HARTZ, Circuit Judge.
Defendant Aaron Heineman was convicted after a bench trial on one count of sending an interstate threat. See 18 U.S.C. § 875(c). The district court found that he knowingly sent an e-mail that caused the recipient to reasonably fear bodily harm. Defendant argues that his conviction violated the First Amendment because the court did not also find that he intended the recipient to feel threatened. We have jurisdiction under 28 U.S.C. § 1291. Agreeing with Defendant, we reverse and remand.1
I. BACKGROUND
In 2010 and 2011 Defendant sent three e-mails espousing white supremacist ideology to a professor at the University of Utah. The first two e-mails did not contain threats, but the third made the professor fear for his safety and the safety of his [972]*972family. Entitled “Poem,” ApltApp., Vol. 1 at 89, it began by addressing the professor by his first name, and contained the following language:
Come the time of the new revolution we will convene to detain you
And slay you, by a bowie knife shoved up into the skull from your pig chin you choke, with blood flooding in your filthily treasonous throat!
We put the noose ring around your neck and drag you as you choke and gasp
The noose laid on the tree branch and the fate hath conferred justice for Treason
You are a filthy traitor along the horde of anti-American and anti-Whitey comrades
whose justice shall come to be delivered
To fuck the traitors, for justice! fuck Mexico! fuck South America!
Fuck your soul to Hell!
Into the furnace pool of MexiShit as the filthily traitorous asshole and puta!
Id. at 90. Law-enforcement officers traced the e-mail to Defendant through his e-mail address, which had the user name “siegheiLneocon.” Id. at 91. When officers contacted him in writing, he responded immediately, “Is this about the email?” Id. He was charged in the United States District Court for the District of Utah with one count of sending an interstate threat, in violation of 18 U.S.C. § 875(c).
Before trial Defendant requested an instruction that “the government must prove that the defendant intended the communication to be received as a threat.” Id. at 18. He asserted that he has Asperger’s Disorder, which impairs his “ability to understand how others will receive the things he says and does.” Aplt. Br. at 2. The district court declined the request. Defendant then moved to dismiss the charge, arguing that § 875(c) was facially unconstitutional if it did not require proof that “the defendant intended to place the hearer in fear of bodily harm or death.” ApltApp., Vol. 1 at 28. After the court denied the motion, the parties agreed to a bench trial on stipulated facts so that Defendant could preserve his legal arguments. He renewed his objections at trial, and the court again rejected them. It found that the government had established that Defendant “knowingly transmitted a communication containing a threat to injure the person of another,” id. at 91, and that the poem was a true threat because it “would cause a reasonable person to conclude that the sender ... intended to cause bodily injury,” id. at 93. The court did not determine whether Defendant intended the professor to feel threatened.
II. DISCUSSION
Defendant was prosecuted under 18 U.S.C. § 875(c); which states in relevant part: “Whoever transmits in interstate or foreign commerce any communication containing ... any threat to injure the person of another, shall be fined under this title or imprisoned not more than five years, or both.” The law in this circuit is settled, and the parties do not dispute, most of what must be proved to establish a violation of the statute. For example, the statement itself must be one that a reasonable person in the circumstances would understand “as a declaration of intention, purpose, design, goal, or determination to inflict [bodily injury] on another.” United States v. Viefhaus, 168 F.3d 392, 395 (10th Cir.1999); see id. at 396; United States v. Dysart, 705 F.2d 1247, 1256 (10th Cir.1983). And “[i]t is not necessary to show that defendant intended to carry out the threat,” although the threat must be a serious one, “as distinguished from words as mere political argument, idle talk or [973]*973jest.” Viefhaus, 168 F.3d at 395 (internal quotation marks omitted).
The issue on appeal is whether § 875(c) requires proof of an additional element—that the defendant intended the recipient to feel threatened. The statutory language contains no mens rea requirement, but as a statute that criminalizes speech, it “must be interpreted with the commands of the First Amendment clearly in mind. What is a threat must be distinguished from what is constitutionally protected speech.” Watts v. United States, 394 U.S. 705, 707, 89 S.Ct. 1399, 22 L.Ed.2d 664 (1969) (per curiam). Thus, we will read into § 875(c) any scienter necessary to satisfy the demands of the First Amendment.2 “We review questions of constitutional law de novo.” ClearOne Commc’ns, Inc. v. Bowers, 651 F.3d 1200, 1216 (10th Cir.2011) (internal quotation marks omitted).
