Madsen, J.
— Defendant J.M. contends that a conviction of felony harassment pursuant to RCW 9A.46.020(1)(a)(i) requires that the State prove the defendant knew or reasonably should have known that his or her threat to cause bodily injury would be communicated to the proposed victim. Both the trial court and the Court of Appeals disagreed, reasoning that the defendant’s knowledge or intent that the threat be communicated to the proposed victim is irrelevant under the statute. We affirm.
FACTS
On April 28, 1999, approximately one week after the school shootings at Columbine High School in Littleton, Colorado, two Denny Middle School students, 14-year-old S.B. and 13-year-old J.T., were walking home from school.1 Thirteen-year-old J.M., who had recently been suspended from Denny, joined them. J.M. was talking about his suspension and saying that he was mad at Mr. Hashiguchi (the principal at Denny), Mr. Boyd (a Denny administrator in charge of seventh grade discipline), and Mr. Sharper (a district security person). J.M. also talked about what happened at Denver, and said that he and his friend would come to the school and “do that” and then probably move out of state; he said “he wanted to do a shooting at Denny [475] like there was in Colorado.” Verbatim Report of Proceedings (RP) at 20, 38. S.B. responded that “if my little brother or somebody gets shot” then he would know who to come to. RP at 20. J.M. then said, “I’d only kill Mr. Sharper, Mr. Hashiguchi, and Mr. Boyd.” Id. When J.M. said this, he was “excited,” “anxious to talk,” “talking loud like he really wanted to do it,” and walking in front and to the side of the others, socking his hand. RP at 21, 22, 39.
J.T. testified he thought J.M. was “blowing off steam” and “just being a jerk.” RP at 40. When he first heard J.M.’s statements, S.B. did not “think that much of it” and also thought that J.M. was just “blowing off steam,” “being a jerk.” RP at 22, 31. However, later, after he thought about it, he thought “there was some possibility that that might happen.” RP at 22. The next day, a teacher overheard S.B. telling another student about what J.M. had said and asking if the other student thought he would do it; the teacher instructed S.B. to tell his counselor, who then told him to tell Mr. Hashiguchi. Mr. Hashiguchi was “shocked, and surprised and concerned” when told. RP at 45. Mr. Hashiguchi had had about 10 contacts with J.M. over the past school year, and was aware of J.M.’s disciplinary problems at school. He had seen J.M. angry, had known him to be emotional, and had seen him crying and noncompliant. Hashiguchi was afraid for his personal safety after hearing about J.M.’s threat. Hashiguchi had never seen J.M. in possession of any weapons, and the suspension at the time was not for any act of violence. He was unaware of any other threats made by J.M. Hashiguchi also testified that he would be concerned if he “heard through the grapevine a threat made by any student.” RP at 52-53.
Hashiguchi reported the incident to the police, and the State charged J.M. with felony harassment. At J.M.’s adjudicatory hearing, he moved for dismissal on the basis that the State had failed to present evidence that he knew his threat would be communicated to the principal. The court denied the motion, and adjudicated J.M. guilty of [476] felony harassment, based on his threat to kill Mr. Hashiguchi.2 J.M. appealed. The Court of Appeals affirmed, holding that the felony harassment statute does not require that the defendant know or should know his or her threat will be communicated to the threatened person. State v. J.M., 101 Wn. App. 716, 6 P.3d 607 (2000). J.M. sought discretionary review by this court, and the American Civil Liberties Union (ACLU) filed a brief in support of his petition for review.
ANALYSIS
J.M. was convicted of felony harassment under- RCW 9A.46.020(1)(a)(i), (b)3:
(1) A person is guilty of harassment if:
(a) Without lawful authority, the person knowingly threatens:
(i) To cause bodily injury immediately or in the future to the person threatened or to any other person;
....
...and
(b) The person by words or conduct places the person threatened in reasonable fear that the threat will be carried out.
(Emphasis added.) “Threat” is defined as, among other things, “to communicate, directly or indirectly the intent . . . [t]o cause bodily injury in the future to the person threatened or to any other person.” RCW 9A.04.110(25)(a).
