 F.2d 1383, 1387 (11th Cir. 1992) (per curiam). But the Commission should have achieved that objective by changing the guideline’s text. It is not difficult to think of ways the Commission could have done so. Many state legislatures have expanded their robbery statutes to reach conduct like Tate’s. Some increase the sentence not just when a robber has a dangerous weapon but also when the robber has “any article used or fashioned in a manner to lead the victim reasonably to believe that it is a dangerous weapon[.]” Wis. Stat. Ann. § 943.32 (2); see Colo. Rev. Stat. Ann. § 18-4-302 (1)(d); Mich. Comp. Laws Ann. § 750.529 (1)(b); Minn. Stat. Ann. § 609.245 (1); Tenn. Code Ann. § 39 - 13-402(a)(1). Others increase the sentence when the robber uses “a simulated deadly weapon,” Ariz. Rev. Stat. Ann. § 13-1904 (A)(2), “a representation of a deadly weapon,” S.C. Code Ann. 16-11-330(A), or “what appears to be a deadly weapon,” Mo. Rev. Stat. § 570.023 (1)(4); see Del. Code Ann. tit. 11, § 832 (a)(2); Ga. Code Ann. § 16-8-41 (a); N.Y. Penal Law § 160.15 (4); Wash. Rev. Code Ann. § 9A.56.200(1)(ii). Still others increase the sentence when the robber “represents by words or other conduct that” the robber has a dangerous weapon. Alaska Stat. No. 20-5071 United States v. Tate Page 23 § 11.41.500(a)(1); see Ark. Stat. Ann. § 41-21-2(1)(a); Conn. Gen. Stat. Ann. § 53a-134(a)(4); 720 Ill. Comp. Stat. Ann. 5/18-1(b)(1); Or. Rev. Stat. Ann. § 164.405 (1)(a). These ubiquitous statutes—many of which predate the guidelines, see LaFave, supra, § 8.11(f), at 455 n.101; Brooks v. State, 552 A.2d 872 , 877 n.5 (Md. 1989)—show that the phrase “possess” a “dangerous weapon” did not have any well-established legal meaning that could cover pretending to possess a dangerous weapon. If “dangerous weapon” had such an idiosyncratic meaning, why would the legislatures need to expand their statutes in this manner? The legislatures began doing so only after their state courts refused to expand them through interpretation rather than legislation. See Lynn Considine Cobb, Annotation, Robbery by Means of Toy or Simulated Gun or Pistol, 81 A.L.R.3d 1006 , § 2[a], Westlaw (databased updated 2021). A few examples prove my point. Consider New Jersey. In 1982, the New Jersey Supreme Court held that a robber did not use a dangerous weapon “by placing his hand in his coat pocket and pretending that he was concealing a handgun.” State v. Butler, 445 A.2d 399, 400 , 403–04 (N.J. 1982). The court reasoned that “the New Jersey statute could not be more clear—the actor must actually possess a dangerous weapon[.]” Id. at 403 . The legislature thus amended the statute to cover the “simulated possession of a deadly weapon.” State v. Chapland, 901 A.2d 351 , 357 n.5 (N.J. 2006). Or consider Arizona. In 1981, the Arizona Supreme Court held that its armed-robbery statute “is not satisfied by the defendant pretending to have a gun or even using a fake gun.” State v. Franklin, 635 P.2d 1213, 1214 (Ariz. 1981) (citation omitted). The legislature again amended the statute to reach simulated deadly weapons. State v. Garza Rodriguez, 791 P.2d 633 , 634–38 (Ariz. 1990). These are not isolated cases. See also, e.g., State v. Ireland, 150 P.3d 532, 535 (Utah 2006); State v. Hopson, 362 N.W.2d 166 , 168–69 (Wis. Ct. App. 1984); State v. Smith, 450 So. 2d 714 , 716 & n.1 (La. Ct. App. 1984). In states that have not expanded their statutes, by contrast, courts continue to require a robber to “actually possess the weapon at the time of the crime.” Gray v. State, 903 N.E.2d 940, 944 (Ind. 2009); see State v. Marshall, 656 S.E.2d 709, 714 (N.C. Ct. App. 2008); Commonwealth v. Howard, 436 N.E.2d 1211 , 1211–13 (Mass. 1982). This background law confirms that the Commission’s commentary expands—rather than interprets—the robbery guideline. Section 2B3.1(b)(2)(E) contains nothing like the broad No. 20-5071 United States v. Tate Page 24 language found in the state statutes. It reaches only those who possessed a dangerous weapon, not those who possessed a simulated weapon or represented that they had one. Contrary to the commentary’s broad reading of “dangerous weapon,” the state cases also show that the phrase lacks the unusual meaning required to cover conduct like Tate’s. Indeed, I could not find any pre-guidelines cases holding that a robber possessed a gun under a (non-expanded) armedrobbery statute when the robber only pretended to possess one by sticking a hand in a pocket or bag. Courts instead noted that “the existence of numerous statutes . . . explicitly addressing the use of an apparent (as opposed to an actual) dangerous or deadly weapon strongly suggests that legislatures use carefully crafted language when they intend to take that route.” Brooks, 552 A.2d at 879 . The Commission unambiguously did not take that route in the robbery guideline itself. And I think that linguistic “deference to the English language” in § 2B3.1(b)(2)(E) must trump administrative deference to the Commission’s expansion of that language in its commentary. Wilburn v. Commonwealth, 312 