Opinion
LIU, J.Penal Code section 422, subdivision (a) prohibits “willfully threaten[ing] to commit a crime which will result in death or great bodily injury to another person, with the specific intent that the statement... is to be taken as a threat . . . which, on its face and under the circumstances in which it is made, is so unequivocal, unconditional, immediate, and specific as to convey to the person threatened, a gravity of purpose and an immediate prospect of execution of the threat, and thereby causes that person reasonably to be in sustained fear for his or her own safety or for his or her immediate family’s safety.” We granted review to determine whether a defendant who utters words to a victim with a subjective intent to threaten may be convicted of an attempt to violate Penal Code section 422 without proof that the intended threat under the circumstances was sufficient to cause a reasonable person to be in sustained fear. For the reasons below, we hold that the crime of attempted criminal threat requires not only proof of a subjective intent to threaten but also proof that the intended threat under the circumstances was sufficient to cause a reasonable person to be in sustained fear.
I.
Defendant Ben Chandler, Jr., lived around the comer from Jamie Lopez. According to Lopez’s testimony at trial, Chandler drove by Lopez in January of 2009, called her a “bitch,” and made comments indicating that he knew when she was alone. Chandler’s comments scared her. The next day he drove by again and said to her, “ ‘fuck you bitch.’ ” Around this time, Lopez also saw Chandler walk up and down the middle of the street using profanity and laughing at her. In addition, Lopez observed “weird” things happening at her house; she heard the sounds of a tennis ball being bounced off her windows and a pipe being thrown at the front door. One day she saw a large quantity of nails spread out in the street and on her driveway, with the word “fuck” spray painted on the street.
Lopez further testified that on January 29, 2009, she saw Chandler walking up the street holding “an object,” saying, “ ‘Fuck you, bitch. I’m going to kill *512you.’ ” Lopez was scared, and she asked a neighbor to take her children because the neighbor’s house had an alarm. Lopez spent the night at the home of another neighbor, Deborah Alva. Lopez and Alva heard Chandler singing a song with lyrics that included the words “somebody’s watching me.” Lopez called the police. The next day, Lopez was in her car with her two children and Alva’s son, and Chandler approached the car and said words to the effect of “I’m going to kill you bitch.” Lopez was frightened and drove away quickly. She again called the police.
Lopez testified that over time she called the police about seven times because of Chandler’s actions. After the incidents, Lopez locked the bedroom, and her family all slept in one room. She also kept an ax and a bat by the windows, and she thought about buying a gun. She bought a video camera and set it up every night to run while she was sleeping. As a result of these incidents, Lopez moved away after two or three months.
Alva testified that she was also a victim of defendant’s threats. Alva had been friends with Chandler but later had a business dispute with him and claimed he owed her money. On January 29, 2009, Alva heard a disturbance, walked outside, and saw Chandler walking up the street swinging a golf club back and forth. Chandler looked at her and said, “ ‘I’m going to kill you, you fucking bitch.’ ” Alva yelled back, “ ‘Bring it on,’ ” because she did not want Chandler to think that he was intimidating her. She first testified that she was not afraid because she “wasn’t going to show him fear.” However, when asked if “inside” she was afraid, she said, “Oh yeah. I was afraid that he would do something to my car.” She also testified that she was afraid Chandler was “capable of carrying out that threat” “[b]ecause of the drugs.” Alva believed he was under the influence of drugs because of his “ranting and raving.” Alva’s husband called the police. Alva turned on the house lights, slept in the living room, and was unable to sleep for the entire weekend.
Defendant testified in his own defense. He said he had never seen Lopez before. He denied placing any nails in the street or writing graffiti on the street. He denied that he owed Alva money, although he acknowledged that he had a business relationship with her and that she thought he owed her money.
In addition, defendant testified that on the evening of January 29, he was chipping golf balls in his backyard when he noticed a laser light on his chest. He was alarmed because he had been shot at the week before. Chandler thought the light was coming from Alva and a group of people gathered at the top of the street on which she lived. He yelled, “ ‘Stop pointing that f’ing thing at me,’ ” and heard Alva and others laughing. Chandler swung his golf club at a tree, then turned and went inside his house.
*513An amended information filed on June 3, 2011, charged defendant in count one with stalking Lopez (Pen. Code, § 646.9, subd. (a)) and in counts two and three with criminal threats against Lopez and Alva, respectively (id., § 422). (All further statutory references are to the Penal Code.) With respect to counts two and three, the trial court instructed the jury with the standard instruction on making a criminal threat, CALCRIM No. 1300, stating in pertinent part: “(1) The defendant willfully threatened to unlawfully kill or to unlawfully cause great bodily injury to Jamie Lopez, she’s in count two, or Deborah Alva, count three; [¶] . . . [¶] (3) The defendant intended that his statement be understood as a threat and intended that it be communicated to Jamie Lopez or Deborah Alva, and that’s count two and then count three; [¶] (4) The threat was so clear, immediate, unconditional, and specific that it communicated to Jamie Lopez or Deborah Alva the serious intention and immediate prospect that the threat would be carried out; [¶] (5) The threat actually caused Jamie Lopez in count two or Deborah Alva in count three to be in sustained fear for her own safety or for the safety of her immediate family; and [¶] (6) Jamie Lopez, that’s count two, Deborah Alva’s that’s count three, fear was reasonable under the circumstances.”