J.M. maintains that the word “knowingly” means that the person making the threat must know, or should know, [477] that the threat will be communicated to the person threatened. The Court of Appeals acknowledgéd that “it may seem intuitive that in order to harass someone the perpetrator must intend that the person threatened find out that he or she has been threatened.” J.M., 101 Wn. App. at 726. The Court of Appeals held, however, that the statute does not require that the perpetrator know or intend that the threat be communicated to the person threatened. Instead, the court held, the statute requires that the perpetrator knowingly communicate the threat either directly or indirectly, that the person threatened finds out about the threat, and that words or conduct of the perpetrator places the person threatened in reasonable fear that the threat will be carried out. J.M., 101 Wn. App. at 730.
J.M. raises a number of statutory construction arguments, as well as First Amendment concerns that he urges require his reading of the statute. Because some of the statutory argument depends in part upon the constitutional claim, we address the First Amendment issue first.
As J.M. contends, RCW 9A.46.020 regulates pure speech. “[A] statute ... which makes criminal a form of pure speech, must be interpreted with the commands of the First Amendment clearly in mind.” Watts v. United States, 394 U.S. 705, 707, 89 S. Ct. 1399, 22 L. Ed. 2d 664 (1969); see State v. Williams, 144 Wn.2d 197, 26 P.3d 890 (2001). J.M. contends that in order to constitutionally construe RCW 9A.46.020(1)(a)(i), it must be construed to encompass only “true threats,” and therefore the court must interpret “knowingly” to mean that the person making the threats must know or intend that the threat be communicated to the person threatened.
“True threats” are not protected speech. See, e.g., United States v. Fulmer, 108 F.3d 1486 (1st Cir. 1997); United States v. Kelner, 534 F.2d 1020, 34 A.L.R. Fed 767 (2d Cir. 1976); United States v. Howell, 719 F.2d 1258 (5th Cir. 1983); United States v. Khorrami, 895 F.2d 1186 (7th Cir. 1990); United States v. Orozco-Santillan, 903 F.2d 1262 (9th Cir. 1990); Williams, 144 Wn.2d at 207-08. A “true [478] threat” is a statement made “ ‘ “in a context or under such circumstances wherein a reasonable person would foresee that the statement would be interpreted... as a serious expression of intention to inflict bodily harm upon or to take the life of [another individual].”’” Id. (quoting State v. Knowles, 91 Wn. App. 367, 373, 957 P.2d 797 (1998) (quoting Khorrami, 895 F.2d at 1192)) (alteration in original). A “true threat” “ ‘is a serious one, not uttered in jest, idle talk, or political argument.’ ” State v. Hansen, 122 Wn.2d 712, 717 n.2, 862 P.2d 117 (1993) (quoting Howell, 719 F.2d at 1260). The reasons threats of violence are outside the First Amendment are the “protect [ion of] individuals from the fear of violence, from the disruption that fear engenders, and from the possibility that the threatened violence will occur.” R.A.V. v. City of St. Paul, 505 U.S. 377, 388, 112 S. Ct. 2538, 120 L. Ed. 2d 305 (1992).
This Court recently stated that “Washington’s criminal harassment statute clearly prohibits true threats,” citing and quoting former RCW 9A.46.020(1)(a)(i) (1992). Williams, 144 Wn.2d at 208. In arguing that RCW 9A.46.020 requires that the defendant knows or should know that the threat will be communicated to the threatened person, J.M. relies, however, on a different statement of what constitutes a “true threat” than set forth above. In Orozco-Santillan, 903 F.2d at 1265-66, the court said that a “true threat” is one “where a reasonable person would foresee that the listener will believe he will be subjected to physical violence upon his person.”