Defendant contends that the Supreme Court’s opinion in Virginia v. Black, 538 U.S. 343, 123 S.Ct. 1536, 155 L.Ed.2d 535 (2003), compels us to adopt his position. But before we examine that opinion, we turn to, and reject, the claims of both parties that we are bound by circuit precedent to adopt their positions.
The government points to four of our decisions. Two can be disposed of summarily. Viefhaus predated Black. Whatever it said, a circuit precedent cannot bind us to the extent that it is inconsistent with a later Supreme Court decision. See Currier v. Doran, 242 F.3d 905, 912 (10th Cir.2001). And United States v. Wolff, 370 Fed.Appx. 888 (10th Cir.2010), [974]*974like all unpublished decisions of this court, is not binding precedent. See 10th Cir. R. 32.1(A).
The government’s third case, Nielander v. Board of County Commissioners, 582 F.3d 1155 (10th Cir.2009), is a post-Black precedent, but it did not address the issue before us. After successfully defending a state criminal-threat charge, Nielander brought a First Amendment retaliation claim in federal court under 42 U.S.C. § 1983. See id. at 1162-63. The district court granted summary judgment against Nielander on the ground that his statements (the alleged threats) leading to the alleged retaliation were not protected by the Constitution. See id. at 1163. We affirmed. See id. at 1165-66. We held that even if Nielander’s statements were protected speech, the defendants were entitled to qualified immunity. See id. at 1166-69. Two of the defendants had “merely provided] the police with then-account of events,” and had not brought any charges against Nielander. Id. at 1166. As for the defendant police officer, we said that he was entitled to qualified immunity because the law was not clearly established that Nielander’s statements were not true threats. See id. at 1167-69. Although we used a pre-Black, definition of true threat that did not include an intent to instill fear, see id. at 1167, we did not consider whether such intent was required, and we fail to see how our ultimate decision would have been affected if we had included that requirement because the evidence would have supported a reasonable belief that Nielander had that intent. See Brecht v. Abrahamson, 507 U.S. 619, 631, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (“[S]ince we have never squarely addressed the issue, and have at most assumed the [standard], we are free to address the issue on the merits.”).
The fourth case relied on by the government is United States v. Teague, 443 F.3d 1310 (10th Cir.2006). Teague was convicted of transmitting an interstate threat under 18 U.S.C. § 875(c) after he sent a series of threatening e-mails to his former divorce attorney. See id. at 1311-13. His defense at trial was that he had not intended a threat, but that his e-mails were meant instead to prod the attorney to take various actions, were meant humorously, or merely stated the biological truth that the attorney and his family were going to die. See id. at 1312-13. On appeal he argued that the jury should have been instructed that he must have intended that his former attorney feel threatened and not just that “he sent the e-mail with knowledge that a reasonable person would take the message as a threat.” Id. at 1318. But Teague had not raised that argument in district court, so we reviewed only for plain error. See id. Accordingly, he needed to show that the requirement of his additional element was “clear under current law.” Id. at 1319 (internal quotation marks omitted). Yet he conceded (incorrectly, as we shall see) that the Supreme Court had not spoken on the issue, and he acknowledged that this circuit had no controlling precedent and that the other circuits were split. See id. We readily concluded that there was no plain error. See id. The parties’ briefs did not cite Black, nor did our opinion. Teague could hardly stand as this court’s interpretation of Black.
Defendant in turn relies on two of our decisions. One is United States v. Pinson, 542 F.3d 822 (10th Cir.2008), which concerned a prosecution under 18 U.S.C. § 871(a) for threatening the life of the President. See id. at 826. Pinson complained that the jury was instructed that an element of the offense was that he “understood and meant the words mailed as a threat.” Id. at 832 (internal quotation [975]*975marks omitted). By making Ms intent an element, argued Pinson, the prosecution was enabled to put on prejudicial evidence of his state of mind. See id. at 831. We affirmed the conviction. See id. at 827. Some language in the opinion certainly supports Defendant’s argument. We said:
The burden is on the prosecution to show that the defendant understood and meant his words as a threat, and not as a joke, warning, or hyperbolic political argument. But a threat violates the law even if the defendant had no actual intention, or even ability, to carry it out.... The proper question for the jury is whether the defendant meant his words as a threat and whether a reasonable person would so regard them. The instruction here conveyed at least the first element of that meaning. It does not imply that the defendant must be shown to have intended to carry out the threat, but it does require that the defendant understood and meant his words to be a threat.