The trial court also instructed the jury on the lesser included crimes of attempt with the language of CALCRIM No. 460: “[T]he People must prove that, [¶] (1) the defendant took a direct but ineffective step towards committing stalking as to count one or criminal threats in counts two and three; and [¶] (2) the defendant intended to commit stalking, that’s count one, or criminal threats, counts two and three.”
On June 10, 2011, a jury found defendant not guilty of making a criminal threat on counts one through three and guilty on the lesser included crime of attempted criminal threat on counts two and three. The jury could not reach a decision as to the lesser included offense on count one, and the court declared a mistrial. In a bifurcated proceeding, the jury also found true two “strike” priors (§§ 667, subds. (b)-(i), 1170.12) and two prior serious felony conviction enhancements (§ 667, subd. (a)). The trial court later struck one of the enhancements, and defendant was sentenced to 33 years to life in prison.
On appeal, defendant argued that the trial court should have instructed the jury that the crime of attempted criminal threat requires a finding that the “intended threat reasonably could have caused sustained fear” under the circumstances, as the court held in People v. Jackson (2009) 178 Cal.App.4th 590, 599 [100 Cal.Rptr.3d 539] (Jackson). The Court of Appeal below rejected the reasoning of Jackson on the ground that nothing in the relevant statutes, this court’s precedent, or the First Amendment to the United States *514Constitution requires an attempted criminal threat to include such a reasonableness element. The Court of Appeal affirmed the convictions and, after correcting a sentencing error, reduced the sentence to 30 years eight months to life in prison.
In light of the conflict between Jackson and the opinion below, we granted review.
II.
In People v. Toledo (2001) 26 Cal.4th 221 [109 Cal.Rptr.2d 315, 26 P.3d 1051] (Toledo), we held that the law of criminal attempt applies to the offense of criminal threat, so that “under California law, there is a crime of attempted criminal threat.” (Id. at p. 224.) Toledo, our first and only encounter with the crime of attempted criminal threat, provides useful background for framing the issue presented here.
In Toledo, a married couple was engaged in a domestic dispute, during which the husband told his wife, “ ‘You know, death is going to become you tonight. I am going to kill you.’ ” (Toledo, supra, 26 Cal.4th at p. 225.) Shortly thereafter, the husband approached his wife with a pair of scissors and “plunged the scissors toward her neck,” but “stopped . . . inches from [her] skin, and said, ‘You don’t want to die tonight, do you? You’re not worth going to jail for.’ ” (Ibid.) The husband walked away, and the wife went to a neighbor’s apartment, “crying . . . and appearing frightened.” (Ibid.) In statements to police that night, the wife said she “ ‘was afraid’ ” her husband “ ‘was going to kill her.’ ” (Ibid.) But at trial, she denied being fearful of her husband during the incident. The husband was charged with various offenses, including criminal threat in violation of section 422. The jury found him not guilty of criminal threat, but guilty of attempted criminal threat. The defendant challenged his conviction on the ground that no such crime exists in California. (Toledo, at p. 226.)
Toledo began by observing that the Legislature enacted section 422 after this court had found a former version of the statute to be unconstitutionally vague. (Toledo, supra, 26 Cal.4th at pp. 228-229.) Then, after reviewing the law of criminal attempt (§§ 21a, 664), we said the plain language of section 664, which prescribes punishment for “[e]very person who attempts to commit any crime” (italics added), “includes those who attempt to commit the crime of criminal threat set forth in section 422.” (Toledo, at p. 230.) Under section 21a, “a defendant properly may be found guilty of attempted criminal threat whenever, acting with the specific intent to commit the offense of criminal threat, the defendant performs an act that goes beyond mere preparation and indicates that he or she is putting a plan into action.” (26 Cal.4th at p. 230.)