Orozco-Santillan is not sound authority for the proposition that only a threat that the threatener knows or should know will be communicated to the victim is unprotected speech. First, the case involved threats made directly to the victims in person and by telephone and the court in Orozco-Santillan was not faced with the issue raised by J.M. in this case. Second, the opinion actually contains more than one formulation of the “true threat” standard, since it also recites the “true threat” standard set forth above. Orozco-Santillan, 903 F.2d at 1265 (“whether a reasonable person [479] would foresee that the statement would be interpreted by those to whom the maker communicates the statement as a serious expression of intent to harm or assault”).4
A number of federal appellate courts have concluded that the defendant need not know or intend that a threat be communicated to the person threatened in order for the threat to constitute a true threat. Many cases involve threats against the President. E.g., United States v. Carrier, 672 F.2d 300, 304 (2d Cir. 1982) (threat to kill the President need not be communicated to the President, and whether words constitute a true threat generally question for the trier of fact); see Howell, 719 F.2d 1258; United States v. Smith, 928 F.2d 740 (6th Cir. 1991); United States v. Manning, 923 F.2d 83 (8th Cir. 1991); United States v. Mitchell, 812 F.2d 1250 (9th Cir. 1987); United States v. Merrill, 746 F.2d 458 (9th Cir. 1984); United States v. Welch, 745 F.2d 614 (10th Cir. 1984); United States v. Callahan, 702 F.2d 964 (11th Cir. 1983). In other contexts, threats communicated to third persons without any indication the defendant knew or intended they would be communicated to the victim have also been held to constitute unprotected true threats. E.g., Kelner, 534 F.2d 1020 (threats to kill Yasser Arafat made to television reporter where there was no evidence that Arafat learned of the threat or that the defendant intended or knew he would learn of the threat); Khorrami, 895 F.2d 1186 (threats to execute Israeli leaders sent to Jewish organization in this country); United States v. Martin, 163 F.3d 1212 (10th Cir. 1998) (threat to murder federal law enforcement officer made to third person).
The conclusions of these courts are consistent with the reasons why threats of violence are not protected speech. An individual who in fact learns of a serious threat of violence, not made in “ ‘jest, idle talk, or political argument,’ ” Hansen, 122 Wn.2d at 717 n.2 (quoting Howell, 719 [480] F.2d at 1260), may well fear violence and endure the disruption that fear engenders, and the possibility still exists that the threatened violence will occur, regardless of whether the speaker intends or knows that the threat will be communicated to the victim. See R.A.V., 505 U.S. at 388.
The weight of authority and the underpinnings of the “true threat” standard do not support J.M.’s position that “knowingly” in RCW 9A.46.020(1)(a) must be read to mean the person making the threat knows or should know the threat will be communicated to the threatened person in order to preserve the statute’s constitutionality.
We next examine whether the language of the statute requires “knowingly” to mean that the person making the threat must know or should know that the threat will be communicated to the victim.
The meaning of a statute is a question of law that is reviewed de novo. Cockle v. Dep’t of Labor & Indus., 142 Wn.2d 801, 807, 16 P.3d 583 (2001). The Court’s fundamental objective in determining what a statute means is to ascertain and carry out the Legislature’s intent. State v. Alvarez, 128 Wn.2d 1, 11, 904 P.2d 754 (1995). If the statute’s meaning is plain on its face, then courts must give effect to its plain meaning as an expression of what the Legislature intended. State v. Chapman, 140 Wn.2d 436, 450, 998 P.2d 282, cert. denied, 531 U.S. 984 (2000). A statute that is clear on its face is not subject to judicial construction. Id.
The word “knowingly” is an adverb, and, as a grammatical matter, an adverb generally modifies the verb or verb phrase with which it is associated. See, e.g., State v. Myles, 127 Wn.2d 807, 813, 903 P.2d 979 (1995) (the word “furtively” is an adverb modifying “carry” in RCW 9.41.250 (“furtively carries”), thus describing the manner in which a dangerous weapon is carried); State v. Warfield, 103 Wn. App. 152, 157, 5 P.3d 1280 (2000) (“knowingly” in “knowingly restrain” in RCW 9A.40.040 is an adverb which modifies the verb “restrain”; “restrain” as defined has four components, thus all four components are modified by [481] “knowingly”). The Court of Appeals here applied this rule, relying on United States v. X-Citement Video, Inc., 513 U.S. 64, 68, 115 S. Ct. 464, 130 L. Ed. 2d 372 (1994) for the proposition that the most natural grammatical reading of the adverb “knowingly” is that it modifies only the verb phrase that it precedes, here, “threatens to cause bodily injury.” J.M., 101 Wn. App. at 725.