Id. at 832 (footnote omitted). But the opinion does not bind us in this case. It concerned a prosecution under 18 U.S.C. § 871(a), which, in contrast to § 875(c), explicitly requires that the offense be committed “knowingly and willfully”; and it did not purport to be stating a proposition of constitutional law. Pinson does not cite any authority (much less First Amendment decisions) for its language supporting Defendant, and it does not otherwise explain why (other than the statutory requirement) a defendant must mean his words to be a threat.
The other case relied on by Defendant is United States v. Magleby, 420 F.3d 1136 (10th Cir.2005). Magleby had filed a motion under 28 U.S.C. § 2255 to challenge his convictions of burning and conspiring to burn a cross, in violation of the civil rights of an interracial couple. See id. at 1138-39. The success of one of his arguments turned on whether his appellate counsel had been ineffective for failing to challenge instructions on the elements of his offenses. In the course of the opinion we paraphrased Black as saying that a threat must have been made “with the intent of placing the victim in fear of bodily harm or death. An intent to threaten is enough.” Id. at 1139 (citation and internal quotation marks omitted). But our ruling on Magleby’s claim of ineffective assistance of appellate counsel hinged on the pre-Black law in effect at the time of his earlier appeal. See id. at 1140. Any statement in the opinion regarding the meaning of Black was irrelevant to the opinion and therefore dictum. “[A] panel of this court ... is not bound by a prior panel’s dicta.” Thompson v. Weyerhaeuser Co., 582 F.3d 1125, 1129 (10th Cir.2009) (brackets and internal quotation marks omitted).
Thus, we are facing a question of first impression in this circuit: Does the First Amendment, as construed in Black, require the government to prove in any true-threat prosecution that the defendant intended the recipient to feel threatened? We conclude that it does.
At issue in Black were three state-law convictions for cross burning (or attempted cross burning) with intent to intimidate. See Black, 538 U.S. at 348-51, 123 S.Ct. 1536. On an open field on private property, defendant Black had led a Ku Klux Klan rally that ended with the participants setting flame to a large cross 300 yards or so from a road, where passersby and neighbors could see it. See id. at 348-49, 123 S.Ct. 1536. In a separate incident, defendants Elliott and O’Mara had driven a truck into the yard of an African-American family that had recently moved into the neighborhood, planted a cross 20 feet [976]*976from the house, and set it on fire. Id. at 350, 123 S.Ct. 1536.
The three were convicted of violating a Virginia statute that provided: “It shall be unlawful for any person or persons, with the intent of intimidating any person or group of persons, to burn, or cause to be burned, a cross on the property of another, a highway or other public place,” and, “Any such burning of a cross shall be prima facie evidence of an intent to intimidate a person or group of persons.” Id. at 348, 123 S.Ct. 1536 (internal quotation marks omitted). The Virginia Supreme Court held the statute facially unconstitutional for two reasons: (1) the statute “selectively eho[se] only cross burning because of its distinctive message,” and was therefore a content-based distinction within the category of true threats, which was impermissible under R.A.V. v. City of St. Paul, 505 U.S. 377, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992); and (2) the prima facie provision rendered it overbroad “because the enhanced probability of prosecution under the statute chills the expression of protected speech.” Black, 538 U.S. at 351, 123 S.Ct. 1536 (brackets and internal quotation marks omitted). It did not decide whether the prima facie provision was severable. See id. at 363, 123 S.Ct. 1536.
The United States Supreme Court affirmed in part and reversed in part. We discuss the opinion at some length because, in our view, it has been misconstrued by some courts that we highly respect. Black devotes little attention to what is required for a threat to be a “true threat” not protected by the First Amendment, and its sentence defining true threat (which appears in the Court’s resolution of the R.A.V. issue) is somewhat ambiguous. Nevertheless, a careful review of the opinions of the Justices makes clear that a true threat must be made with the intent to instill fear.