*515This result, we explained, comports with the legislative purpose underlying section 422. The crime of attempted criminal threat encompasses situations where a defendant intends to commit a criminal threat “but is thwarted from completing the crime by some fortuity or unanticipated event.” (Toledo, supra, 26 Cal.4th at p. 232.) “For example, if a defendant takes all steps necessary to perpetrate the completed crime of criminal threat by means of a written threat, but the crime is not completed only because the written threat is intercepted before delivery to the threatened person, the defendant properly may be found guilty of attempted criminal threat. Similarly, if a defendant, with the requisite intent, orally makes a sufficient threat directly to the threatened person, but for some reason the threatened person does not understand the threat, an attempted criminal threat also would occur. Further, if a defendant, again acting with the requisite intent, makes a sufficient threat that is received and understood by the threatened person, but, for whatever reason, the threat does not actually cause the threatened person to be in sustained fear for his or her safety even though, under the circumstances, that person reasonably could have been placed in such fear, the defendant properly may be found to have committed the offense of attempted criminal threat.” (Id. at p. 231.) In such situations, “imposing criminal liability upon the defendant for attempted criminal threat in no way will undermine the legislative purpose of prohibiting threats of the specific nature and severity of those identified in section 422.” (Id. at p. 232.)
We went on to “reject the contention that the crime of attempted criminal threat is unconstitutionally overbroad” in violation of the First Amendment. (Toledo, supra, 26 Cal.4th at p. 233.) Citing the examples above, we said the crime “in most instances” will involve circumstances where “the defendant in fact has engaged in all of the conduct that would support a conviction for criminal threat, but where the crime of criminal threat has not been completed only because of some fortuity outside the defendant’s control or anticipation . . . .” (Id. at p. 234.) Because a defendant’s speech is not constitutionally protected in “the general circumstances to which the crime . . . ordinarily will apply” (id. at p. 233), we reasoned, the crime of attempted criminal threat is not overbroad on its face. Nor was the offense unconstitutional as applied in Toledo, where the evidence might have led the jury to conclude that the defendant’s threat “reasonably could have caused [the victim] to be in sustained fear” even though it did not “actually cause[] [the victim] to be in such fear.” (Id. at p. 235.)
Toledo thus confirmed the existence and constitutional validity of the crime of attempted criminal threat. However, we had no occasion to decide whether the crime requires that the intended threat be objectively threatening. Toledo contemplated that “in most instances” the crime will involve an intended threat that is “sufficient” to cause a reasonable person to be in sustained fear but, for whatever reason, fails to cause the victim to “actually” be in *516sustained fear. (Toledo, supra, 26 Cal.4th at pp. 231, 234.) But Toledo did not purport to identify every possible scenario that could support a conviction of attempted criminal threat. Nor did we address any contention that an intended threat must be sufficient to cause a reasonable person to be in sustained fear. That contention is now squarely before us.
III.
Under the criminal attempt statute, attempted criminal threat requires “a specific intent to commit the crime” of criminal threat “and a direct but ineffectual act done toward its commission.” (§ 21a.) Tracking the language of the criminal threat statute (§ 422), we said in Toledo that the intent required for an attempted criminal threat is a specific intent “to threaten to commit a crime resulting in death or great bodily injury with the further intent that the threat be taken as a threat, under circumstances sufficient to convey to the person threatened a gravity of purpose and an immediate prospect of execution so as to reasonably cause the person to be in sustained fear for his or her own safety or for his or her family’s safety.” (Toledo, supra, 26 Cal.4th at pp. 230-231.)
Neither Toledo nor the language of sections 21a and 422, subdivision (a) makes clear whether an attempted criminal threat requires a specific intent to make a threat that is objectively sufficient to cause the victim reasonably to be in sustained fear or whether it is enough that the offender subjectively intends the threat to be sufficient to cause the victim reasonably to be in sustained fear. The latter is a plausible construction of the criminal attempt statute as applied to section 422. But the former is also plausible on the ground that if reasonableness were purely a matter of subjective belief instead of an objective standard, then a “specific intent” (§ 21a) to cause a “person reasonably to be in sustained fear” (§ 422, subd. (a), italics added) would be no different than a specific intent to simply cause a person to be in sustained fear, with the word “reasonably” omitted. The ambiguity arises from the application of the term “specific intent” in section 21a to the word “reasonably” in section 422. (See Toledo, supra, 26 Cal.4th at p. 232 [“a defendant can be found to have committed the crime of attempted criminal threat only if he or she acts with the specific intent to make the very kind of threat... to which section 422 applies”].) Unlike our dissenting colleague (dis. opn., post, at p. 530), we find no unambiguous meaning here.
In resolving this ambiguity, we first discuss generally applicable principles of the law of attempt. We then examine constitutional concerns that arise specifically from the fact that attempted criminal threat, unlike other attempt crimes, criminalizes a category of speech. We ultimately construe the elements of attempted criminal threat in light of those constitutional concerns.
*517A.
In general, the specific intent required by the law of attempt does not require a showing that the intended act would be effective in completing the target crime. Attempted murder, for example, requires a specific intent to unlawfully kill another human being with malice aforethought. (§§ 187, 664.) If a defendant had tried to kill another person by using poison but misjudged the amount of poison necessary to kill, the defendant’s misapprehension would not be a legal barrier to an attempted murder conviction. The defendant might argue that his use of insufficient poison casts doubt on whether he actually intended to kill, but assuming sufficient evidence to the contrary, a jury could find that the defendant, despite his misapprehension, had a specific intent to kill. A conviction of attempted murder in such circumstances would satisfy the two elements of an attempt: “a specific intent to commit the crime, and a direct but ineffectual act done toward its commission.” (§ 21a.)