“Knowingly” is also a statutorily defined term. RCW 9A.08.010(1)(b)(i) states in relevant part that a person “acts knowingly . . . when ... he [or she] is aware of a fact, facts, or circumstances or result described by a statute defining an offense,” that is, has subjective knowledge. Alternatively, “knowingly” also means that a trier of fact may, but is not required to, infer actual knowledge if a reasonable person in the same circumstances would believe that facts exist which are described by statute as defining an offense. RCW 9A.08.010(1)(b)(ii) (as construed to meet constitutional standards regarding presumptions in State v. Shipp, 93 Wn.2d 510, 516-17, 610 P.2d 1322 (1980) (thus requiring that the jury still find subjective knowledge)).
From these two premises — the grammatical function of adverbs and the definition of “knowingly,” we conclude that the statute’s intent is expressed in its plain language: under RCW 9A.46.020(1)(a)(i), the defendant must subjectively know that he or she is communicating a threat, and must know that the communication he or she imparts directly or indirectly is a threat of intent to cause bodily injury to the person threatened or to another person. That is, “knowingly” modifies “threaten,” and thus relates to each part of the applicable definition of “threat.” See Warfield, 103 Wn. App. at 157. As to communicating the threat, the defendant must be aware that he or she is communicating a threat. Therefore, the individual who writes a threat in a personal diary, or who mutters a threat without awareness that it is heard, or who loudly makes a threat when he or she thinks no one else is around, does not “knowingly threaten.” As to the nature of the threat, whether or not the speaker actually intends to carry out the [482] threat is not relevant. However, the communication must be of the intent to cause bodily injury. Thus, the defendant must be aware that the threat is of such an intent. It must, therefore, be a real or serious threat. Idle talk, joking, or puffery does not constitute a knowing communication of an actual intent to cause bodily injury. Thus, the statute as a whole requires that the perpetrator knowingly threaten to inflict bodily injury by communicating directly or indirectly the intent to inflict bodily injury; the person threatened must find out about the threat although the perpetrator need not know nor should know that the threat will be communicated to the victim; and words or conduct of the perpetrator must place the person threatened in reasonable fear that the threat will be carried out.
J.M. maintains, however, that the Court’s analysis in X-Citement Video, in the end, requires that the word “knowingly” means that the defendant must or should know that the threat will be communicated to the intended victim. After stating its agreement with the Ninth Circuit that the most grammatical reading of the word “knowingly” is as an adverb modifying the following verb, the Court in X-Citement then found for a number of reasons that the rule should not be followed. Essentially, the court engaged in an analysis of whether certain elements of the Protection of Children Against Sexual Exploitation Act, at issue in the case, were “strict liability” elements, or whether the scienter requirement of knowledge which clearly applied to other elements also applied to the elements in question.
In X-Citement, the Court interpreted 18 U.S.C. § 2252, which makes it illegal for any person to “ ‘knowingly transport[] or ship [] . . . any visual depiction, if — (A) the producing of such visual depiction involves the use of a minor engaging in sexually explicit conduct; and (B) such visual depiction is of such conduct.’ ” X-Citement Video, 513 U.S. at 68 (quoting 18 U.S.C. § 2252). The Court held that the term “knowingly,” which clearly applied to the verbs it precedes (“ship” and “transport”), also applied to the phrase “use of a minor.” There were five reasons for the Court’s conclusion: [483] (1) the activities involved, shipping and transporting of magazines and film, were not activities generally subject to regulation, and thus citizens would not expect their conduct to be proscribed; (2) the statute provided for harsh penalties, indicating Congress intended there to be some evil intent; (3) clearly innocent conduct would violate the statute if knowledge of the minor’s age were not required; (4) absent a knowledge requirement, the statute would impinge on free speech; and (5) in light of unclear legislative history, a construction lacking any scienter requirement would raise constitutional doubts.
J.M. maintains that similar concerns are raised in this case. First, J.M. says that there is nothing unlawful about uttering threatening words to others who do not take them seriously, and a threat is only unlawful when conveyed to a person who then is in reasonable fear. This contention misapprehends the Court’s analysis in X-Citement. The question is whether the activity is generally the sort of activity which citizens rightfully do not expect to be regulated. To the contrary, a threat of violence upon another person is not the kind of activity which citizens should expect will be unregulated.