To resolve the R.A.V. issue, the Court, in an opinion by Justice O’Connor for herself and four other Justices, began by reiterating the fundamental protection of speech embodied in the First Amendment. “The hallmark of the protection of free speech is to allow free trade in ideas— even ideas that the overwhelming majority of people might find distasteful or discomforting.” Id. at 358, 123 S.Ct. 1536 (internal quotation marks omitted). But it also noted that the Amendment “permits restrictions upon the content of speech in a few limited areas, which are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.” Id. at 358-59, 123 S.Ct. 1536 (internal quotation marks omitted). As examples of permissible content-based restrictions, it included incitement to imminent breach of the peace, fighting words, and “true threats.” Id. Critical to the resolution of our case, the Court then defined true threats, stating:
True threats encompass those statements where the speaker means to communicate a serious expression of an intent to commit an act of unlawful violence to a particular individual or group of individuals. The speaker need not actually intend to carry out the threat. Rather, a prohibition on true threats protects individuals from the fear of violence and from the disruption that fear engenders, in addition to protecting people from the possibility that the threatened violence will occur. Intimidation in the constitutionally pre-scribable sense of the word is a type of true threat, where a speaker directs a threat to a person or group of persons with the intent of placing the victim in fear of bodily harm or death.
Id. at 359-60, 123 S.Ct. 1536 (emphasis added) (brackets, citations, and internal quotation marks omitted).
[977]*977Turning to the Virginia statute, the Court acknowledged that cross-burning is symbolic expression, and therefore governed by First Amendment principles. See id. at 360, 123 S.Ct. 1536. And it further acknowledged that R.A.V. held that even when a statute bans speech within an unprotected category (such as true threats), it ordinarily cannot discriminate by targeting only a subset of speech within the category. See id. at 361-62, 123 S.Ct. 1536. Thus, the ordinance in R.A.V. was invalidated because it banned “certain symbolic conduct, including cross burning, when done with the knowledge that such conduct would arouse anger, alarm or resentment in others on the basis of race, color, creed, religion or gender,” id. at 361, 123 S.Ct. 1536 (internal quotation marks omitted), but it did not cover, for example, “those who wish to use fighting words in connection with other ideas—to express hostility, for example, on the basis of political affiliation, union membership, or homosexuality,” id. (brackets and internal quotation marks omitted). R.A.V. did not hold, however, that prohibiting only a subset of “a proscribable area of speech” is always barred by the First Amendment. Id. In particular, “the First Amendment permits content discrimination based on the very reasons why the particular class of speech at issue is proscribable.” Id. at 362, 123 S.Ct. 1536 (ellipsis and internal quotation marks omitted). Thus, “[t]he Federal Government can criminalize only those threats of violence that are directed against the President since the reasons why threats of violence are outside the First Amendment have special force when applied to the person of the President.” Id. (ellipses and internal quotation marks omitted). (But the government cannot ban “only those threats against the President that mention his policy on aid to inner cities.” Id. (internal quotation marks omitted).) Accordingly, the Virginia statute’s “ban on cross burning carried out with intent to intimidate [was] fully consistent” with the dictates of R.A.V. because “burning a cross is a particularly virulent form of intimidation.” Id. at 363, 123 S.Ct. 1536.
On the First Amendment overbreadth issue, there was no Court majority. Three Justices did not address the issue; they would have invalidated the Virginia statute under R.A.V. See Black, 538 U.S. at 380-87, 123 S.Ct. 1536 (Souter, J.). Of the remaining six, four (in an opinion by Justice O’Connor for those who formed the RAV.-issue majority except for Justice Scalia) would have held that the statute was rendered facially unconstitutional by the provision that “any such burning of a cross shall be prima facie evidence of an intent to intimidate a person or group of persons.” Id. at 363, 123 S.Ct. 1536 (brackets and internal quotation marks omitted) (O’Connor, J.). In their view, the flaw was that the provision “strip[ped] away the very reason why a State may ban cross burning with the intent to intimidate.” Id. at 365, 123 S.Ct. 1536. The plurality explained:
The prima facie evidence provision permits a jury to convict in every cross-burning case in which defendants exercise their constitutional right not to put on a defense. And even where a defendant like Black presents a defense, the prima facie evidence provision makes it more likely that the jury will find an intent to intimidate regardless of the particular facts of the case. The provision permits the Commonwealth to arrest, prosecute, and convict a person based solely on the fact of cross burning itself.
Id. “[T]he act of cross burning may mean that a person is engaging in eonstititutionally proscribable intimidation,” it said, “[b]ut that same act may mean only that [978]*978the person is engaged in core political speech.” Id. Thus, the prima facie provision “ignore[d] all of the contextual factors that are necessary to decide whether a particular cross burning is intended to intimidate.” Id. at 367, 123 S.Ct. 1536. “The First Amendment,” concluded the plurality, “does not permit such a shortcut.” Id.