Many other examples confirm that a person may be convicted of an attempt to commit a crime he never could have completed under the circumstances. (See, e.g., People v. Rojas (1961) 55 Cal.2d 252, 256-258 [10 Cal.Rptr. 465, 358 P.2d 921] [the defendants could be convicted of attempting to receive stolen property by receiving items that they believed to be stolen but that had actually been recovered by police]; People v. Pham (2011) 192 Cal.App.4th 552, 560-561 [121 Cal.Rptr.3d 458] [“[Defendant cannot escape liability for his attempt to kill [the victims] just because, contrary to his belief, it turned out his intended victims were not where he thought they were. His crimes were complete when, with the intent to kill the two teenagers, he fired shots into a group in which he thought they were.”]; In re Ryan N. (2001) 92 Cal.App.4th 1359, 1380-1384 [112 Cal.Rptr.2d 620] [substantial evidence of an attempt to assist a suicide where the defendant encouraged the victim to take pills that were not actually lethal]; People v. Thompson (1993) 12 Cal.App.4th 195, 202-203 [15 Cal.Rptr.2d 333] [the defendant could be convicted of attempted rape for having sex with a victim he believed was alive but was actually dead]; People v. Reed (1996) 53 Cal.App.4th 389, 396-397 [61 Cal.Rptr.2d 658] [affirming attempted child molestation conviction where the defendant was involved in a sting and there were no actual children].)
Further, we have held that the commission of an attempt does not require proof of any particular element of the completed crime. (See People v. Scott (2011) 52 Cal.4th 452, 488 [129 Cal.Rptr.3d 91, 257 P.3d 703] [attempted rape]; People v. Lindberg (2008) 45 Cal.4th 1, 30 [82 Cal.Rptr.3d 323, 190 P.3d 664] [attempted robbery]; see also People v. Dillon (1983) 34 Cal.3d 441, 454-455 [194 Cal.Rptr. 390, 668 P.2d 697].) Here it is true that the *518Legislature enacted section 422 in order “to describe and limit the type of threat” comprising the completed offense of criminal threat. (Toledo, supra, 26 Cal.4th at p. 229; see id. at pp. 228-229.) But neither the text nor legislative history of section 422 addresses the interplay between the offense of criminal threat and the generally applicable law of attempt.
In addressing the question before us, we leave undisturbed the foregoing principles of the law of attempt, which are well established. Those principles, without more, tend to suggest that a statement that is intended as a threat, but is ineffectual as such, may fall within the scope of an attempted criminal threat. However, we are confronted with an additional consideration that is specific to this case; Attempted criminal threat, unlike other attempt crimes, penalizes speech. The parties have devoted most of their briefing to whether it violates the First Amendment to criminalize speech that, though intended as a threat, is not actually threatening enough under the circumstances to cause a reasonable person to be in sustained fear. We now examine this constitutional issue and its implications for the offense of attempted criminal threat.
B.
At first blush, a statement intended by its speaker as a threat might not immediately call to mind core First Amendment values. But, as the United States Supreme Court has long recognized, the First Amendment “demands that content-based restrictions on speech be presumed invalid [citation] and that the Government bear the burden of showing their constitutionality.” (Ashcroft v. American Civil Liberties Union (2004) 542 U.S. 656, 660 [159 L.Ed.2d 690, 124 S.Ct. 2783].) This principle does not depend on a court’s independent valuation of the suppressed speech: “The First Amendment’s guarantee of free speech does not extend only to categories of speech that survive an ad hoc balancing of relative social costs and benefits. The First Amendment itself reflects a judgment by the American people that the benefits of its restrictions on the Government outweigh the costs. Our Constitution forecloses any attempt to revise that judgment simply on the basis that some speech is not worth it.” (United States v. Stevens (2010) 559 U.S. 460, 470 [176 L.Ed.2d 435, 130 S.Ct. 1577] (Stevens).) Indeed, u[m]ost of what we say to one another lacks ‘religious, political, scientific, educational, journalistic, historical, or artistic value’ (let alone serious value), but it is still sheltered from Government regulation.” (Id. at p. 479.)