J.M. also urges that the knowledge requirement as construed by the Court of Appeals infringes on First Amendment rights, another consideration in X-Citement Video. As indicated above, however, the First Amendment does not require that the defendant know or should know the threat will be communicated to the victim, and serious threats made in private conversations can be unprotected speech.
J.M. also contends, though, that the Court of Appeals’ reading of RCW 9A.46.020 is not the most grammatical construction of the statute. He points out that the Court in X-Citement Video also observed that under the most natural grammatical reading of 18 U.S.C. § 2252, “knowingly” would not apply to the minority of the performers because the age was set forth in an independent clause separated by interrupting punctuation from the clause in which “knowingly” appears. X-Citement Video, 513 U.S. at 68. J.M. [484] argues that under RCW 9A.46.020(1)(a)(i), language relating to the victim is not in an independent clause, and thus the actor must know he is threatening that person. However, given the definition of “threaten,” which in the context of subsection (1)(a)(i) means a direct or indirect communication of an intent to cause bodily injury to the person threatened or to any other person, the subsection says only that the threatener must communicate the intent to cause bodily injury. One can know he or she is communicating an intent to cause such harm regardless of whether he or she knows that the victim will learn of the threat.
J.M. also argues that the definition of “knowingly” in RCW 9A.08.010(1)(b) means that the defendant must have knowledge of all of the elements of the offense. Initially, this argument is undercut by X-Citement Video, because if a scienter element in one part of a statute necessarily applies to all elements, the Court would not have needed to engage in any analysis to determine if knowingly applied to the age of the minor performers. Moreover, the definition of acting “knowingly” in RCW 9A.08.010(1)(b)(i), that a defendant is aware “of a fact, facts, or circumstances or result described by a statute defining an offense” contemplates that the knowledge requirement may apply to only part of a statute as, at the least, the reference to “result” shows. In addition, the structure of the harassment statute, RCW 9A.46.020(1), indicates the “knowingly” requirement applies to only subsection (1)(a), identifying threats within the statute’s purview, and does not apply to subsection (1)(b), setting forth the requirement that the defendant’s words or conduct places the person threatened in reasonable fear, since “knowingly” appears in the first section, but not in the second. Thus, the “knowingly” requirement in section(1)(a) does not apply to all elements of the crime.
J.M. maintains that the Court of Appeals’ construction of the statute is erroneous because it renders the knowledge requirement superfluous.5 He contends that the fact that a [485] threat involves a “communicated intent,” RCW 9A-.04.110(25)(a) (defining threat), means that a threat is inherently intentional because “one cannot innocently threaten harm to another.” City of Seattle v. Ivan, 71 Wn. App. 145, 157, 856 P.2d 1116 (1993).6 Then, he reasons, because a threat is inherently intentional, the knowledge element must mean something more them the fact defendant knew he was uttering a threat, or else it is superfluous. As the State argues, however, there is no scienter requirement in the definition of “threat.”7
J.M. also reasons that legislative intent is contrary to the Court of Appeals construction of “knowingly.” As noted, RCW 9A.46.010 states the Legislature’s finding that protection of persons from harassment can be accomplished without infringing on constitutionally protected speech. The constitutional arguments have been addressed. RCW 9A.46.010 also states that
[t]he legislature finds that the prevention of serious, personal harassment is an important governmental objective. Toward that end, this chapter is aimed at making unlawful the repeated invasions of a person’s privacy by acts and threats which show a pattern of harassment designed to coerce, intimidate, or humiliate the victim.
(Emphasis added.)
J.M. urges that the emphasized language must be read to mean that the defendant must have knowledge that the [486] threat will reach the victim in order to bring the statute within concerns about words or behavior designed to affect the victim. We were faced with a similar claim in State v. Alvarez, 128 Wn.2d 1, 904 P.2d 754 (1995), where it was argued that the legislative findings in RCW 9A.46.010 relating to repeated invasions of privacy and a pattern of harassment meant that a single threat could not sustain a conviction under RCW 9A.46.020. We rejected the argument, reasoning that the elements of RCW 9A.46.020 unambiguously require only one act for conviction. In a similar vein, RCW 9A.46.020 does not contain any language suggesting that an element of the crime is a “design to coerce, intimidate, or humiliate” the victim. (The statute does require that the victim be placed in reasonable fear by words or conduct of the person making the threat.) Accordingly, as in Alvarez, the findings in RCW 9A.46.020 do not compel an interpretation of RCW 9A.46.020(1) to include such element.