Justice Scalia, who was a member of the Court majority on the R.A.V. issue, dissented from the plurality on overbreadth. He disagreed with the plurality’s interpretation of the prima facie provision, contending that it permitted an inference of the requisite intent only if the defendant put on no rebuttal evidence. See id. at 368-71, 123 S.Ct. 1536 (Scalia, J.). And he argued that (1) so construed, the statute was not unconstitutionally overbroad because it would be highly unlikely to lead to any convictions for constitutionally protected conduct; and (2) the plurality should not have assumed its own construction of the prima facie provision but instead should have remanded to the Virginia Supreme Court for a definitive interpretation. See id. at 371-79, 123 S.Ct. 1536. He agreed, however, with setting aside Black’s conviction because the jury had been instructed that it could infer his intent from the act of cross-burning itself, so the jury may have ignored evidence showing no such intent. See id. at 379-80, 123 S.Ct. 1536.
Finally, Justice Thomas would have upheld the statute, saying that cross-burning is not expressive conduct and the prima facie provision established a permissible inference. See id. at 388-400, 123 S.Ct. 1536.
We read Black as establishing that a defendant can be constitutionally convicted of making a true threat only if the defendant intended the recipient of the threat to feel threatened. The majority of the Court said that “ ‘[t]rue threats’ encompass those statements where the speaker means to communicate a serious expression of an intent to commit an act of unlawful violence to a particular individual or group of individuals.” Id. at 359, 123 S.Ct. 1536. When the Court says that the speaker must “mean[] to communicate a serious expression of an intent,” it is requiring more than a purpose to communicate just the threatening words. Id. It is requiring that the speaker want the recipient to believe that the speaker intends to act violently. The point is made again later in the same paragraph when the Court applies the definition to intimidation threats: “Intimidation in the constitutionally proscribable sense of the word is a type of true threat, where a speaker directs a threat to a person or group of persons with the intent of placing the victim in fear of bodily harm or death.” Id. at 360, 123 S.Ct. 1536 (emphasis added).
Moreover, the plurality’s overbreadth analysis is predicated on the understanding that the First Amendment requires the speaker to intend to place the recipient in fear. According to the plurality, at least one First Amendment flaw in the prima facie provision was that a jury could infer an “intent to intimidate” from the act of cross-burning itself. Id. at 363, 123 S.Ct. 1536. The prima facie provision, wrote Justice O’Connor, “does not distinguish between a cross burning done with the purpose of creating anger or resentment and a cross burning done with the purpose of threatening or intimidating a victim.” Id. at 366, 123 S.Ct. 1536. But how could that be a First Amendment problem if the First Amendment is indifferent to whether the speaker had an intent to threaten? The First Amendment overbreadth doctrine does not say simply that laws restricting speech should not prohibit too much speech. It says that laws restricting [979]*979speech should not prohibit too much speech tkat is protected by the First Amendment.
True, Justice O’Connor’s overbreadth analysis was not adopted by a Court majority. But that portion of her opinion did not include an analysis of what the First Amendment requires to convict someone of a true threat. The plurality obviously assumed that the discussion of the R.A.V. issue had already established that an intent to threaten was required. The one Justice (Justice Scalia) who had departed from the rest of the R.AV.-issue majority to dissent on overbreadth in no way challenged the underlying assumption by the plurality that the First Amendment required an intent to threaten. On the contrary, he agreed with the reversal of Black’s conviction on the ground that the instruction at his trial based on the prima facie provision could have led the jury to convict without considering evidence that he had no intent to intimidate. See id. at 379-801, 112 S.Ct. 2538 (Scalia, J.). Thus, the overbreadth discussion confirms our reading of the definition of true threat in the discussion of the R.A.V. issue. We also note that Justice Souter’s opinion for the R.A.V.-issue dissenters also seems to have assumed that intent to instill fear is an element of a true threat required by the First Amendment. In support of the view that the Virginia statute could not survive First Amendment scrutiny, the opinion pointed out that the prima facie provision could encourage juries to convict despite weak evidence of an intent to intimidate. See id. at 384-87, 112 S.Ct. 2538 (Souter, J.).