Thus, for example, video games in which “ ‘[v]ictims are dismembered, decapitated, disemboweled, set on fire, and chopped into little pieces’ ” retain First Amendment protection. (Brown v. Entertainment Merchants Assn. (2011) 564 U.S. 786, _ [180 L.Ed.2d 708, 131 S.Ct. 2729, 2738].) The high court has also held that virtual child pornography is protected speech. *519(Ashcroft v. Free Speech Coalition (2002) 535 U.S. 234, 250-251 [152 L.Ed.2d 403, 122 S.Ct. 1389] (Free Speech Coalition).) In reaching these conclusions, the high court has underscored that it is not the speaker’s burden to demonstrate why such speech merits protection, but rather the government’s burden to put forward a sufficient justification for excluding the speech from the First Amendment’s protection. (See Brown, at p. _ [131 S.Ct. at p. 2738]; Free Speech Coalition, at p. 246; see also United States v. Playboy Entertainment Group, Inc. (2000) 529 U.S. 803, 816-817 [146 L.Ed.2d 865, 120 S.Ct. 1878] (Playboy).)
Consistent with these principles, the high court has made clear that any statute that “makes criminal a form of pure speech . . . must be interpreted with the commands of the First Amendment clearly in mind.” (Watts v. United States (1969) 394 U.S. 705, 707 [22 L.Ed.2d 664, 89 S.Ct. 1399] (Watts).) In Watts, the defendant said at a public rally on the Washington Monument grounds that he would not respond to his draft notice and that “ ‘If they ever make me carry a rifle the first man I want to get in my sights is L. B. J.’ ” (Id. at p. 706.) Reversing the defendant’s conviction under a statute prohibiting threats against the President, the high court said the defendant’s speech, considered in context, was “political hyperbole” rather than a “true ‘threat’ ” and was therefore constitutionally protected. (Id. at p. 708.)
Citing Watts, the high court in Virginia v. Black (2003) 538 U.S. 343 [155 L.Ed.2d 535, 123 S.Ct. 1536] (Black) said “the First Amendment . . . permits a State to ban a ‘true threat.’ [Citations.] [¶] ‘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. [Citations.] 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.’ [Citation.] 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 pp. 359-360.) Black examined a Virginia statute prohibiting cross burning “ ‘with the intent of intimidating any person or group of persons.’ ” (Id. at p. 348.) While invalidating a portion of the statute that treated the burning of a cross as prima facie evidence of an intent to intimidate (id. at pp. 363-367), the high court held that “Virginia’s statute does not run afoul of the First Amendment insofar as it bans cross burning with intent to intimidate” (id. at p. 362).
The Court of Appeal below implicitly read Blade to mean that “[outlawing speech that is subjectively intended as a threat cannot possibly have any *520chilling effect on protected speech.” But it is arguable whether the high court ruled so broadly. It is true, as the Court of Appeal noted, that “the Virginia law at issue did not require that anybody actually be intimidated, much less that it be reasonable for someone to be intimidated under the circumstances.” However, there was no suggestion in either of the two cases consolidated in Black that the victims of the cross burnings were not actually or reasonably intimidated. Nor would such a contention have had much force given the high court’s recognition that when cross burning is done with an intent to intimidate, it is an especially effective form of intimidation. (See Black, supra, 538 U.S. at p. 357 [“[W]hen a cross burning is used to intimidate, few if any messages are more powerful.”]; ibid. [“[W]hen a cross burning is directed at a particular person not affiliated with the Klan, the burning cross often serves as a message of intimidation, designed to inspire in the victim a fear of bodily harm. Moreover, the history of violence associated with the Klan shows that the possibility of injury or death is not just hypothetical.”]; id. at p. 363 [“burning a cross is a particularly virulent form of intimidation”]; ibid. [referring to “cross burning’s long and pernicious history as a signal of impending violence”].) Because the high court understood that cross burning done with an intent to intimidate is particularly effective at achieving its intended effect, and because there was no dispute in Black as to whether the victims were actually or reasonably put in fear, Black had no occasion to decide whether a subjective intent to threaten is sufficient by itself to establish a true threat, at least outside the context of cross burning.
Nor does United States v. Williams (2008) 553 U.S. 285 [170 L.Ed.2d 650, 128 S.Ct. 1830] (Williams) resolve the issue before us. Williams held that a federal statute generally banning offers to provide or requests to obtain child pornography was not overbroad under the First Amendment. In response to the contention that “it would be unconstitutional to punish someone for mistakenly distributing virtual child pornography as real child pornography,” the high court said: “We disagree. Offers to deal in illegal products or otherwise engage in illegal activity do not acquire First Amendment protection when the offeror is mistaken about the factual predicate of his offer. The pandering and solicitation made unlawful by the Act are sorts of inchoate crimes — acts looking toward the commission of another crime, the delivery of child pornography. As with other inchoate crimes — attempt and conspiracy, for example — impossibility of completing the crime because the facts were not as the defendant believed is not a defense.” (Id. at p. 300.)