The parties also debate the extent to which the decision in State v. Hansen, 122 Wn.2d 712, 862 P.2d 117 (1993) affects this case. That case involves RCW 9A.72.160(1), intimidating a judge, which provides in part that a person is guilty of the crime if he or she “directs a threat to a judge because of a ruling or decision of the judge in any official proceeding.” “Threat” is defined for purposes of the statute to include direct and indirect communications of intent to use force, as well as direct and indirect communications of intent to cause injury. RCW 9A.72.160(2); see RCW 9A.04.110(25). The court held that the statute does not contain a mens rea requirement that the defendant intend or know that the threat will actually reach the judge. The court reached this conclusion in part because the definition of “threat” includes indirect communications, and partly because the purpose of the statute was determined to be prevention of retaliation against judges, which does not require that the threatened judge be aware of the threat.
J.M. maintains that because the Legislature did not include either “knowingly” or “intentionally” in the intimi[487] dating a judge statute, but did in the harassment statute, it must have intended that a different construction apply insofar as whether the defendant must know or intend that the threat reach the victim. J.M. reads too much into this one difference in the statutes (inclusion of “knowingly” at one place in the harassment statute while neither that word nor “intentionally” appears in the intimidating a judge statute). The two statutes are different in structure, language and purpose.
As to purpose, however, J.M. points out that the purpose of the harassment statute is not to protect the intended victim from actual threatened harm, in contrast to the statute at issue in Hansen which the court determined was intended to protect judges from retaliatory acts.8 Thus, he reasons, while in Hansen it did not matter whether the judge actually heard the threat, the harassment statute specifically requires that the victim reasonably fear that the threat will be carried out. While true, this does not mean that the defendant must know or should know that the threatened person will learn of the threat. The reasonable fear arises (or does not) regardless of whether the defendant has such knowledge. The harassment statute’s plain language does not require such knowledge.
Neither the differences in the two statutes nor our analysis in Hansen supports J.M.’s argument.
J.M. contends that the rule of lenity should be applied because, even if the Court of Appeals’ reading of the statute is reasonable, so is his. This ambiguity should be resolved in the defendant’s favor, he contends.9 The meaning of “knowingly” is clear from the plain language of the statute, [488] however, and the rule of lenity does not apply.
Finally, J.M. submitted a statement of additional authority citing State v. G.S., 104 Wn. App. 643, 17 P.3d 1221 (2001). The decision in G.S. reversed a conviction under RCW 9A.46.020(1)(a)(i). The State did not seek discretionary review of the Court of Appeals decision. The issues in G.S. and those in this case are not the same, and thus the propriety of the holdings in that case is not before us. Nonetheless, lest confusion ensue, we do note that the court in G.S. appears to have equated the person threatened with the person to whom the communication of the threat is made. That conclusion is, of course, at odds with our decision here. Under RCW 9A.46.020(1)(a)(i), the person threatened is generally the victim of the threat, i.e., the person against whom the threat to inflict bodily injury is made. The person to whom the threat is communicated may or may not be the victim of the threat. Here, for example, Mr. Hashiguchi was the person threatened, and J.T. and S.B. were the persons to whom J.M. communicated the threat. The statute also contemplates that a person may be threatened by harm to another. An example that comes readily to mind is a communication of intent to harm the child of the person threatened. Again, however, the person to whom the perpetrator communicates the threat may be someone other than the person threatened.
We hold that RCW 9A.46.020(1)(a)(i) does not require that the person making a threat to cause bodily injury know or should know that the threat will be communicated to the victim of the threat. The Court of Appeals is affirmed.
Alexander, C.J., and Smith, Johnson, Sanders, Ireland, Bridge, Chambers, and Owens, JJ., concur.