The Ninth Circuit has read Black as we do. See United States v. Bagdasarian, 652 F.3d 1113, 1116-18 (9th Cir.2011); United States v. Cassel, 408 F.3d 622, 630-33 (9th Cir.2005). It said that a “natural reading” of Black’s definition of true threats “embraces not only the requirement that the communication itself be intentional, but also the requirement that the speaker intend for his language to threaten the victim.” Cassel, 408 F.3d at 631. We also find some support from the Seventh Circuit. In United States v. Parr, 545 F.3d 491, 500 (7th Cir.2008), it wrote: “It is possible that the Court was not attempting a comprehensive redefinition of true threats in Black; the plurality’s discussion of threat doctrine was very brief. It is more likely, however, that an entirely objective definition is no longer tenable.”
Other circuits have declined to read Black as imposing a subjective-intent requirement. See United States v. Clemens, 738 F.3d 1, 9-12 (1st Cir.2013) (on plain-error review); United States v. Elonis, 730 F.3d 321, 327-32 (3rd Cir.2013), cert. granted, — U.S.-, 134 S.Ct. 2819, — L.Ed.2d - (2014); United States v. White, 670 F.3d 498, 506-12 (4th Cir.2012); United States v. Jeffries, 692 F.3d 473, 477-81 (6th Cir.2012), cert. denied, — U.S. -, 134 S.Ct. 59, 187 L.Ed.2d 25 (2013); United States v. Nicklas, 713 F.3d 435, 438-40 (8th Cir.2013); United States v. Martinez, 736 F.3d 981, 986-88 (11th Cir.2013). But the reasons for their conclusions do not persuade us.
We discuss the opinion in Jeffries, 692 F.3d 473, because it expresses the various reasons for rejecting our reading of Black. To begin with, the Sixth Circuit said that Black had no need to impose a subjective-intent requirement because the Virginia statute already required that intent. It wrote:
[Black ] merely applies—it does not innovate—the principle that what is a threat must be distinguished from what is constitutionally protected speech. It says nothing about imposing a subjective standard on other threat-prohibiting statutes, and indeed had no occasion to [980]*980do so: the Virginia law itself required subjective intent. The problem in Black thus did not turn on subjective versus objective standards for construing threats. It turned on overbreadth—that the statute lacked any standard at all. The prima facie evidence provision failed to distinguish true threats from constitutionally protected speech because it ignored all of the contextual factors that are necessary to decide whether a particular cross burning is intended to intimidate, and allowed convictions based solely on the fact of cross burning itself.
Jeffries, 692 F.3d at 479-80 (emphasis added) (brackets, citations, and internal quotation marks omitted); accord Clemens, 738 F.3d at 11; Martinez, 736 F.3d at 986-87; Nicklas, 713 F.3d at 439-40; see Elonis, 730 F.3d at 330. But we do not read the plurality’s overbreadth analysis as Jeffries does. As described above, one of the predicates for the plurality’s overbreadth ruling was the Court’s view that a threat was unprotected by the First Amendment only if the speaker intended to instill fear in the recipient. If the First Amendment does not require subjective intent, how could it invalidate the statute for allowing a jury to find subjective intent on improper or inadequate grounds? See Cassel, 408 F.3d at 631 (Black’s over-breadth analysis made clear that “intent to threaten [is] the sine qua non of a constitutionally punishable threat”); White, 670 F.3d at 523 (Floyd, J., dissenting) (“If the First Amendment did not impose a specific intent requirement, ‘Virginia’s statutory presumption was superfluous to the requirements of the Constitution, and thus incapable of being unconstitutional in the way that the majority understood it.’ ” (quoting Frederick Schauer, Intentions, Conventions, and the First Amendment: The Case of Cross-Burning, 55 Sup.Ct. Rev. 197, 217 (2003))). Why would the First Amendment care how a jury goes about finding an element that is a matter of indifference to the Amendment? We cannot read the plurality’s overbreadth ruling as derived from a lack of standards simpliciter.