Williams also noted that the federal statute at issue targeted “ ‘[a]ny person who . . . [¶] . . . [¶] . . . knowingly . . . [¶] . . . [¶] . . . advertises, promotes, presents, distributes, or solicits . . . any material or purported material in a manner that reflects the belief, or that is intended to cause another to believe, that the material or purported material is, or contains’ ” a visual depiction of a “ ‘minor engaging in sexually explicit conduct.’ ” (Williams, supra, 553 U.S. *521at pp. 289-290.) The statute defined “ ‘ “sexually explicit conduct” ’ ” to include “ ‘lascivious exhibition of the genitals or pubic area of any person.’ ” (Id. at p. 290.) In response to the argument that “the statute could apply to someone who subjectively believes that an innocuous picture of a child is ‘lascivious,’ ” the high court said: “That is not so. The defendant must believe that the picture contains certain material, and that material in fact (and not merely in his estimation) must meet the statutory definition. Where the material at issue is a harmless picture of a child in a bathtub and the defendant, knowing that material, erroneously believes that it constitutes a ‘lascivious exhibition of the genitals,’ the statute has no application.” (Id. at p. 301.)
Notably, this latter reasoning in Williams was not compelled by the statutory text, which refers only to a panderer’s or solicitor’s “ ‘belief’ ” that the material contains depictions of a minor engaged in sexually explicit conduct. (Williams, supra, 553 U.S. at p. 290.) The statute does not contain an objective requirement that the material in fact depict sexually explicit conduct. But the high court read such a requirement into the statute in the course of rejecting First Amendment concerns about the statute’s potential breadth. (Williams, at pp. 300-301.) Thus, we do not believe it is clear from Williams that a defendant’s subjective belief in the illegality of his conduct is, in all instances, sufficient to remove speech-based conduct from First Amendment concern. And we are not aware of any case that has extended the principle that “impossibility of completing the crime because the facts were not as the defendant believed is not a defense” (Williams, at p. 300) to allow punishment of speech that the speaker believed to be threatening but was not objectively so.
Our own cases in this area have held that speech may be punished when it is threatening to a reasonable person, but we have not addressed whether a subjective intent to threaten is alone sufficient to establish a true threat. In Toledo, we held that “the type of threat satisfying the criminal threat provisions of section 422 — that is, a threat ‘to commit a crime which will result in death or great bodily injury to another person . . . which, on its face and under the circumstances in which it is made, is so unequivocal, unconditional, immediate, and specific as to convey to the person threatened, a gravity of purpose and an immediate prospect of execution of the threat’— constitutes speech that falls outside the protection of the First Amendment.” (Toledo, supra, 26 Cal.4th at p. 233.) Toledo relied on In re M.S. (1995) 10 Cal.4th 698 [42 Cal.Rptr.2d 355, 896 P.2d 1365], where we upheld a pair of hate crime statutes and said: “When a reasonable person would foresee that the context and import of the words will cause the listener to believe he or she will be subjected to physical violence, the threat falls outside First Amendment protection.” (Id. at p. 710; see Toledo, at p. 233.)
*522In People v. Lowery (2011) 52 Cal.4th 419 [128 Cal.Rptr.3d 648, 257 P.3d 72] (Lowery), we examined section 140, subdivision (a), which prohibits “willfully” threatening a witness. The defendant argued that his conviction violated the First Amendment “because the statute lacked a specific intent requirement.” (Lowery, at p. 421.) We acknowledged that the statute requires only “ ‘a purpose or willingness to commit the [threatening] act’ ” and does not require a specific intent to threaten. (Id. at p. 427.) But in order “to ensure the constitutionality of section 140(a),” we construed the statute “as applying only to those threatening statements that a reasonable listener would understand, in light of the context and surrounding circumstances, to constitute a true threat, namely, ‘a serious expression of an intent to commit an act of unlawful violence’ [citation], rather than an expression of jest or frustration.” (Ibid.) Lowery thus held that a subjective intent to threaten is not necessary to remove threatening speech from constitutional protection; the objectively threatening character of the speech suffices. (Accord, U.S. v. Elonis (3d Cir. 2013) 730 F.3d 321, cert. granted June 16, 2014, 573 U.S. _.) But Lowery did not address whether a subjective intent to threaten is sufficient to remove speech that is not objectively threatening from constitutional protection.
Five justices in Lowery signed a concurring opinion that explained why U.S. v. Bagdasarian (9th Cir. 2011) 652 F.3d 1113 was “mistaken” in holding that a “ ‘true threat’ ” requires “proof that the speaker subjectively intended the statements to be taken as a threat.” (Lowery, supra, 52 Cal.4th at p. 428 (conc. opn. of Baxter, J.).) The concurring opinion emphasized that “decisions prior to Black ‘ “almost uniformly” ’ applied an objective standard, not a subjective standard, to determine whether a statement was a true threat and thus outside of the protections afforded by the First Amendment. [Citations.]” (Id. at p. 429 (conc. opn. of Baxter, J.) The opinion went on to say: “One might also question the logic of resting the constitutional determination whether speech qualifies as a true threat on the subjective understanding of the speaker, without regard to whether the speech objectively would be viewed as threatening. [Citation.] A statement that is subjectively intended to be a threat but which presents no objective indicators of its threatening nature would not trigger fear in the recipient or cause disruption.” (Id. at p. 432 (conc. opn. of Baxter, J.).)