S.W.3d 321, 327 (Ky. 2010). C The pre-guidelines cases on which my colleagues rely do not convince me that the phrase “possessed” a “dangerous weapon” could reach a robber who pretended to possess a gun by concealing a hand in a bag. Start with the cases addressing similar conduct. Some state courts indicated that a robber’s threat of a concealed gun was sufficient evidence for a rational jury to infer that the robber possessed an actual gun. See Commonwealth ex rel. Johnson v. Myers, 189 A.2d 331, 332 (Pa. Super. Ct. 1963) (per curiam); see also, e.g., State v. Elam, 312 So. 2d 318, 322 (La. 1975); State v. Sherman, 335 N.E.2d 753 , 755–56 (Ohio 1973). And some state statutes similarly indicated that the prosecution could establish a “prima facie” case that the defendant was actually armed by using evidence of “any verbal or other representation by the defendant that he is then and there so armed[.]” Ala. Stat. § 13A-8-41(b) (1977); Stewart v. State, 443 So. 2d 1362, 1364 (Ala. Crim. App. 1983); James v. State, 405 So. 2d 71, 73 (Ala. Crim. App. 1981). But these sufficiency-of-the-evidence rules (whether statutory or judge-made) do not help the government here. It did not attempt to prove that Tate had a real gun. And the state decisions did not reach the more remarkable conclusion that a concealed hand itself qualified as the dangerous weapon (at least not under a statute without an expanded reach). Elam, for No. 20-5071 United States v. Tate Page 25 example, disclaimed such a reading: “nor do we hold that a ‘hand in a pocket’ is a dangerous weapon.” 312 So. 2d at 322 n.*. That leaves McLaughlin—the Supreme Court case on which my colleagues primarily rely. But that case was not about pretending to possess a weapon. Lamont McLaughlin actually had a gun, albeit an unloaded one. The Court held that his unloaded gun qualified as a “dangerous weapon” under the federal bank-robbery statute. 476 U.S. at 17–18; 18 U.S.C. § 2113 (d). I would not expand this holding to cover Tate’s conduct. Even when unloaded, guns (unlike hands) are still commonly called “weapons.” So the case turned on the statutory adjective: Could a gun be considered “dangerous” even when unloaded? The Court gave three “independently sufficient” reasons for answering “yes.” See McLaughlin, 476 U.S. at 17 . The first and third reasons considered the question from the perspective of the “class of weapons” to which McLaughlin’s gun belonged. Wilburn, 312 S.W.3d at 328 . The Court held that “a gun is an article that is typically and characteristically dangerous; the use for which it is manufactured and sold is a dangerous one, and the law reasonably may presume that such an article is always dangerous even though it may not be armed at a particular time or place.” McLaughlin, 476 U.S. at 17 . It added that a gun can be used as a bludgeon. Id. at 18 . Neither reason covers Tate’s hand in a bag. The second reason requires more discussion. The Court stated that “the display of a gun instills fear in the average citizen; as a consequence, it creates an immediate danger that a violent response will ensue.” Id. at 17–18. Does this statement give us license to treat as a “dangerous weapon” anything that instills fear or risks violence? Any statement that a robber has a gun? Any threat of violence? I do not think so. That interpretation would elevate the statute’s implied purpose (to discourage fear-inducing conduct) over its enacted text (“use of a dangerous weapon or device”). But, as the Supreme Court has reminded many times since McLaughlin, “no legislation pursues its purposes at all costs.” Rodriguez v. United States, 480 U.S. 522 , 525–26 (1987) (per curiam). Legislation is instead “the art of compromise[.]” Henson v. Santander Consumer USA Inc., 137 S. Ct. 1718, 1725 (2017). A reasonable legislature might think it best to enhance the punishment only for a robber who uses a weapon (not one who pretends to) because a statute that treats both defendants alike could increase the risk that a robber brings a No. 20-5071 United States v. Tate Page 26 real gun to a robbery. Cf. LaFave, supra, § 8.11(f), at 456 n.108. Rather than read McLaughlin as pulling within the statute any and all fear-inducing behavior, I would read it to retain the requirement of a “dangerous weapon.” See United States v. Dixon, 790 F.3d 758, 761 (7th Cir. 2015) (Easterbrook, J.). And although a concealed hand with a threat “may lead to fear,” it is not a weapon, let alone a dangerous one. See id. Two other factors support my instinct that we should not interpret McLaughlin as adopting the purpose-over-text approach necessary to cover Tate’s conduct. To begin with, even before McLaughlin, the “great weight of authority” had already held that unloaded guns were dangerous weapons. LaFave, supra, § 8.11(f), at 456. As noted, however, these same courts refused to treat a concealed hand as a “dangerous weapon” (some even called that conclusion “absurd”). Brooks, 552 A.2d at 879–880; see People v. Skelton, 414 N.E.2d 455 , 456–57 (Ill. 1980); Cooper v. State, 297 S.W.2d 75 , 76–78 (Tenn. 1956). These different outcomes show that an unloaded gun can be described as a “dangerous weapon” in a way that a concealed hand cannot be. Even in the context of the bank-robbery statute itself, courts have not read McLaughlin as broadly as the government needs here. See 18 U.S.C. § 2113 (d). Admittedly, several courts, including our own, have used McLaughlin to hold that the display of a toy gun can qualify as a “dangerous weapon” because it may induce fear. See United States v. Medved, 905 F.2d 935 , 939–40 (6th Cir. 1990); United States v. Martinez-Jimenez, 864 F.2d 664 , 666–67 (9th Cir. 1989); cf. United States v. Perry, 991 F.2d 304, 308 (6th Cir. 1993). Count me as “skeptical” of these decisions. Dixon, 790 F.3d at 761 . But this precedent is still a significant step removed from a concealed empty hand plus a threat of a gun. I am not aware of any case reading § 2113(d)’s enhancement to reach that sort of deception; courts have instead suggested that it would not do so. See United States v. Wolfe, 245 F.3d 257, 259 , 261–62 (3d Cir. 2001); United States v. Ray, 21 F.3d 1134 , 1135–40 (D.C. Cir. 1994). At day’s end, neither McLaughlin nor any other precedent can justify the atextual conclusion that a concealed hand is a “dangerous weapon.” No. 20-5071 United States v. Tate Page 27 D The commentary’s history cements my view. It shows that even the Commission did not believe that the guideline’s text could be read in the broad way that my colleagues interpret it— at least not without an amendment. The original definition of “dangerous weapon” covered only “an instrument capable of inflicting death or serious bodily injury”; it did not cover an object used to create the impression that one had such a weapon. U.S.S.G. § 1B1.1 cmt. n.1(d) (1987). Cases decided shortly after this definition recognized that the Commission had “elected not to authorize increased punishment for claiming to be armed.” United States v. Coe, 891 F.2d 405, 411 (2d Cir. 1989); see United States v. Hawkins, 901 F.2d 863, 865 (10th Cir. 1990). And the Commission unofficially suggested that a “toy gun [did] not meet the requirements of a . . . dangerous weapon” and thus did not trigger the dangerous-weapon enhancement. See U.S. Sentencing Commission Answers Questions Most Frequently Asked About the Sentencing Guidelines, 1 Fed. Sent. R. 423 , 423, 425–26 (Apr. 1989) (Question 36). We, too, acknowledged that “fake weapons had not been taken into consideration adequately” under this initial set of guidelines. Medved, 905 F.2d at 941 . But the Commission quickly had a change of heart. In November 1989, it added the following language to the commentary’s dangerous-weapon definition: “Where an object that appeared to be a dangerous weapon was brandished, displayed, or possessed, treat the object as a dangerous weapon.” U.S.S.G. App. C, amend. 71 (effective Nov. 1, 1989); U.S.S.G. App. C Supp., amend. 110 (effective Nov. 1, 1989) (applying to § 2B3.1). Courts relied on this “expansion on the meaning” of dangerous weapon to cover such things as a hand in a towel, Dixon, 982 F.2d at 121–24, a toy gun, Shores, 966 F.2d at 1387–88, or a shoebox claimed to be a bomb, United States v. Hart, 226 F.3d 602, 605 (7th Cir. 2000). As these courts reasoned, it was the commentary that “equate[d] appearance with reality,” not the guideline. Dixon, 982 F.2d at 122–23. And as we noted, “if the use of toy weapons had been taken into account by the Commission from the beginning, there would have been no need for the amendment.” Medved, 905 F.2d at 943 . In 2000, the Commission added its present definition to the commentary to incorporate the holdings of Dixon and Shores. See U.S.S.G. App. C, Vol. II, amend. 601 (effective Nov. 1, No. 20-5071 United States v. Tate Page 28 2000). When doing so, the Commission effectively acknowledged that the commentary defined dangerous weapon to mean not a dangerous weapon: “The definition of ‘dangerous weapon’ in Application Note 1(d) of §1B1.1 also is amended to clarify under what circumstances an object that is not an actual, dangerous weapon should be treated as one for purposes of guideline application.” Id. (emphasis added). As our recent cases teach, however, this expansion to the guideline belonged in the guideline, not in its commentary. See Riccardi, 989 F.3d at 488 . We have held, for example, that the commentary impermissibly expanded a guideline that increased the punishment for completed crimes by interpreting it to cover attempt crimes. Havis, 927 F.3d at 386–87. The same logic applies here. This commentary impermissibly expands a guideline increasing the punishment for possessing a dangerous weapon by interpreting it to cover pretending to have such a weapon. And the few decisions that have actually addressed the validity of the commentary engaged in the type of “reflective deference” to the commentary that we have now jettisoned. Riccardi, 989 F.3d at 485 (citation omitted); see Stitman, 472 F.3d at 987 ; Dixon, 982 F.2d at 121–22. For these reasons, I respectfully concur in the judgment.