Also, Jeffries did not read Black’s definition of true threat to include subjective intent. It said that Black’s language— “ ‘ “True threats” encompass those statements where the speaker means to communicate a serious expression of an intent to commit an act of unlawful violence’ ”— conveys “only that a defendant ‘means to communicate’ when she knowingly says the words.” Jeffries, 692 F.3d at 480 (quoting Black, 538 U.S. at 359, 123 S.Ct. 1536); accord Martinez, 736 F.3d at 987; Elonis, 730 F.3d at 329; White, 670 F.3d at 508-09. Perhaps. There is certainly some ambiguity in the language quoted from Black about what the speaker must intend. But the natural reading is that the speaker intends to convey everything following the phrase means to communicate, see Cassel, 408 F.3d at 631; White, 670 F.3d at 522 (Floyd, J., dissenting), rather than just to convey words that someone else would interpret as a “ ‘serious expression of an intent to commit an act of unlawful violence,’ ” Jeffries, 692 F.3d at 480. And later in the same paragraph of Black two sentences resolve any ambiguity. The sentence immediately after the quote is, “The speaker need not actually intend to carry out the threat.” Black, 538 U.S. at 359-60, 123 S.Ct. 1536. The proposition that the speaker need not intend to carry out the threat is a helpful qualification if there is a requirement that the defendant intend the victim to feel threatened. See White, 670 F.3d at 522 (Floyd J., dissenting). But no such qualification is called for if the preceding sentence means that the only requisite mens rea is that the defendant “knowingly says the words.” Jeffries, 692 F.3d at 480. Once it [981]*981is established that the sole requisite intent is to say the (threatening) words, no reasonable person (juror) would then need to be informed that the defendant need not intend to carry out the threat. If there is no requirement that the defendant intend the victim to feel threatened, it would be bizarre to argue that the defendant must still intend to carry out the threat.
A later sentence in the paragraph is still more definitive about Black’s meaning. It says, “Intimidation in the constitutionally proscribable sense of the word is a type of true threat, where a speaker directs a threat to a person or group of persons with the intent of placing the victim in fear of bodily harm or death.” Black, 538 U.S. at 360, 123 S.Ct. 1536 (emphasis added). The Court was not referring to intimidation as defined by the Virginia statute or even as it might be defined by other statutes, but the meaning that is required by the First Amendment. Jeffries did not dispute that this sentence means that intimidation cannot be proscribed unless the speaker utters the threatening words “with the intent of placing the victim in fear of bodily harm or death.” 692 F.3d at 480 (internal quotation marks omitted). Rather, it disposed of the sentence by saying that it only “shows that intimidation is one ‘type of true threat,’ a reality that does little to inform § 875(c), which prohibits all types of threats to injure a person.” Id.; accord Elonis, 730 F.3d at 329 n. 4 (“[T]his sentence explains when intimidation can be a true threat, and does not define when threatening language is a true threat.”); Martinez, 736 F.3d at 987 (“[T]he general class of true threats does not require such an inquiry.... After all, intimidation is but one type of true threat____[Explicitly requiring subjective intent for one discrete type of true threat makes little sense if the Court intended all true threats to require such intent.”). Yet why should the First Amendment require a subjective intent for intimidation but not other true threats? See White, 670 F.3d at 522 (Floyd, J., dissenting). Nothing in Black so much as hints at a reason for such a distinction. What is it about nonin-timidation threats that makes them so much worse than threats of bodily harm or death that the First Amendment allows them to be prosecuted even when the speaker did not intend to instill fear? One would have thought the opposite—that there should be less First Amendment protection for threats of bodily harm or death. The sentence in Black about “intimidation” is best read as merely applying the propositions stated earlier in the paragraph to the specific statute before the Court.
Jeffries found further support for an objective standard in the rationale for denying First Amendment protection to true threats:
While the First Amendment generally permits individuals to say what they wish, it allows government to “protect[ ] individuals” from the effects of some words—“from the fear of violence, from the disruption that fear engenders, and from the possibility that the threatened violence will occur.” R.A.V., 505 U.S. at 377, 388, 112 S.Ct. 2538; Black, 538 U.S. at 344, 123 S.Ct. 1536. What is excluded from First Amendment protection— threats rooted in their effect on the listener—works well with a test that focuses not on the intent of the speaker but on the effect on a reasonable listener of the speech.
692 F.3d at 480 (citations modified); accord Martinez, 736 F.3d at 987-88; Elonis, 730 F.3d at 329-30. Well said. But to say that the effect on the listener supports a “threat” exception to the freedom of speech does not mean that no other considerations come into play. For example, it may be worth protecting speech that ere-[982]*982ates fear when the speaker intends only to convey a political message. As we understand Black, the Supreme Court has said as much. When the speaker does not intend to instill fear, concern for the effect on the listener must yield.
In short, despite arguments to the contrary, we adhere to the view that Black required the district court in this case to find that Defendant intended to instill fear before it could convict him of violating 18 U.S.C. § 875(c).3
III. CONCLUSION
We REVERSE Defendant’s conviction and REMAND for the district court to determine whether he intended his e-mail to be threatening.