What was true when we decided Lowery remains true today: The great weight of authority, both before and after Black, has focused on the objective character of a threat as the dividing line between protected speech and an unprotected true threat. (See, e.g., U.S. v. Jeffries (6th Cir. 2012) 692 F.3d 473, 480 [“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.”]; U.S. v. White (4th Cir. 2012) 670 F.3d 498, 511 [“First Amendment principles distinguish protected speech from unprotected speech *523based on an objective view of the speech, not its mens rea.”]; U.S. v. Mabie (8th Cir. 2011) 663 F.3d 322, 332-333; Porter v. Ascension Parish School Bd. (5th Cir. 2004) 393 F.3d 608, 616; U.S. v. Malik (2d Cir. 1994) 16 F.3d 45, 49; U.S. v. Kosma (3d Cir. 1991) 951 F.2d 549, 556-557; but see U.S. v. Bagdasarian, supra, 652 F.3d at pp. 1117-1118 & fn. 14 [reading Black to set forth a subjective test while noting that the statements at issue in Bagdasarian failed to pass either an objective or subjective test].) All of these cases, like Lowery, addressed whether a subjective intent to threaten is necessary, not whether it is sufficient, to establish a true threat. But it is evident from the case law that, if we were to uphold a conviction for attempted criminal threat with no requirement that the defendant’s statement caused a person reasonably to be in sustained fear, our court would be the first to hold that a subjective intent to threaten is sufficient by itself to eliminate First Amendment protection from speech that is not objectively threatening.
The absence of firm authority for such a rule gives us pause, as does its potential breadth. Suppose the defendant in Watts had been deluded as to his own efficacy and had subjectively believed his threat against President Johnson was sufficient to induce reasonable fear, even though it was objectively insufficient. Would it impinge on protected speech to punish his intended threat, even if any reasonable person would have understood it as “political hyperbole” (Watts, supra, 394 U.S. at p. 708)? Or consider the facts of NAACP v. Claiborne Hardware Co. (1982) 458 U.S. 886 [73 L.Ed.2d 1215, 102 S.Ct. 3409] (Claiborne Hardware). In “an impassioned plea for black citizens to unify, to support and respect each other, and to realize the political and economic power available to them” (id. at p. 928), Charles Evers (the older brother of Medgar Evers) said in a speech to several hundred people: “ ‘If we catch any of you going in any of them racist stores, we’re gonna break your damn neck’ ” (id. at p. 902). Could the police have arrested Evers based on probable cause of a subjective intent to threaten, even if it was clear in context that the speech was not objectively threatening? Such exposure would be in tension with the precept that “[t]he government ‘cannot constitutionally premise legislation on the desirability of controlling a person’s private thoughts.’ ” (Free Speech Coalition, supra, 535 U.S. at p. 253.) It would also risk a chilling effect, since “a speaker might still be worried about being prosecuted for a careless . . . statement, even if he does not have the intent required to render him liable.” (United States v. Alvarez (2012) 567 U.S. _, _ [183 L.Ed.2d 574, 132 S.Ct. 2537, 2555] (conc. opn. of Breyer, J.).)
To be sure, defendant’s statements in this case do not resemble the political speech at issue in Watts and Claiborne Hardware. But if we are to posit a category of constitutionally proscribable speech based on a subjective intent to threaten, it is not obvious that the First Amendment permits “the further content discrimination of proscribing only” those statements we deem to have no social value while protecting those statements we deem socially valuable. *524(R. A. V. v. St. Paul (1992) 505 U.S. 377, 384 [120 L.Ed.2d 305, 112 S.Ct. 2538]; see id. at p. 386 [“the power to proscribe [speech] on the basis of one content element . . . does not entail the power to proscribe it on the basis of other content elements”].) The high court has expressly declined to authorize such value-based content discrimination outside the context of obscenity. (See Stevens, supra, 559 U.S. at pp. 479-480.) In evaluating sexual material not alleged to be obscene, the high court has said: “We cannot be influenced . . . by the perception that the regulation in question is not a major one because the speech is not very important. The history of the law of free expression is one of vindication in cases involving speech that many citizens may find shabby, offensive, or even ugly. It follows that all content-based restrictions on speech must give us more than a moment’s pause.” (Playboy, supra, 529 U.S. at p. 826.)
Further, we note that the high court in Black identified three rationales for excluding true threats from the ambit of the First Amendment: protecting people “ ‘from the fear of violence,’ ” protecting people “ ‘from the disruption that fear engenders,’ ” and protecting people “ ‘from the possibility that the threatened violence will occur.’ ” (Black, supra, 538 U.S. at p. 360.) Although all three rationales apply to burning a cross with an intent to intimidate (see ante, at p. 519), it is questionable whether any of them applies to statements that are not objectively threatening. As to the first two rationales, “[a] statement that is subjectively intended to be a threat but which presents no objective indicators of its threatening nature would not trigger fear in the recipient or cause disruption.” (Lowery, supra, 52 Cal.4th at p. 432 (conc. opn. of Baxter, J.).) And as to the third rationale, the possibility that threatened violence will occur is a reason not to protect speech that is objectively threatening. But it is not clear how speech that is not objectively threatening would trigger, except in the most speculative way, a possibility that the threatened violence will occur. (Cf. Free Speech Coalition, supra, 535 U.S. at p. 253 [“The government may not prohibit speech because it increases the chance an unlawful act will be committed ‘at some indefinite future time.’ ”].)
C.
As the discussion above indicates, criminalizing a statement that is intended as a threat but is not objectively threatening raises serious constitutional issues. We need not resolve those issues in this case, however, because “a statute must be construed, if reasonably possible, in a manner that avoids a serious constitutional question.” (People v. Engram (2010) 50 Cal.4th 1131, 1161 [116 Cal.Rptr.3d 762, 240 P.3d 237].)
*525As noted (ante, at p. 516), neither the criminal threat statute (§ 422) nor the criminal attempt statutes (§§ 21a, 664) unambiguously indicate that an attempted criminal threat requires only a subjective intent to threaten, with no objective component. To avoid substantial First Amendment concerns associated with criminalizing speech, we construe the offense of attempted criminal threat to require proof that the defendant had a subjective intent to threaten and that the intended threat under the circumstances was sufficient to cause a reasonable person to be in sustained fear. Accordingly, when a defendant is charged with attempted criminal threat, the jury must be instructed that the offense requires not only that the defendant have an intent to threaten but also that the intended threat be sufficient under the circumstances to cause a reasonable person to be in sustained fear.
Again, we emphasize that our holding is specific to attempts to violate section 422; it does not disturb the generally applicable law of criminal attempt.
IV.
In this case, defendant contends that the trial court’s instructions did not convey to the jury that it had to find the intended threats sufficient to cause reasonable fear under the circumstances. But whether or not the instructions adequately conveyed this element of the offense, reversal is not warranted because any error was harmless beyond a reasonable doubt. (See Neder v. United States (1999) 527 U.S. 1, 8-15 [144 L.Ed.2d 35, 119 S.Ct. 1827]; People v. Cole (2004) 33 Cal.4th 1158, 1208-1209 [17 Cal.Rptr.3d 532, 95 P.3d 811].)
Upon reviewing the record, we conclude that no reasonable juror could have failed to find defendant’s threats sufficient under the circumstances to cause a reasonable person to be in sustained fear. Neither the prosecution nor the defense ever suggested that defendant could be convicted of attempted criminal threat based solely on his subjective intent to threaten. Nor does the evidence suggest that the jury convicted defendant on that basis, since defendant expressly threatened to kill both victims. Moreover, the defense theory at trial did not contest the reasonableness of the victims’ fear. Instead, defendant argued that there was reasonable doubt as to whether he made any of the alleged threats and that the threats, if made, did not cause actual or sustained fear. For example, defense counsel argued in closing, “When you look at this case, you have to look at the credibility of the witnesses. You have to look at whether or not, first of all, did Mr. Chandler even do anything. Did he even say anything. And if he did, what was his intent and what was the effect on the listener. Basically, did it put them in sustained fear? Not just fear for a fleeting second, but sustained fear.” In rebuttal, the *526prosecutor responded, “I’m not really sure what the defense is here. Is it that he didn’t do it, or is it that those victims were not afraid?”
The facts here differ from those in Jackson. There the defendant stood outside the victims’ house while the victims were inside, and one victim testified “she believed that defendant had mentioned both ‘blowing our heads off’ and ‘chopping our heads off.’ ” (Jackson, supra, 178 Cal.App.4th at p. 594.) The Court of Appeal, in reversing the conviction, reasoned that “the jury might have concluded, since [the victims] were safely inside the house with a telephone to call the police while defendant sat out front, or since defendant’s threats were so outlandish, that defendant’s statements could not reasonably have caused the victims to suffer sustained fear.” (Id. at p. 600.) Here, by contrast, Lopez and Alva testified that defendant, a neighbor, made explicit threats that he was going to kill each of them, and defendant made the threats while face-to-face with the victims (and, in Alva’s case, while swinging a golf club) on the street where the victims lived.
In sum, defendant’s threats were sufficient under the circumstances to cause a reasonable person to be in sustained fear — indeed, defendant did not argue otherwise at trial — and no reasonable juror could have concluded otherwise.
CONCLUSION
For the reasons above, we affirm the judgment of the Court of Appeal.
Cantil-Sakauye, C. J., Werdegar, J., and Chin, J., concurred.